Datasets:

gabrielaltay
/

pmcoa

Dataset card Data Studio Files Files and versions Community

Split (2)

train · 4.94M rows

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 PLoS Comput BiolPLoS Comput. BiolpcbiplcbploscompPLoS Computational Biology1553-734X1553-7358Public Library of Science San Francisco, USA 1638929910.1371/journal.pcbi.001007805-PLCB-RA-0183R2plcb-01-07-10Research ArticleNeuroscienceVertebratesSegregation of the Brain into Gray and White Matter: A Design Minimizing Conduction Delays Brain Segregation into Gray/White MatterWen Quan 12Chklovskii Dmitri B 21 Department of Physics and Astronomy, State University of New York at Stony Brook, Stony Brook, New York, United States of America 2 Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, United States of America Friston Karl EditorUniversity College London, United Kingdom To whom correspondence should be addressed. E-mail: [email protected] 2005 30 12 2005 1 7 e781 8 2005 28 11 2005 Copyright: © 2005 Wen and Chklovskii.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 ubiquitous feature of the vertebrate anatomy is the segregation of the brain into white and gray matter. Assuming that evolution maximized brain functionality, what is the reason for such segregation? To answer this question, we posit that brain functionality requires high interconnectivity and short conduction delays. Based on this assumption we searched for the optimal brain architecture by comparing different candidate designs. We found that the optimal design depends on the number of neurons, interneuronal connectivity, and axon diameter. In particular, the requirement to connect neurons with many fast axons drives the segregation of the brain into white and gray matter. These results provide a possible explanation for the structure of various regions of the vertebrate brain, such as the mammalian neocortex and neostriatum, the avian telencephalon, and the spinal cord. Synopsis Vertebrate brains generally contain two kinds of tissue: gray matter and white matter. Gray matter contains local networks of neurons that are wired by dendrites and mostly nonmyelinated local axons. White matter contains long-range axons that implement global communication via often myelinated axons. What is the evolutionary advantage of segregating the brain into white and gray matter rather than intermixing them? In this study, the authors postulate that brain functionality benefits from high synaptic connectivity and short conduction delays—the time required for a signal from one neuron soma to reach another. Using this postulate, they show quantitatively that the existence of many fast, long-range axons drives the segregation of the brain into gray and white matter. The theory not only provides a possible explanation for the structure of various brain regions such as cerebral cortex, neostriatum, and spinal cord, but also makes several testable predictions such as the scaling estimate of the cortical thickness. Citation:Wen Q, Chklovskii DB (2005) Segregation of the brain into gray and white matter: A design minimizing conduction delays. PLoS Comput Biol 1(7): e78. ==== Body Introduction A ubiquitous feature of the vertebrate brain is its segregation into white and gray matter (http://www.brainmuseum.org). Gray matter contains neuron somata, synapses, and local wiring, such as dendrites and mostly nonmyelinated axons. White matter contains global, and in large brains mostly myelinated, axons that implement global communication. What is the evolutionary advantage of such segregation [1]? Networks with the same local and global connectivity could be wired so that global and local connections are finely intermixed. Since such design is not observed, and invoking an evolutionary accident as an explanation has agnostic flavor, we searched for an explanation based on the optimization approach [2–6], which is rooted in the evolutionary theory [7–9]. We started with the assumption that evolution “tinkered” with brain design to maximize its functionality. Brain functionality must benefit from higher synaptic connectivity, because synaptic connections are central for information processing as well as learning and memory, thought to manifest in synaptic modifications [10,11]. However, increasing connectivity requires adding wiring to the network, which comes at a cost. The cost of wiring is due to metabolic energy required for maintenance and conduction [12–15], guidance mechanisms in development [16], conduction time delays and attenuation [17,18], and wiring volume [6]. Two pioneering studies, by Ruppin et al. [19] and Murre and Sturdy [20], have proposed that the segregation of white and gray matter could be a consequence of minimizing the wiring volume. They modeled the brain by a network consisting of local and global connections, which give rise to gray and white matter correspondingly. Although wiring volume minimization is an important factor in the evolution of brain design, their results remain inconclusive because predictions of the volume minimization model for the present problem are not robust and are difficult to compare with empirical observations (see Discussion). In this paper, we adopted the model of connectivity introduced in Ruppin et al. [19] and Murre and Sturdy [20], including local and global connections, but minimized the conduction delay, i.e., the time that takes a signal (such as action potential and/or graded potential) to travel from one neuron's soma to another. To see that high connectivity and short conduction delay are competing requirements, note that adding wiring to the network increases not only its volume, but also the distance between neurons. In turn, this requires longer wiring, which, for the same conduction velocity, introduces longer delays. Longer delays are detrimental because fewer computational steps can be performed within the time frame imposed on animals by the environment, making the brain a less powerful computational machine [12]. We show that the competing requirements for high connectivity and short conduction delay may lead naturally to the observed architecture of vertebrate brain as seen in mammalian neocortex and bird telencephalon. As in any other theoretical analysis, we make several major assumptions. First, given that exact connectivity is not known, we characterized the interneuronal connectivity statistically by requiring a fixed number of connections per neuron. Second, although conduction delays are known to differ between connections, we minimized the mean conduction delay. Finally, it is likely that, in the course of evolution, minimization of conduction delay was accompanied by the increase in connectivity. However, it is not known how to quantify the benefits of increased connectivity in comparison with conduction delay increase. Therefore, we adopted a mathematically sound approach of minimizing conduction delay while keeping network connectivity fixed. To obtain quantitative results, we used two analytical (nonnumerical) tools borrowed from theoretical physics. First, most of the derivations were done using the scaling approach. In this approach, a relationship between variables takes the form of proportionality rather than equality. In other words, numerical factors of order one are ignored. One can manipulate and combine such proportionality relationships and still get an estimate that is correct by an order of magnitude. A long history of successful applications of the scaling approach supports its validity. Second, we used a perturbation theory approach, which is helpful when the exact analytical solution to a problem is unavailable. In this approach, a simpler problem is solved exactly. Then the exact solution is modified to fit the actual problem by taking advantage of the fact that such modification is minor. Again, the long history of this approach supports its validity, as long as the difference between the exactly solvable and the actual problem is characterized by a parameter that is much smaller than one. We present our theory in Results, which is organized into seven sections. In the first, we consider competing requirements between small conduction delays and high connectivity in local circuits. We show that local conduction delay limits the size of the local network with all-to-all potential connectivity to the size of the cortical column. The second section models full brain architecture as a small-world network, which combines high local connectivity with small conduction delay. We derive a simple estimate of conduction delay in global connections as a function of the number of neurons. In the third section, we consider spatially integrating local and global connections. We argue that mixing local and global connections substantially increases local conduction delay, while the global conduction delay may be unaffected. In the fourth section, by minimizing local conduction delay we derive a condition under which white/gray matter segregation reduces conduction time delays. The fifth section gives a necessary condition for the segregated design to be optimal, and an example of such design is given in the sixth section. Finally, the seventh section restates our results in terms of the numbers of neurons, interneuronal connectivity, and axon diameter. Results Conduction Delays Limit the Size of a Highly Connected Network We begin by considering the time delay in the local circuits of neocortex, because their mode of operation—thought to involve recurrent computations [21,22]—seems most sensitive to the detrimental impact of time delay. We derive a scaling relationship between local conduction delay and the number of neurons that can have all-to-all potential connectivity. By assuming that the tolerable delay is on the order of a millisecond, we show that the maximum size of such network is close to that of the cortical column. Local cortical circuits may be viewed as a network of n neurons with all-to-all potential synaptic connectivity, meaning that the axons and dendrites of most neurons come close enough to form a synapse [23–25]. In the following we do not distinguish between axons and dendrites in local circuits, and we refer to them as “local wires.” Mathematical symbols used in this paper are shown in Table 1. The mean conduction delay t in local circuits is given by the average path length between two connected neurons (via potential synapses), ℓ, divided by the conduction velocity, s: Table 1 Mathematical Symbols Used in the Main Text Experimental measurements [26,27] and theoretical arguments [28,29] suggest that conduction velocity, s, scales sublinearly with the diameter, d, of local wires (nonmyelinated axons and dendrites): where β is a constant coefficient and θ is a positive power smaller than one (however, see [30]). By combining Equations 1 and 2, we arrive at the expression for the conduction delay: Equation 3 may give an impression that the conduction delay decreases monotonically with wire diameter d. But this is not necessarily the case because ℓ can be a function of d. The following argument [17] shows that the conduction delay, t, as a function of wire diameter, d, has a minimum (provided 0 < θ < 1), which defines the optimal wire diameter. Given the branching structure of axons and dendrites and uniform distribution of neurons, ℓ can be approximated by the linear size of the network [6], which can be easily estimated in the two limiting cases. In the limit when the wire diameter approaches zero, all the nonwire components (such as synapses) are compressed together and take up the space vacated by shrinking wires. Because the volume of the network approaches the volume of the nonwire components, which is constant, the conduction delay diverges as 1/d θ according to Equation 3 [17]. In the opposite limit when the wire diameter is large, the network volume is determined mostly by the wiring [17]. Because wires run in all directions, they must get longer as they get thicker, and the linear size of the network grows proportionally to the wire diameter. Then, according to Equation 3, the conduction delay increases as d 1-θ. Therefore, conduction delay is minimized by the optimal wire diameter, for which the nonwire occupies a fixed fraction of the neuropil volume [17] (see also the first section in Materials and Methods). As a result, the optimal volume of the network is of the same order as the nonwire volume. Assuming that nonwire consists mostly of synaptic components, such as axonal boutons and spine heads, the optimal network volume is of the same order as the total synaptic volume. Therefore, the local network volume is given by: where vs is the average synapse volume and n is the total number of neurons in the local network. (In a network with all-to-all connectivity, n is also the number of local connections made by a neuron via potential synapses.) For the sake of clarity, we ignore the fact that only a fraction (0.1–0.3) of potential synapses are converted into actual synapses [23]. Such numerical factors are ignored in the equations of the main text of this paper, but can be included straightforwardly (see the first section in Materials and Methods). One consequence of Equation 4 is that the optimal wire diameter is on the same order of magnitude as the synaptic linear size, consistent with anatomical observations [31]: By using Equations 3–5 and assuming θ = 1/2, suggested by the cable theory [28,29], we find that the smallest possible mean conduction delay in local networks is given by As the smallest possible conduction delay grows with the number of neurons in the network, fixing conduction delay imposes a constraint on the maximum size of the network. It seems reasonable to assume that the biggest tolerable conduction delay is on the order of a millisecond, a time scale corresponding to physiological events such as the extent of an action potential and the rise-time of an excitatory postsynaptic potential [32]. This time scale could be dictated by the metabolic costs [33]. If we approximate the synaptic volume at a fraction of a cubic micrometer, and β ~ 1 m/s μm−1/2 [28,34,35], the maximum number of neurons in the all-to-all connected network is on the order of 104. This corresponds to roughly the size of a cortical column, which is then the largest network that can combine all-to-all potential synaptic connectivity with tolerable conduction delay. Small-World Network Combines High Local Connectivity with Small Conduction Delay Human neocortex contains about 1010 neurons—many more than could possibly be wired in an all-to-all fashion with a physiologically tolerable conduction delay. In particular, substituting this neuron number into Equation 6, we find that the delay would be on the order of seconds—clearly too slow. Given that the brain is too large to combine high interconnectivity with short conduction delay [36,37] how can it maintain high functionality? In this section, we consider the architecture of the brain as a whole and show that much shorter global conduction delay can be achieved by sacrificing all-to-all connectivity. Anatomical evidence suggests that the brain maintains short conduction delays by implementing sparse global interconnectivity while preserving high local interconnectivity [31]. Such design resembles the small-world network [38], as has been noticed by several authors [39–42]. In a small-world network, a high degree of clustering (the probability of a connection between two neighbors of one neuron) is combined with a small network diameter (the average number of synapses on the shortest path connecting any two neurons). In a neurobiological context this means a combination of high computational power in local circuits with fast global communication [31,36,37,39,40,42]. Thus it is not surprising that evolution adopted this architecture when the size of the network made all-to-all connectivity impractical [36,39,43–45]. How fast could global connections be? Global conduction delay T in a connection of length L with conduction velocity S is given by Here and below, upper-case letters are reserved for parameters of global connections and lower-case letters for parameters of local connections. In big brains, global axons are mostly myelinated as would be expected, given higher demand on their conduction velocity (unpublished data) [28]. In myelinated axons, conduction velocity S scales linearly with diameter [28,46], D as where B is a proportionality coefficient. Combining Equations 7 and 8, we find that the conduction delay is given by The average length of global connections is given by where V is brain volume. In turn, brain volume can be estimated by adopting the following model. Based on anatomical data [31], we assume that most neurons send one global connection to another local network in the brain. Initially, we ignore the volume occupied by local connections. We denote the number of neurons in the brain as N, which can be much larger than the number of local connections (via potential synapses) per neuron, n. Global connections have length L and diameter D. Thus the total volume of the brain can be approximated as Combining Equations 10 and 11, we find Substituting this expression into Equation 9, we obtain Equation 13 can be used to estimate conduction delay in global axons. By substituting B ~ 5 m/s μm−1 [46,47] and the number of neurons in human neocortex, N ~ 1010, we find that the delay is around 20 ms. Compared with the several-second delay expected in a human brain if it had all-to-all connectivity, this is a significant improvement. For the mouse neocortex, by substituting N ~ 107 we find that the delay is around 0.6 ms. This is much better than the 50-ms delay expected, according to Equation 6, if the mouse cortex had all-to-all connectivity. As these estimates are based on the scaling approach, they are reliable only up to an order of magnitude. Yet, they demonstrate that sparse global connections can be much faster than a fully connected network with a comparable number of neurons. Combining Local and Global Connections Increases Conduction Delays After having considered conduction delays in local and global connections separately, now we are in a position to analyze how they are combined in the brain. Here we argue that the main difficulty in integration arises when introducing global connections into local networks. We adopt a model combining both local and global connections proposed by Ruppin et al. [19] and Murre and Sturdy [20]. In this model, each neuron connects (via potential synapses in our case) with n local neurons and sends a global axon to another arbitrarily chosen local network in the brain. For simplicity, we neglect specificity and assume that local connections are made with nearest n neurons located in a sphere of radius ℓ centered on a given neuron, where ℓ is given by Equation 4. Although local and global connections may be highly specific [22,48–50], this approximation is sufficient to understand brain segregation into white and gray matter. The effect of combining local and global connections on the conduction delays can be analyzed in two steps. First, consider the effect of introducing local connections into the network of global connections. This leads to the swelling of the brain volume beyond that in Equation 11. Thus, global axons must be longer, and Equation 13 gives only the lower bound for global conduction delay (see the second section in Materials and Methods). Yet the increase in the global conduction delay caused by the swelling of network can be offset via speeding up global axons by making them thicker, (Equation 8). We show in the second section in Materials and Methods that the global network can absorb local connections and preserve the required global conduction delay. Second, introduction of global connections into local circuits increases local conduction delay and is impossible to compensate by making local connections thicker (see the third section in Materials and Methods). While conduction velocity depends linearly on the global myelinated axon diameter (Equation 8), it scales sublinearly with the local wire diameter (Equation 2). Thus, the smallest possible mean local conduction delay increases when more global connections are mixed with local connections. To describe this quantitatively, we introduce the ratio of global axon volume that is finely intermixed with local connections to the initial unperturbed gray matter (i.e., total local circuits) volume, λ. When λ is much smaller than one, we can argue that the initial minimum local conduction delay is only slightly affected by the penetration of global connections in the gray matter. As shown in the third section in Materials and Methods, because of intermixing global connections and local connections, the increase in local conduction delay, Δt, is proportional to the ratio λ: where t is conduction delay in unperturbed local circuits given by Equation 6. As before, numerical factors are neglected in the spirit of the scaling estimate. According to our original assumption, brain functionality is maximized when conduction delay is minimized. According to Equation 14, the smallest possible conduction delay in local circuits is achieved when λ = 0, i.e., when global and local connections are fully segregated. But full segregation does not lead to a feasible design, because global connections originate and terminate on neurons in local circuits. Thus, we must find a design that spatially integrates local and global connections. We note that minimization of local and global conduction delays are competing desiderata, as can be illustrated by varying the global axon diameter, D. Increasing D speeds up signal propagation along global connections and, therefore, reduces global conduction delay. Yet, thicker global axons are detrimental for local conduction delay because of an increase in λ (Equation 14). As the relative contributions to functionality of conduction delays in local and global connections are unknown, we searched for the optimal design that minimizes local conduction delay as a function of D. Our analysis begins with considering small values of D, i.e., λ ≪ 1. Comparison of the Homogeneous Design and Designs with Gray and White Matter Segregation In order to determine the optimal design we need to compare local conduction delays in different designs combining gray and white matter. In general, this problem is difficult to solve analytically. Yet, when global connections that are intermixed with the gray matter take less volume than does local, i.e., λ ≪ 1, the perturbation theory approach allows us to compare local conduction delays in homogeneous design (HD), in which gray matter and white matter are finely intermixed, to designs in which gray and white matter are segregated. In HD, local and global connections are finely and uniformly intermixed (Figure 1). Then, according to Equation 14, the relative conduction delay increase due to the penetration of global axons of diameter D in the gray matter is given by Figure 1 Homogeneous Design In HD, local and global connections are uniformly and finely intermixed. Inset shows a typical local network containing local axons (thin gray lines) and dendrites (gray and black tree-like structures), and global axons (thick, light-blue lines spanning the whole circle) that perforate gray matter. When the volume of global axons is small, the linear size of the network can be approximated as G 1/3. where N is the total number of neurons in the network. In this expression, we use Equation 11 for the volume of global connections and the fact that the average length of global axons is given by the linear size of the network, which, for a small λ, is given by the linear size of gray matter, G 1/3. We note that the perturbation approach remains valid while the relative conduction increase in HD is less than one, i.e., ND 2 ≪ G 2/3. Another contribution to the mean local conduction delay comes from the boundary effect. Recall that the model requires each neuron in the gray matter to establish connections with n nearest neighbors. If a neuron is far from the boundary of the gray matter, these connections can be implemented in a sphere of radius ℓ given by Equation 4 (Figure 2). Yet neurons within distance ℓ of the gray matter boundary cannot find n neighbors within the sphere of the same size. Therefore, the radius of the local connections sphere must be expanded to find n nearest neighbors (Figure 2). Figure 2 Boundary Effects in the Gray Matter The red full circle illustrates the local connection sphere of a neuron that does not experience the boundary effect. Neurons near external boundary must inflate their local connection sphere to implement the required local connectivity, as illustrated by thin yellow semicircle. Neurons near white matter tracts penetrating gray matter must also inflate their local connection sphere to implement the required local connectivity, as illustrated by the thick red semicircle. Blue line with arrowhead shows typical routing of global axons. R is the size of gray matter modules, where global and local connections are finely intermixed. Expanding the range of local connections for neurons near the boundary increases average local conduction delay. The fraction of neurons that experience the boundary effect is proportional to the volume within distance ℓ from the boundary. As the boundary area in HD is given by G 2/3, the fraction of affected neurons is given by ℓG 2/3/G ~ ℓ/G 1/3, which is less than one because the linear size of the gray matter G 1/3 ≫ ℓ. Since the relative increase in delay for each neuron in the affected volume is of order one, this expression also gives a relative increase in the average local conduction delay. As this boundary effect is determined by the external boundary, it is independent of the design and can be ignored. Yet, the logic of this calculation will be used in the following to estimate the effect of gray and white matter boundary on local conduction delay. Can segregation of gray and white matter reduce local conduction delay in HD? In HD, global axons are straight and are finely intermixed with the local connections. The contribution of global axons to local conduction delays could be reduced by decreasing the length of global axonal segments within the gray matter, according to Equation 14. Rather than connecting neurons with a straight axon, a typical global axon would go toward the nearest white matter tract (region occupied only by global axons) and travel in the white matter until it is close to the target neuron. Then the axon would leave the white matter and traverse the gray matter toward its target (Figure 2). Such routing may increase the length of global axons, but it would minimize impact on local conduction delays. To calculate the relative local delay increase in the segregated design we estimate the relative volume of global axons in the gray matter, λ. We introduce the mean distance between a neuron and the nearest white matter tract, R, which also gives the linear size of gray matter modules (Figure 2). Then the relative volume of nonfasciculated global axons inside the gray matter in the segregated design is given by Comparing Equation 16 with Equation 15, one can see that segregation may be advantageous to HD if R ≪ G 1/3. In other words, introducing a sufficient number of white matter tracts into the gray matter may reduce the length of nonfasciculated global axonal segments in the gray matter and, hence, the local conduction delay. Although segregation of gray and white matter may reduce local conduction delay, it has a disadvantage compared to HD in that it may induce a larger boundary effect because of the white matter tracts inside the gray matter. This effect is similar to the external boundary effect in HD, but it cannot be ignored, because it is different for different designs. If a neuron is far from the gray and white matter interface, its local connections can be implemented in the sphere of radius ℓ (Equation 4; Figure 2). If a neuron is close to the interface, the white matter occupies part of the sphere, meaning that the local sphere radius ℓ must be expanded so that a neuron can still find its n nearest neighbors (Figure 2). Therefore, whether the segregated design is preferred or not depends on whether the relative local conduction delay increase through the boundary effect is much smaller than the local delay increase in HD (Equation 15). To evaluate the mean local conduction delay increase through the boundary effect in the segregated design, we need to specify the geometry of the white matter tracts, because the boundary effect generally depends on the surface area of the tracts. For a typical tract that spans the whole brain (i.e., has length L), we can relate its minimal surface area At to its cross-sectional area, Φ: In turn, the cross-sectional area of a tract depends on the global axon diameter D, and one may conjecture that whether the segregated designs are advantageous or not depends on D. Indeed, we can formulate the following theorem, which is valid to the first order of ND 2/G 2/3 (Equation 15) and while our perturbation approach is valid (i.e., provided ND 2 ≪ G/ℓ, as will be shown later). Theorem 1. In the regime ND2 ≪ ℓ2, local conduction delays in the optimal segregated design and HD are equivalent. In the regime ND2 ≫ ℓ2, there is at least one segregated design with local delays less than those in HD. To prove the first part of the theorem, we calculate the local conduction delay through the boundary effect in the segregated designs and compare it with HD. The length of the global tract segment inside the local sphere is ℓ. The other two dimensions of global tracts are much less than ℓ (Figure 3A), as the minimal boundary effect is achieved by the minimal surface area in Equation 17. Since the total cross-sectional area of the global tracts is ND 2 ≪ ℓ2, each tract's cross-sectional area, Φi, is much less than the cross-sectional area of the local connection sphere (Figure 3A). Inclusion of such a tract into a local sphere increases its radius to (ℓ2 + Φi)1/2. Then, the relative increase in the local conduction delay for neurons in that sphere is [(ℓ2 + Φi)1/2 − ℓ]/ℓ ≃ Φi /ℓ2 ≪ 1. Figure 3 Boundary Effect Induced by White Matter Tracts with Different Cross-Sectional Areas (A) In the case Φ ≪ ℓ2, two dimensions of the white matter tracts (shown in white) can be much smaller than ℓ. Red circle illustrates local connection sphere of a neuron. (B) In the case Φ ≫ ℓ2, neurons within distance ℓ from the white matter tract experience the boundary effect. Now we add up conduction delays contributed by all the tracts to neurons in affected spheres. As the number of spheres affected by one tract is given by L/ℓ, the fraction of neurons experiencing the boundary effect induced by one tract is given by ℓ2 L/G, and the relative local conduction delay increase is given by (Φi/ℓ2)ℓ2 L/G ~ Φi/G 2/3. The total relative increase in local delay is the sum of the boundary effects induced by different tracts, Notice that even if there are multiple tracts within the local connection sphere (i.e., the sphere radius can be larger than ℓ), the above result is still correct. By comparing local conduction delay increase for segregated designs (Equation 18) with that for HD (Equation 15), one can see that they are the same. Therefore, when ND 2 ≪ ℓ2, the optimal segregated designs and HD are equivalent to the first order of ND 2/G 2/3. To prove the second part of the theorem (the ND 2 ≫ ℓ2 regime), we specify a segregated design with smaller local delays than that in HD. In such a design, global axons belong to M (M ≫ 1) tracts with cross-sectional area Φ ≫ ℓ2 each and length L ~ G 1/3. The distance between two tracts is much larger than ℓ. Then, the total affected neuropil volume through the boundary effect is the product of the total surface area of the tracts, MΦ1/2 G 1/3, and ℓ. For a typical neuron within the affected volume, a fraction of its local connection sphere with volume ~ℓ3 is occupied by the white matter tract, as illustrated in Figure 3B. To implement the required local connectivity, the local sphere radius ℓ should expand by a numerical factor of order one. Next, we add up the relative local delay increase induced by all global tracts affecting all the neurons in a volume, given by ℓMΦ1/2 G 1/3/G. Because the total cross-sectional area MΦ ~ ND 2, the relative local delay increase is By comparing relative conduction delay in segregated design (Equation 19) with that in HD (Equation 15), one can see that because Φ ≫ ℓ2 as specified, segregated design is advantageous in the regime ND 2 ≪ G 2/3. Although in the regime ND 2 ≫ G 2/3 we do not have a closed-form expression for the local conduction delay in HD, we can still show that it has longer conduction delays than the segregated design. We show in the third section in Materials and Methods that local conduction delay in HD is a monotonically increasing function of λ, and hence is a monotonically increasing function of ND 2. Thus, the relative delay increase in HD exceeds one when ND 2 ≫ G 2/3. Yet, in the regime ND 2 ≫ G 2/3, the relative local delay increase in a segregated design can still be much smaller than one. To prove this, we note that in a segregated design, the local conduction delay increase because of the nonfasciculated global axons intermixed with gray matter, i.e., λ ~ ND 2 R/G (Equation 16), can be much smaller than one, if R ≪ G 1/3. In addition, the relative local delay increase through the boundary effect can also be much smaller than one. To see this, we specify the tracts in such a way that the total surface area of the white matter tracts is the surface area of the gray matter G/R. Then, using an analysis similar to that illustrated in Figure 3B, the relative local delay increase through the boundary effect is given by ℓG/(RG), which can be much smaller than one if ℓ/R ≪ 1. We note that λ ≪ 1 and R ≫ ℓ could both be satisfied if ND 2 ≪ G/ℓ. Thus, when ND 2 ≫ G 2/3 and ND 2 ≪ G/ℓ, there is at least one segregated design with a local delay less than that in HD. Having considered both the ND 2 ≪ G 2/3 regime and ND 2 ≫ G 2/3 regime, we have proven the second part the theorem. Optimality Condition for Segregated Designs In the previous section, we showed that in the regime ND 2 ≫ ℓ2, there is at least one segregated design with local conduction delay shorter than that in HD. However, we did not specify which design is the optimal one. In this section, we give a necessary condition for a segregated design to be optimal in the regime ND 2 ≫ ℓ2 and ND 2 ≪ G/ℓ. As the advantage of segregation becomes apparent when the total cross-section of global axons ND 2 ~ ℓ2, it is natural to expect that a similar condition defines the optimal gray matter module size R 0, which minimizes local conduction delays. In other words, the number of neurons in the gray matter module is such that the total cross-sectional area of their global axons is given by ℓ2. As the number of neurons in the sphere of radius R 0 is ℓ2/D 2 and the number of neurons in the sphere of radius ℓ is n, we have Thus, we can formulate the following theorem: Theorem 2. In the regime ND2 ≫ ℓ2 and ND2 ≪ G/ℓ, the minimum local conduction delay is achieved by the segregated design with the gray matter module containing ℓ2/D2 neurons. To prove this theorem, we consider designs with gray matter module size smaller and greater than R 0, and show that they have a local conduction delay greater than that in the design with module size R 0. In the case R 0 ≪ R, by applying Theorem 1 to any module one can see that converting that module from HD to segregated designs can reduce local conduction delay. For example, fasciculating global axons within that module into multiple tracts would reduce local conduction delay. In the other case, if modules with size R 0 contain only global axons from the neurons inside the module, by applying Theorem 1 one can see that any optimal segregated designs containing modules with size R ≪ R 0 is equivalent to designs containing modules with size R 0. Moreover, if the tracts inside the module of size R 0 contain external global axons (i.e., global axons that do not belong to the neurons inside the module with size R 0 and/or do not innervate the neurons inside the module), converting segregated designs with module size R ≪ R 0 to designs with module size R 0 reduces the local conduction delay. This happens because merging all the tracts within the module of size R 0 into one reduces the boundary effect. To see this, note that the minimal surface area of the big tract inside the module with size R 0 is on the order of (∑Φi)1/2 R 0 ≪ ∑(Φi 1/2)R 0, where Φi is the mean cross-sectional area of a small tract containing external global axons, and ∑(Φi 1/2)R 0 is the total surface area of the smaller tracts inside the module with size R 0. Even if the tracts run in different directions, most of the tracts can be merged together at the scale R 0, because the typical length of a tract is much greater than that, and a small curvature would not affect the total length by an order of magnitude. Taken together, by considering the two possible cases, we have proven that the minimum conduction delay in segregated designs is achieved with module size R 0. Such designs may be further classified by the relative dimensions of the gray matter. The total boundary area between gray and white matter (i.e., the total surface area of the white matter tracts), A, could satisfy either A ~ G/R 0 or A ≪ G/R 0. As the local conduction delay through the boundary effect grows with A, the latter design has the shorter delay. In the following, we call segregated designs satisfying A ≪ G/R 0 the perforated design (PD). Branching Pipe Design—An Example of Perforated Design In the previous section, we have shown that in the optimal segregated designs, the size of the module, in which global and local connections are finely intermixed, is given by R 0. However, Theorem 2 does not specify other dimensions of the segregated design, such as the total surface area of the white matter tracts. In this section, by considering a specific example, which we name the branching pipe design, we show that the condition A ≪ G/R 0 can be satisfied in the regime in which our perturbation approach is valid. In other words, we prove that PD exists in the regime ND 2 ≪ G/ℓ. We specify the branching pipe design as follows (Figure 4). Global axons belong to several cylindrical white matter pipes perforating the gray matter. Higher-order branches split off lower-order pipes at regular intervals. Different order branches have different lengths and different pipe diameter. The length of the zeroth-order branches (i.e., the main pipes) is given by the linear size of the brain. The length of k + 1st-order branches is given by the interpipe distance among the kth order branches, forming a space-filling structure. The interpipe distance among the finest branches is given by R 0 in Equation 20 (Figure 4). Figure 4 Branching Pipe Design Schematic illustration of branching pipe design with three orders of branches. The distance between kth order branches determines the length of the k + 1st-order branches. The distance between highest-order branches is given by R 0. Although we can calculate the total surface area of the branching pipes for any given order k (as discussed in the fourth section in Materials and Methods), for simplicity we present the main results from the branching pipe design in which only first-order branches exist. We minimize the total surface area of such branching pipes and the local conduction delay by searching for the optimal length and the diameter of the first-order branches and the optimal diameter of zeroth-order branches. We find that the expression for the minimal total surface area of the first-order branching pipes A depends on whether the total white matter volume is greater than the total gray matter volume or not. In the regime ℓ2 ≪ ND 2 ≪ G 2/3, the gray matter occupies most of the brain volume, and A is calculated (see the fourth section in Materials and Methods) as: In turn, λ can be found by substituting G ~ (N/n) ℓ3 and optimal R ~ R 0 (Equation 20) into Equation 16: Then the minimal local conduction delay is given by This dependence of Δt/t on ND 2 is plotted on log-log scale in Figure 5 (represented by the thick blue line). Figure 5 Local Conduction Delay As a Function of Global Axon Diameter in HD and PD Local conduction delay is calculated for specific values ℓ = 0.5 mm, N = 108, and G = 103 mm3 and plotted in log-log coordinates. Thin red line, local conduction delay in HD; thick blue line, local conduction delay in PD. Delay in PD is calculated for the branching pipe design containing only first-order branches. In the regime G 2/3 ≪ ND 2 ≪ G/ℓ, white matter occupies most of the volume, and the specified segregated design has a different appearance: The gray matter is confined to a thin sheet. Sheet thickness is given by the length of the highest-order branches. Then, the minimum surface area of the branching pipes (as calculated in the fourth section in Materials and Methods) is given by In this regime, the minimal local conduction delay is given by (Figure 5) As λ ≪ 1 is equivalent to ND 2 ≪ G/ℓ (to see this, substitute G ~ (N/n)ℓ3 into ND 2 ≪ G/ℓ and compare it with Equation 22), we show that for such a branching pipe design, A ≪ G/R 0 in the regime where our perturbation approach is valid. In other words, we verify the existence of PD in the regime ND 2 ≪ G/ℓ. We note that when λ is approaching one, according to Equations 20 and 22, R 0 2 ~ ℓ2 ~ nD 2, meaning that the total surface area of the gray matter with size ℓ is taken up by the global axons. Therefore, when λ → 1, we must have A ~ G/R 0 ~ G/ℓ ~ ND 2. This can also be seen from the expressions for A in the branching pipe design, i.e., Equations 21 and 24. Moreover, λ ~ 1 (i.e., ND 2 ~ G/ℓ) is when our perturbation approach to calculating the local conduction delay in PD breaks down (Figure 5). When ND 2 ≫ G/ℓ, i.e., λ ≫ 1, we may consider clusters with discrete spatial arrangement, and each cluster has n neurons to implement local connectivity. In this case, we can estimate the lower limit of the cluster size, given by n 1/2 D, assuming that cluster volume is filled by tightly packed global axons. Because of local connections, the actual cluster size must be even greater. Alternatively, clusters may abut each other to form a sheet, and the sheet thickness could be much smaller than ℓ. In this case, however, we cannot determine the necessary conditions for the design to be optimal. Fortunately, existing anatomical data suggest that actual brains are not even close to the regime where λ ≫ 1, as will be shown later. Phase Diagram of Optimal Designs In previous sections we derived conditions under which various designs are optimal in terms of minimizing conduction delays. Specifically, HD is optimal if ND 2 ≪ ℓ2 and PD is optimal if ND 2 ≫ ℓ2 and λ ≪ 1. Next, we illustrate these results on a phase diagram (Figure 6) in terms of basic network parameters such as the local wire diameter d, the number of local connections (via potential synapses) per neuron n, global axon diameter D, and the total number of neurons in the brain N. To obtain the phase diagram, in the first-order perturbation theory, we substitute the expression for ℓ (Equations 4 and 5) into ND 2 ≫ ℓ2, and find that PD is optimal when (N/n)1/2 D/n 1/6 d ≫ 1. In the linear-log space of Figure 6, this expression corresponds to the regime above the thick green line. Figure 6 Phase Diagram of Optimal Designs In this phase diagram, we show parameter regimes in which HD or PD are optimal in terms of the global axon diameter D, local wire diameter d, total neuron number N, and the number of local connections per neuron n. We assume n = 104 and d = 1 μm for all empirical data points. Values of D in mammalian brains are from S. S. H. Wang (personal communication) and [60], and values of N in the neocortex are from [44]. Value of N in rat neostriatum is from [62]. For birds, we assume N = 107. Next, we estimate where perturbation theory fails by setting λ to one. By substituting Equations 4 and 5 into the expression for λ (Equation 22), we find that λ can be rewritten as Then condition λ ~ 1 is equivalent to n 1/6 d/D ~ 1, corresponding to the thin red line in Figure 6. Discussion We have shown that the segregation of the brain into gray and white matter may be a natural consequence of minimizing conduction delay in a highly interconnected neuronal network. We related the optimal brain design to the basic parameters of the network, such as the numbers of neurons and connections between them, as well as wire diameters. Although we do not know whether competing desiderata of short time delay and high interconnectivity were crucial factors driving evolution of vertebrate brains, our theory makes testable predictions. Below, we compare these predictions with known anatomical facts. Scaling Estimate of the Cortical Thickness As fasciculated fibers are usually not observed in neocortical gray matter (according to Nissl and myelin stains), we identify cortical thickness with gray matter module size, R. Our prediction for the optimal module size R 0 (Equation 20) can be rewritten by using Equations 4 and 5 Using n ~ 104 [22,31], d ~ 1 μm [31], and D ~ 1 μm [31] (also measured in the corpus callosum of macaque monkey; S. S.-H Wang, personal communication), we predict cortical thickness R 0 ~ 1 mm. This estimate agrees well with the existing anatomical data [45,51,52], despite being derived using scaling. By substituting these values into Equation 26, we find that λ is smaller than one, justifying our perturbation theory approach. Next, we apply our results to the allometric scaling relationship between cortical thickness, R 0, and brain volume, V. We assume that n and D both increase with brain size [8,39,40] according to the following power laws: n ~ V 1/3 [8,39,40,44] and D ~ V 1/6 (see the fifth section in Materials and Methods). Then, by using Equation 27 and the constancy of the optimal local wire diameter d across different species [31], we predict that R 0 ~ V 4/27. This prediction agrees well with the empirically obtained power law relationship (with exponent 1/9) between cortical thickness and brain volume [39,45,51–53]. Thus, our theory explains why the cortical thickness changes little while brain volume varies by several orders of magnitude between different species. Two previous studies [39,53] also discussed the nature of the scaling law between cortical thickness and brain volume. One study [39] relies on the assumption that the number of neurons in a module of the neocortex is constant. The volume of the module might be cubic R 0. Because the neuronal density may scale inversely as the cubic root of brain volume (see the fifth section in Materials and Methods), R 0 should scale as one-ninth of the brain volume to ensure that the number of neurons in a module is independent of brain volume. The other study [53] relies on the assumption that the number of such modules scales as two-thirds of the total gray matter volume. Hence, the volume of the module scales as one third of the gray matter volume. As the total cortical gray matter volume may scale linearly with the brain volume (see the fifth section in Materials and Methods), the size of the module scales as one-ninth of the brain volume. In this paper, we take a different approach by deriving the expression for the cortical thickness based on the optimization principle. However, we obtain a scaling exponent close to, but not exactly equal to, one-ninth. Comparison of the Cortical Structure and PD Neocortex has a sheet-like appearance, and the total area of the gray and white matter boundary is given by A ~ G/R 0, where G is the total gray matter volume. According to our theory, such design is optimal when λ becomes close to one, which may be the case in big brains. Cortical convolutions may correspond to the geometry expected in the pipe design. However, when λ ≪ 1, our theory predicts that the optimal design satisfies A ≪ G/R 0. This prediction does not seem to be consistent with empirical observations from small brains, such as the smooth, sheet-like mouse cortex. It would be interesting to know whether different requirements on connectivity or other developmental and/or functional constraints could resolve this discrepancy. Comparison of Mammalian Neostriatum and PD Neostriatum is named for its striated appearance (in Nissl- and myelin-stained material [54,55]) caused by axons of neostriatal neurons gathering into fiber fascicles and perforating the gray matter [56]. Areas with higher cell density, or lower global fiber density (myelin-poor [54,55]), are called striosomes or patches [57–59]. Because this structure resembles PD, we identify patch size with R 0 (Equation 27). In a typical rodent (rat or mouse) neostriatum, each principal neuron may locally contact thousands other neurons [56]. Taking n ~ 103, d ~ 1 μm [31], and D ~ 0.6 μm [60], we estimate that R 0 ~ 300 μm. This estimate agrees well with existing anatomical data [61]. In addition, we may estimate the average axonal fascicle size. Given the total number of neurons in the rat neostriatum is about 106 [62], we find that the fascicle diameter is of the same order of ℓ, approximately 100 μm (see Equation 58 in the fourth section in Materials and Methods). This estimate agrees well with fascicle size [55] (see also http://www.hms.harvard.edu/research/brain/atlas.html). Comparison of the Avian Telencephalon and PD Bird brains also exhibit segregation into gray and white matter and may resemble PD. Distinct fiber fascicles have been identified that connect different brain regions (see http://avianbrain.org/boundaries.html), such as the connections from HVC to RA in songbirds, which are presumably myelinated axons [63]. Interestingly, unlike in mammals, which have a large cortex on the top of other brain structures, in birds the white matter fascicles can be scattered throughout the whole forebrain. However, more precise data would be desirable, such as measurements of large-scale myelin distribution in serial sections of bird telencephalons. Comparison of the Spinal Cord and PD While the inner core of the spinal cord contains gray matter, the outer shell contains the white matter consisting of long axons from spinal and cortical neurons [56]. According to our theory, such organization is optimal if the inner core diameter is on the same order as R 0. To see if this is the case, note that a principal (motor) neuron in the spinal cord has a very large arbor span [56,64] and may receive 105–106 potential connections. Given n ~ 105, d ~ 1 μm, and D ~ 1 μm, we find R 0 ~ 8 mm according to Equation 27, which is on the same order as the inner core diameter [56]. Related Work Our work builds upon several insights from recent studies. In particular, the idea of minimizing conduction delay has been used to explain why axons and dendrites take a certain fraction of the neuropil [17]. The main result in that paper is further extended in this study to show that local conduction delay must increase after mixing gray and white matter (see Materials and Methods). Also, in our model local circuits are approximated by the network with all-to-all connectivity, which relies on the concept of potential synapses [23]. Adopting this model allowed us to derive explicit results for the total length of local connections (see the first section in Materials and Methods) [6]. We benefited from several previous studies of anatomical and functional connectivity between different cortical areas. These studies helped conceptualize network connectivity by revealing many interesting features of the network [65–71], such as hierarchal [72], clustering [73], and small-world properties [41,74], which helped to generate new models to address functional specialization and integration [75–80]. We adopted (with the potential synapse caveat) the connectivity model used by Ruppin et al. [19] and Murre and Sturdy [20]. These authors applied the wiring optimization approach to explain the segregation of white and gray matter in the brain. Given a network with local and global connections, they searched for a design having minimum total wiring volume. They attempted to show that a segregated cortex-like design has a smaller volume than does a homogeneous structure. Murre and Sturdy [20] used the scaling approach to calculate network volume for several network connectivity patterns and layouts. We verified their calculation of the interior (homogeneous) structure volume. However, their calculation of the external (cortex-like) structure volume does not seem to be self-consistent. The volume of axons in the external structure was calculated by using the expression that was unjustifiably adapted from the internal structure calculation, thus undermining their conclusion. Ruppin et al. [19] did not rely on scaling arguments and calculated the volume of brain structures given their geometric characteristics, under reasonable assumptions of connectivity parameters. These authors showed that segregation of the network into the inner core organization, which has an inner core of gray matter surrounded by white matter, does not lead to volume efficiency compared to a homogeneous structure. They also showed that the external sheet (cortex-like) structure has a smaller volume than the inner core organization. However, this does not prove that the cortex-like structure has a smaller volume than the homogeneous structure, a conclusion relying on a fine balance of numerical factors. We analyzed the advantages of gray and white matter segregation from the conduction delay perspective. Our results complement previous studies in some respects but differ in many others. Here, we summarize several novel points. First, we showed that the segregation of white and gray matter is consistent with minimizing conduction delay. Second, we determined the maximum number of neurons in the all-to-all connected network with a reasonable conduction delay and showed that local cortical networks are close to that limit. Third, we proposed a possible explanation for the thickness of the neocortex, which varies surprisingly little among mammalian species. Unlike Murre and Sturdy [20], who suggested that cortical thickness is determined by the maximum density of incoming and outgoing global axons (condition indicated by the thin red line in Figure 6), we argue that in most brains it is the result of minimizing local conduction delay. Fourth, our theory is based on the scaling approach and yields a phase diagram of optimal designs for a wide range of parameters. This allowed us to apply the theory to several different structures other than the neocortex. Derived scaling relationships can be tested by future experimental measurements. Wiring Volume and Conduction Delay Minimization As features of brain design have been explained by minimizing both the total volume and the conduction time delay, it is natural to wonder how these approaches relate to each other. In general, the evolutionary cost is likely to include both the volume and the time delay. Hopefully, such unified framework will emerge eventually. In the meanwhile, since the exact form of the cost function is not known, we sought to construct theories to explain features of brain architecture based on the simplest possible assumptions. Next, we proposed how time delay and volume can be related based on the current theory. In our model, conduction delay in local circuits is minimal when the local wire diameter is at its optimal value, which corresponds to an optimum gray matter volume. (For details, see the first section in Results.) The local conduction delay increases when the local wire diameter d is smaller than the optimum value. In this case, volume cost and conduction delay cost are competing requirements. In the opposite case, when the local wire diameter is thicker than the optimal value, invoking additional conduction delay cost is accompanied by additional volume cost. Therefore, as long as the gray matter volume is greater than its optimal volume, e.g., because of intermixing global axons with gray matter, we may associate the additional conduction delay cost with the volume cost, named the effective volume cost. However, in the white matter, the relationship between volume and delay is different. Increasing white matter volume by making the global axon diameter thicker does not increase the global conduction delay (see the second section in Materials and Methods). Thus, the effective volume cost of white matter is just the tissue cost. From this perspective, we propose that gray matter has a greater effective volume cost than does white matter. This may have several biological implications: (1) Initial segments of axons originating from pyramidal neurons head straight toward (and are perpendicular to) the boundary between the white and gray matter. Once axons cross the white/gray matter border, they change direction. Although such design may increase the length of global axons, it largely reduces the effective volume cost of gray matter, because the volume of global axons in the gray matter is minimal. (2) Another implication of differential effective volume costs in the gray and white matter is that the global axons in gray matter may be thinner than in white matter. Such variation in diameter could preserve short conduction delays in local and global connection. Of course, global axons cannot be made infinitesimally small without sacrificing global conduction delay. Further exploration of this effect would require more experimental measurements of diameter changes at the white/gray matter border. (3) In abutting topographically organized cortical sensory areas, the maps are mirror reflections of each other relative to the border of the areas. The purpose of such organization remains unclear, because interarea connections in the white matter do not benefit from this organization. In particular, placing two cortical areas next to each other (without mirror reflection) would not increase the length of interarea connections in the white matter. Yet, according to our theory, neurons close to the border would be at a disadvantage, because their local connections would have to reach further to find appropriate targets. Mirror-reflecting maps relative to the interarea border would eliminate a discontinuity in a map and place neurons with similar receptive fields closer to each other. Such an arrangement would benefit intracortical connections. Materials and Methods Minimization of conduction delay in a local network with branching axon and dendrite design. Here we revisit the analysis from [17] using more specific information about the network. Consider wiring up a local network of n neurons with all-to-all potential connectivity. The mean conduction delay in local circuits is given by where d is the local wire diameter; v 1/3, the linear size of the local network, approximates the average path length between two potentially connect two neurons. We assume a sublinear relationship between local wire diameter and conduction velocity, and β is a proportionality coefficient. From Equation 28, we want to find the minimal local conduction delay and the corresponding optimal local network volume. Therefore, we have to eliminate wire diameter d from the previous equation and rewrite it as a function of local network volume. To get this expression, we first notice that the total volume of the local network is given by where vn is the nonwire volume, which is assumed to be a constant, and χ is the total wire length per neuron. Second, for an all-to-all potentially connected network, by applying the branching axon and dendrite design [6], we also have This expression is derived as follows [6]. First, the local network volume, v, is divided into cubes of volume, d 3, i.e., into v/d 3 voxels. Then, the number of potential contacts between an axon and a dendrite is given by the number of voxels that contain them both. Each axon occupies χ/d voxels, the same number as a dendrite. The fraction of voxels containing the axon is (χ/d)/(v/d 3), the same as the fraction containing the dendrite. Then, the fraction of voxels containing both the axon and the dendrite is the product of the two fractions, χ 2 d 4/v 2. By multiplying this fraction by the total number of voxels, we find the number of voxels containing axon and dendrite, χ 2 d/v. Then, the condition for having at least one potential contact is given by Equation 30. Combining Equation 29 with Equation 30 and excluding χ yields By combining Equation 28 with Equation 31, we obtain In Equation 32, by setting the first derivative of v to zero, we find the optimal network volume, or gray matter volume, should be And the minimal local conduction delay is given by We assume that nonwire consists mostly of synaptic components, such as axonal boutons and spine heads. In addition, only a fraction, f(0.1–0.3), of potential synapses are actual synapses [23]. Therefore, the nonwire volume can be estimated as where vs is a single synapse volume. Assuming that θ = 1/2 from classical cable theory and substituting it into Equations 34 and 35, we find the minimal local conduction delay is proportional to For simplicity, after neglecting f, this expression is used in Equation 6. Furthermore, the optimal wire diameter can also be calculated by combining Equations 31, 33, and 35, which gives After neglecting f, this expression also appears in Equation 5. Global conduction delay can be preserved after intermixing gray and white matter. After introducing the local connections (gray matter) into the global connections, the total network volume swells and Equation 11 changes to where G is the total gray matter volume. After substituting L ~ V 1/3, D ~ L/(BT), i.e., Equation 9 and 10, into Equation 38, the expression for V can be rewritten as After substituting Equation 39 into D ~ L/(BT) ~ V 1/3/(BT), we find the global axon diameter is given by Therefore, as long as T > N 1/2/B, we can find the corresponding global axon diameter D. Local conduction delay increases after intermixing gray and white matter. Consider again the network described in above with n neurons and all-to-all potential connectivity. After white matter perforates the neuropil, its volume inside gray matter can be expressed by vλ, where v is the unperturbed optimal local gray matter volume given by Equation 33 and λ is a positive dimensionless parameter. After such perturbation, the volume of the local network, i.e., Equation 29, changes to Second, for an all-to-all potentially connected network, by applying the branching axon and dendrite design [6], Equation 30 changes to By combining Equations 28, 41, and 42 and excluding χ and d, we can express the local conduction delay as a function of the total local network volume v′: Equation 43 shows that t′ is a monotonically increasing function of λ, and we recover the expression for t in Equation 32 as λ = 0. Moreover, when λ ≪ 1, the local network is still close to the unperturbed optimal state, i.e., v′ ≃ v, and we can expand Equation 43 to the first order of λ, which yields After combining Equation 44 with Equations 32 and 33, we obtain the expression for local conduction delay from the perturbation theory, or After neglecting the numerical coefficient in the spirit of scaling estimate, the last expression also appears in Equation 14. Local conduction delay and surface area in the branching pipe design. We will address stepwise the process by which we developed this design; first, we present general considerations; second, we develop the first-order branching design; and third, we describe the nonbranching pipes design. First, to calculate the local conduction delay in the branching pipes, we consider a general model in which the white matter pipes have total J branching orders. A branch at order k (0 ≤ k ≤ J) has length Lk and pipe diameter Pk. The total number of kth order branches within the neuropil with linear size Lk is given by Mk. Then, we can evaluate the relative local conduction delay increase through the boundary effect introduced by the kth order branches. The affected neuropil volume through the boundary effect is given by the product of total pipe surface area, MkPkLk, and distance ℓ. This means that the ratio of the affected volume to the total gray matter volume, or the relative local conduction delay increase is given by However, Equation 47 does not tell us what the total local conduction delay is, as different branching orders can have different branching length and diameter. To examine this further, we assume that the branching structure has a space-filling feature. In particular, the length of the main branch L 0 is given by the linear size of the network, G 1/3, and the length of k + 1st order branch is given by the interpipe distance among the kth order branches. For the terminal branches k = J, the interpipe distance between them is given by R 0 (Equation 20). If the length of the k + 1st order branches is much larger than the diameter of the kth order branches, i.e., Lk + 1 ≫ Pk, the interpipe distance between kth order branches is given by Lk/Mk 1/2. Thus, we have where LJ + 1 is the interpipe distance among the terminal branches, given by R 0 (Equation 20). By denoting Nk as the number of neurons in the neuropil with linear size Lk, Nk and Nk + 1 also have the following relationship according to Equation 48, where NJ + 1 is the total neuron number in the neuropil with linear size R 0. In addition, because the pipes with length Lk contains the global axons from the neurons inside the neuropil with linear size Lk, we should also have By substituting Equations 48–50 into Equation 47, we find that where ℓNJ + 1 1/2 D/LJ + 1 2 ~ ℓ2/R 0 2 ~ λ, because according to Theorem 2, ℓ2 is the total cross-sectional area of the global axons inside the module with size R 0. Then, the total local conduction delay increase through the boundary effect is given by This expression can be minimized as a function of Mk. As a result, we obtain For J > 1, we also have Given the total number of neurons in the gray matter N = N 0 and the total branching orders J, by substituting Equations 53 and 54 into Equation 49, we can also obtain Mk explicitly. Next, by using Equations 48–50, we can find the optimal branching length and diameter for different branching orders. Second, we consider a simple branching model in which only the first-order branches exist. In this case, J = 1, and by substituting Equation 53 into Equation 49, we obtain By substituting Equations 55 and 53 into Equation 52, the relative local conduction delay increase through the boundary effect is given by where we neglect the numerical factor of the order of one in the spirit of the scaling estimate. The total local conduction delay increase is the sum of Equation 56 and the expression for relative local conduction delay increase due to intermixing nonfasciculated global axonal segments and gray matter, i.e., λ. However, for the scaling estimate, the second term could be ignored, and we obtain Equation 23. Next, we calculate the total surface area of the branching pipes A. According to Equation 56 and Δt/t ~ ℓA/G, we then obtain the total surface area of the branching pipes where the last expression uses the relationship ℓ2/R 0 2 ~ λ. This expression also appears in Equation 21. In addition, we can also estimate the diameter and length of the first-order branching pipes. P 1 can be obtained by combining Equations 49 and 50, which yields And according to Equation 48, L 1 is given by In the previous analysis, we assume that the length of the first-order branches is much larger than the diameter of the main branches, i.e., L 1 ≫ P 0, which allows us to use Equation 48. This assumption holds when the total white matter volume is much smaller than the gray matter volume, i.e., ND 2 ≪ G 2/3. In the opposite regime, however, L 1 ≪ P 0 must hold, as the volume of the main branching pipe is much larger than the gray matter volume surrounding it. To see this, note that the volume of the main branching pipe is given by P 0 2 L 0, where L 0 is the length of the main branch and the volume of the gray matter surrounding an individual pipe is given by (P 0 + L 1)2 L 0 − P 0 2 L 0. Then, it is easy to check that if the gray matter volume is much larger than the white matter pipe volume, we have L 1 ≫ P 0, while in the opposite case we have L 1 ≪ P 0. Geometrically, when ND 2 ≫ G 2/3, the gray matter resembles a sheet, and the sheet thickness is given by the length of the first-order branches. As the pipe design exhibits a different configuration when ND 2 ≫ G 2/3, we expect that the expressions for the total surface area of the pipes and the minimal local conduction delay are different from what we derived above. In this case, the total surface area of the main branching pipes is equal to the surface area of the gray matter sheet G/L 1, and the relative local conduction delay increase through the boundary effect of the main branches is given by Δt 0/t ~ ℓG/L 1 G ~ ℓ/L 1. To calculate the boundary effect induced by the terminal branches, we assume that R 0 ≫ P 1, where P 1 is the diameter of the terminal branches. This condition allows us to use Equations 48–50. Later, we will confirm that R 0 ≫ P 1 holds. Then, L 1 ~ M 1 1/2 R 0, P 1 ~ M 1 1/4ℓ, and the relative delay increase due to the terminal branches is given by Δt 1/t ~ ℓP 1 M 1/L 1 2 ~ λM 1 1/4. By adding up the local delay from the main and the first-order branches, we find that in the regime ND 2 ≫ G 2/3, the total local conduction delay increase is given by Minimizing this expression as a function of M 1, we obtain M 1 ~ λ −2/3, and Δt/t ~ λ 5/6, as appeared in Equation 25. Next, we calculate the total surface area of the pipes A. As Δt/t ~ ℓA/G ~ λ 5/6, we then obtain the total surface area of the branching pipes as appeared in Equation 24. To check whether R 0 ≫ P 1, we note that P 1 ~ M 1 1/4ℓ. Then, R 0 ≫ P 1 requires R 0 ≫ λ −1/6ℓ, as M 1 ~ λ −2/3. In turn, this requires λ ≪ 1 as ℓ/R 0 ~ λ 1/2. Thus, R 0 ≫ P 1 if λ ≪ 1. This condition should always be satisfied for the PD. Third, the nonbranching pipe model corresponds to J = 0. It does not belong to the PD, because A ≪ G/R 0 does not always hold in such a design when λ ≪ 1, i.e., ND 2 ≪ G/ℓ. To see this, we note that in the regime ND 2 ≫ G 2/3, A ~ G/R 0 must hold in the nonbranching pipe model, because the pipe diameter P 0 is much larger than R 0. In other words, when G/ℓ ≫ ND 2 ≫ G 2/3, the gray matter in the nonbranching pipe model resembles a sheet with thickness R 0. Scaling of the mammalian neocortex. The theoretical framework developed in this paper allows us to derive several scaling laws for the neocortex. Provided our perturbation theory is valid, the total neocortical volume G should be proportional to the total nonwire volume. Assuming that nonwire contains mostly synapses, we have First, from Equation 62, we find that the synaptic density, ρs, is a constant, since ρs ~ Nn/G ~ 1/vs, where the average synapse volume vs is assumed to be a constant in different cortical areas and across different species. The prediction of constant synapse density is supported by experimental observations [31,40,81,82] from a small number of taxa so far, and was used as a starting point to derive scaling laws of the mammalian brains in several theoretical papers [39,45]. Second, we find the neuronal density ρ ~ N/G ~ N/(Nnvs) ~ 1/n. Since ρ scales inversely as the cubic root of total brain volume V across different mammalian species (ρ ~ V −1/3) [40,83], and the cortical volume is loosely proportional to the brain volume (G ~ V) [84], we find n ~ V 1/3, N ~ V 2/3, and n ~ N 1/2. We note that Braintenberg [31,44] has previously proposed the square-root relationship between n and N. He assumed that the cerebral cortex could be divided into N 1/2 compartments and each compartment contains N 1/2 neurons. The local connectivity within a compartment is almost all-to-all, and every compartment is connected to every other one by a global axon. Third, we find that the global axon diameter D scales as V 1/6. To see this, we note that the total white matter volume W is given by ND 2 V 1/3, where the average length of global axons in the white matter is assumed to be proportional to the brain size, V 1/3. Since N ~ V 2/3, and it has also been reported that W ~ V 4/3 across different mammalian species [3,39,84–86], we find D ~ V 1/6. This is consistent with recent measurements from corpus callosum, which indicates that the average diameter of global axons scales monotonically with the brain size [40]. Then, using n ~ V 1/3, D ~ V 1/6 and Equation 27, we obtain R0 ~ V 4/27, an expression from the first section in Discussion. We are grateful to Alexei Koulakov, Sen Song, and Samuel S. H. Wang for helpful discussions and to Georg Streidter for helpful comments on the manuscript. We also thank Maxim Nikitchenko for suggestions about the figures. This research was supported by the Swartz Foundation, the Klingenstein Foundation, and the National Institutes of Health/National Institute of Mental Health grant 69838. Competing interests. The authors have declared that no competing interests exist. Author contributions. QW and DBC conceived the theory, performed the calculations, and wrote the paper. Abbreviations HDhomogeneous design PDperforated design ==== Refs References Wright RD 1934 Some mechanical factors in the evolution of the central nervous system J Anat 69 86 88 17104520 Mitchison G 1991 Neuronal branching patterns and the economy of cortical wiring Proc R Soc Lond B Biol Sci 245 151 158 Zhang K Sejnowski TJ 2000 A universal scaling law between gray matter and white matter of cerebral cortex Proc Natl Acad Sci U S A 97 5621 5626 10792049 Van Essen DC 1997 A tension-based theory of morphogenesis and compact wiring in the central nervous system Nature 385 313 318 9002514 Cherniak C 1992 Local optimization of neuron arbors Biol Cybern 66 503 510 1586674 Chklovskii DB 2004 Synaptic connectivity and neuronal morphology: Two sides of the same coin Neuron 43 609 617 15339643 Allman JM 1999 Evolving brains New York Scientific American Library 235 p. Striedter GF 2005 Principles of brain evolution. Donini G, editor Sunderland (Massachusetts) Sinauer Associates 436 p. Jerison HJ 1973 Evolution of the brain and intelligence New York Academic Press 496 p. Squire LR Kandel ER 2000 Memory: From mind to molecules New York Scientific American Library 246 p. Hebb DO 1949 The organization of behavior: A neuropsychological theory New York Wiley 354 p. Laughlin SB Sejnowski TJ 2003 Communication in neuronal networks Science 301 1870 1874 14512617 Lennie P 2003 The cost of cortical computation Curr Biol 13 493 497 12646132 Levy WB Baxter RA 1996 Energy efficient neural codes Neural Comput 8 531 543 8868566 Attwell D Laughlin SB 2001 An energy budget for signaling in the grey matter of the brain J Cereb Blood Flow Metab 21 1133 1145 11598490 Dickson BJ Cline H Polleux F Ghosh A 2001 Making connections. Meeting: Axon guidance and neural plasticity EMBO Rep 2 182 186 11266357 Chklovskii DB Schikorski T Stevens CF 2002 Wiring optimization in cortical circuits Neuron 34 341 347 11988166 Chklovskii DB Stepanyants A 2003 Power-law for axon diameters at branch point BMC Neurosci 4 18 12946281 Ruppin E Schwartz EL Yeshurun Y 1993 Examining the volume efficiency of the cortical architecture in a multi-processor network model Biol Cybern 70 89 94 8312401 Murre JM Sturdy DP 1995 The connectivity of the brain: Multi-level quantitative analysis Biol Cybern 73 529 545 8527499 Douglas RJ Koch C Mahowald M Martin KA Suarez HH 1995 Recurrent excitation in neocortical circuits Science 269 981 985 7638624 Binzegger T Douglas RJ Martin KA 2004 A quantitative map of the circuit of cat primary visual cortex J Neurosci 24 8441 8453 15456817 Stepanyants A Hof PR Chklovskii DB 2002 Geometry and structural plasticity of synaptic connectivity Neuron 34 275 288 11970869 Chklovskii DB Mel BW Svoboda K 2004 Cortical rewiring and information storage Nature 431 782 788 15483599 Kalisman N Silberberg G Markram H 2005 The neocortical microcircuit as a tabula rasa Proc Natl Acad Sci U S A 102 880 885 15630093 Ritchie JM 1995 The axon: Structure, function, and pathophysiology Physiology of axons New York Oxford University Press 68 96 Hodgkin AL 1954 A note on conduction velocity J Physiol 125 221 224 13192767 Rushton WA 1951 A theory of the effects of fibre size in medullated nerve J Physiol 115 101 122 14889433 Koch C 1999 Biophysics of computation: Information processing in single neurons New York Oxford University Press 585 p. Hoffmeister B Janig W Lisney SJ 1991 A proposed relationship between circumference and conduction velocity of unmyelinated axons from normal and regenerated cat hindlimb cutaneous nerves Neuroscience 42 603 611 1896136 Braitenberg V Schuz A 1998 Cortex: Statistics and geometry of neuronal connectivity Berlin Springer 249 p. Markram H Lubke J Frotscher M Sakmann B 1997 Regulation of synaptic efficacy by coincidence of postsynaptic APs and EPSPs Science 275 213 215 8985014 Attwell D Gibb A 2005 Neuroenergetics and the kinetic design of excitatory synapses Nat Rev Neurosci 6 841 849 16261178 Swadlow HA Waxman SG 1976 Variations in conduction velocity and excitability following single and multiple impulses of visual callosal axons in the rabbit Exp Neurol 53 128 150 964334 Griff ER Greer CA Margolis F Ennis M Shipley MT 2000 Ultrastructural characteristics and conduction velocity of olfactory receptor neuron axons in the olfactory marker protein-null mouse Brain Res 866 227 236 10825498 Ringo JL 1991 Neuronal interconnection as a function of brain size Brain Behav Evol 38 1 6 1657274 Ringo JL Doty RW Demeter S Simard PY 1994 Time is of the essence: A conjecture that hemispheric specialization arises from interhemispheric conduction delay Cereb Cortex 4 331 343 7950307 Watts DJ Strogatz SH 1998 Collective dynamics of “small-world” networks Nature 393 440 442 9623998 Changizi MA 2001 Principles underlying mammalian neocortical scaling Biol Cybern 84 207 215 11252638 Harrison KH Hof PR Wang SS 2002 Scaling laws in the mammalian neocortex: Does form provide clues to function? J Neurocytol 31 289 298 12815248 Sporns O Kotter R 2004 Motifs in brain networks PLoS Biol 2 e369. DOI: 10.1371/journal.pbio.0020369 . 15510229 Karbowski J 2003 How does connectivity between cortical areas depend on brain size? Implications for efficient computation J Comput Neurosci 15 347 356 14618069 Deacon T 1990 Rethinking mammalian brain evolution Amer Zool 30 629 705 Braitenberg V 2001 Brain size and number of neurons: An exercise in synthetic neuroanatomy J Comput Neurosci 10 71 77 11316341 Stevens CF 1989 How cortical interconnectedness varies with network size Neural computation 1 473 479 Hursh J 1939 Conduction velocity and diameter of nerve fibers Amer J Physiol 127 131 139 Boyd IA Kalu KU 1979 Scaling factor relating conduction velocity and diameter for myelinated afferent nerve fibres in the cat hind limb J Physiol 289 277 297 458657 Stepanyants A Tamas G Chklovskii DB 2004 Class-specific features of neuronal wiring Neuron 43 251 259 15260960 Carmichael ST Price JL 1996 Connectional networks within the orbital and medial prefrontal cortex of macaque monkeys J Comp Neurol 371 179 207 8835726 Felleman DJ Van Essen DC 1991 Distributed hierarchical processing in the primate cerebral cortex Cereb Cortex 1 1 47 1822724 Prothero JW Sundsten JW 1984 Folding of the cerebral cortex in mammals. A scaling model Brain Behav Evol 24 152 167 6466964 Hofman MA 1985 Size and shape of the cerebral cortex in mammals. I. The cortical surface Brain Behav Evol 27 28 40 3836731 Prothero J 1997 Cortical scaling in mammals: A repeating units model J Hirnforsch 38 195 207 9176732 Goldman-Rakic PS 1982 Cytoarchitectonic heterogeneity of the primate neostriatum: Subdivision into Island and Matrix cellular compartments J Comp Neurol 205 398 413 7096628 Herkenham M Edley SM Stuart J 1984 Cell clusters in the nucleus accumbens of the rat, and the mosaic relationship of opiate receptors, acetylcholinesterase and subcortical afferent terminations Neuroscience 11 561 593 6325999 Shepherd GM 1998 The synaptic organization of the brain New York Oxford University Press 648 p. Herkenham M Pert CB 1981 Mosaic distribution of opiate receptors, parafascicular projections and acetylcholinesterase in rat striatum Nature 291 415 418 6165892 Graybiel AM Ragsdale CW Jr. Yoneoka ES Elde RP 1981 An immunohistochemical study of enkephalins and other neuropeptides in the striatum of the cat with evidence that the opiate peptides are arranged to form mosaic patterns in register with the striosomal compartments visible by acetylcholinesterase staining Neuroscience 6 377 397 6164013 Gerfen CR 1989 The neostriatal mosaic: Striatal patch-matrix organization is related to cortical lamination Science 246 385 388 2799392 Partadiredja G Miller R Oorschot DE 2003 The number, size, and type of axons in rat subcortical white matter on left and right sides: A stereological, ultrastructural study J Neurocytol 32 1165 1179 15044847 Brown LL Feldman SM Smith DM Cavanaugh JR Ackermann RF 2002 Differential metabolic activity in the striosome and matrix compartments of the rat striatum during natural behaviors J Neurosci 22 305 314 11756514 Oorschot DE 1996 Total number of neurons in the neostriatal, pallidal, subthalamic, and substantia nigral nuclei of the rat basal ganglia: A stereological study using the cavalieri and optical disector methods J Comp Neurol 366 580 599 8833111 Vates GE Broome BM Mello CV Nottebohm F 1996 Auditory pathways of caudal telencephalon and their relation to the song system of adult male zebra finches J Comp Neurol 366 613 642 8833113 Cullheim S Fleshman JW Glenn LL Burke RE 1987 Membrane area and dendritic structure in type-identified triceps surae alpha motoneurons J Comp Neurol 255 68 81 3819010 Friston KJ Frith CD Liddle PF Frackowiak RS 1993 Functional connectivity: The principal-component analysis of large (PET) data sets J Cereb Blood Flow Metab 13 5 14 8417010 Sporns O Tononi G Edelman GM 2000 Theoretical neuroanatomy: Relating anatomical and functional connectivity in graphs and cortical connection matrices Cereb Cortex 10 127 141 10667981 Kotter R Stephan KE Palomero-Gallagher N Geyer S Schleicher A 2001 Multimodal characterisation of cortical areas by multivariate analyses of receptor binding and connectivity data Anat Embryol (Berl) 204 333 350 11720237 Hilgetag CC Grant S 2000 Uniformity, specificity and variability of corticocortical connectivity Philos Trans R Soc Lond B Biol Sci 355 7 20 10703041 Stephan KE Hilgetag CC Burns GA O'Neill MA Young MP 2000 Computational analysis of functional connectivity between areas of primate cerebral cortex Philos Trans R Soc Lond B Biol Sci 355 111 126 10703047 Stephan KE Kamper L Bozkurt A Burns GA Young MP 2001 Advanced database methodology for the collation of connectivity data on the macaque brain (CoCoMac) Philos Trans R Soc Lond B Biol Sci 356 1159 1186 11545697 Young MP Scannell JW O'Neill MA Hilgetag CC Burns G 1995 Non-metric multidimensional scaling in the analysis of neuroanatomical connection data and the organization of the primate cortical visual system Philos Trans R Soc Lond B Biol Sci 348 281 308 8577827 Hilgetag CC O'Neill MA Young MP 2000 Hierarchical organization of macaque and cat cortical sensory systems explored with a novel network processor Philos Trans R Soc Lond B Biol Sci 355 71 89 10703045 Hilgetag CC Burns GA O'Neill MA Scannell JW Young MP 2000 Anatomical connectivity defines the organization of clusters of cortical areas in the macaque monkey and the cat Philos Trans R Soc Lond B Biol Sci 355 91 110 10703046 Sporns O Zwi JD 2004 The small world of the cerebral cortex Neuroinformatics 2 145 162 15319512 Tononi G Sporns O 2003 Measuring information integration BMC Neurosci 4 31 14641936 Tononi G Sporns O Edelman GM 1994 A measure for brain complexity: Relating functional segregation and integration in the nervous system Proc Natl Acad Sci U S A 91 5033 5037 8197179 Friston K 2002 Functional integration and inference in the brain Prog Neurobiol 68 113 143 12450490 Penny WD Stephan KE Mechelli A Friston KJ 2004 Modelling functional integration: a comparison of structural equation and dynamic causal models Neuroimage 23 Suppl 1 S264 S274 Kotter R Stephan KE 2003 Network participation indices: characterizing component roles for information processing in neural networks Neural Netw 16 1261 1275 14622883 Passingham RE Stephan KE Kotter R 2002 The anatomical basis of functional localization in the cortex Nat Rev Neurosci 3 606 616 12154362 Cragg BG 1967 The density of synapses and neurones in the motor and visual areas of the cerebral cortex J Anat 101 639 654 4964696 Schuz A Demianenko GP 1995 Constancy and variability in cortical structure. A study on synapses and dendritic spines in hedgehog and monkey J Hirnforsch 36 113 122 7751602 Tower D 1954 Structural and functional organization of mammalian cerebral cortex: The correlation of neuronal density with brain size Journal of Comparative Neurology 101 19 52 13211853 Hofman MA 1989 On the evolution and geometry of the brain in mammals Prog Neurobiol 32 137 158 2645619 Hofman MA 1988 Size and shape of the cerebral cortex in mammals. II. The cortical volume Brain Behav Evol 32 17 26 3056571 Bush EC Allman JM 2003 The scaling of white matter to gray matter in cerebellum and neocortex Brain Behav Evol 61 1 5 12626858	16389299	PMC1323466	CC BY	2021-01-05 09:18:23	no	PLoS Comput Biol. 2005 Dec 30; 1(7):e78	utf-8	PLoS Comput Biol	2,005	10.1371/journal.pcbi.0010078	oa_comm
==== Front PLoS Comput BiolPLoS Comput. BiolpcbiplcbploscompPLoS Computational Biology1553-734X1553-7358Public Library of Science San Francisco, USA 1638930010.1371/journal.pcbi.001007905-PLCB-RA-0224R4plcb-01-07-09Research ArticleBioinformatics - Computational BiologyEvolutionEukaryotesAnimalsYeast and FungiPlantsNew Maximum Likelihood Estimators for Eukaryotic Intron Evolution New Estimators for Intron EvolutionNguyen Hung D Yoshihama Maki Kenmochi Naoya Frontier Science Research Center, University of Miyazaki, Kiyotake, Miyazaki, JapanBourne Philip EditorUniversity of California San Diego, United States of America To whom correspondence should be addressed. E-mail: [email protected] 2005 30 12 2005 1 7 e791 9 2005 29 11 2005 Copyright: © 2005 Nguyen 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 evolution of spliceosomal introns remains poorly understood. Although many approaches have been used to infer intron evolution from the patterns of intron position conservation, the results to date have been contradictory. In this paper, we address the problem using a novel maximum likelihood method, which allows estimation of the frequency of intron insertion target sites, together with the rates of intron gain and loss. We analyzed the pattern of 10,044 introns (7,221 intron positions) in the conserved regions of 684 sets of orthologs from seven eukaryotes. We determined that there is an average of one target site per 11.86 base pairs (bp) (95% confidence interval, 9.27 to 14.39 bp). In addition, our results showed that: (i) overall intron gains are ~25% greater than intron losses, although specific patterns vary with time and lineage; (ii) parallel gains account for ~18.5% of shared intron positions; and (iii) reacquisition following loss accounts for ~0.5% of all intron positions. Our results should assist in resolving the long-standing problem of inferring the evolution of spliceosomal introns. Synopsis When did spliceosomal introns originate, and what is their role? These questions are the central subject of the introns-early versus introns-late debate. Inference of intron evolution from the pattern of intron position conservation is vital for resolving this debate. So far, different methods of two approaches, maximum parsimony (MP) and maximum likelihood (ML), have been developed, but the results are contradictory. The differences between previous ML results are due predominantly to differing assumptions concerning the frequency of target sites for intron insertion. This paper describes a new ML method that treats this frequency as a parameter requiring optimization. Using the pattern of intron position in conserved regions of 684 clusters of gene orthologs from seven eukaryotes, the authors found that, on average, there is one target site per ~12 base pairs. The results of intron evolution inferred using this optimal frequency are more definitive than previous ML results. Since the ML method is preferred to the MP one for large datasets, the current results should be the most reliable ones to date. The results show that during the course of evolution there have been slightly more intron gains than losses, and thus they favor introns-late. These results should shed new light on our understanding of intron evolution. Citation:Nguyen HD, Yoshihama M, Kenmochi N (2005) New maximum likelihood estimators for eukaryotic intron evolution. PLoS Comput Biol 1(7): e79. ==== Body Introduction Twenty-eight years have passed since the discovery of spliceosomal introns [1], but their evolution remains poorly understood. In the ongoing debate on intron evolution, the central issues include: introns-early versus introns-late, the mini-gene hypothesis, the proto-splice site hypothesis, rates of intron gain and loss, and ratio of parallel gain. Reconstruction of intron evolution from observed data is an important step toward resolution of these issues. Although introns have high mutation rates, making it difficult to trace lineages through sequence homology, their positions are well conserved [2,3]. Therefore, it is possible to reconstruct intron evolution by comparing intron positions among sets of orthologous genes from different species. There are five basic steps involved in this reconstruction: step 1, sequence genomes and annotate the genes; step 2, select sets of orthologous genes among species; step 3, align orthologous genes to produce an intron presence/absence matrix; step 4, reconstruct a phylogenetic tree of the species studied; and step 5, make inferences of intron evolution based on the intron matrix and the phylogenetic tree. Among these, step 5 is perhaps the most critical, since erroneous inferences will lead to misunderstandings of intron evolution. However, to date this step remains the least investigated. Two main approaches, maximum parsimony (MP) and maximum likelihood (ML), have been applied to determine intron evolution from the pattern of intron position conservation. The MP approach, based on the assumption that intron gain and loss events occur rarely in evolution, infers the most parsimonious scenario (measured as a function of gains and losses). Conversely, the ML approach infers a scenario with the highest probability of producing the observed data, for a given model of intron evolution. In a large-scale study of intron evolution, Rogozin et al. [4] applied an MP method to infer intron gains and losses along each branch of a phylogenetic tree of eight eukaryotes using 684 gene orthologs. Their results demonstrated that intron density at the crown (plant–animal) ancestor is relatively high (nearly one-third of the density found in humans). Gains outweighed losses in some lineages; in others, the opposite trend was observed. However, overall there were two gains per loss, suggesting a more important role for intron gain than loss during the course of eukaryotic evolution. In another large-scale study of 2,073 sets of orthologs from four fungal species, Nielsen et al. [5] used an MP method to infer intron gains and losses, then used a probabilistic model to correct the numbers. They found that intron gains are on a par with intron losses in the four species investigated. Roy and Gilbert [6,7] applied an ML method to the above mentioned dataset of eight eukaryotes (excluding one species). Their results were somewhat surprising: the genes of the crown ancestor were rich in introns (about two-thirds the density found in humans), and many lineages exhibited a notable excess of intron losses over gains. The ratio between gains and losses in this study was 0.7, suggesting a general trend toward decreased intron density. In another study, Qiu et al. [8] applied a Bayesian network method, also based on the ML principle, to infer the evolution of introns in ten gene families containing a total of 677 sequences. Their results suggest that many of the intron positions shared across various species are the result of independent gains, and are not due to conservation of intron position. The results of the two ML methods clearly contradict each other and also differ significantly from results of the MP methods. This may be because of the different assumptions they made about the number of target sites; a target site is defined here as a possible position for intron insertion in the multiple-sequence alignment of all species. The number of target sites thereby includes all observed sites, plus possible sites that had introns in the past or may have introns in the future. The method of Roy and Gilbert [6,7] assumes that parallel gains do not occur at all, which will happen only when the number of target sites is many orders of magnitude larger than the number of observed sites. In contrast, the method of Qui et al. [8] assumes that the number of target sites is the number of observed sites. In this paper, we propose a new ML method, the underlying basis of which is the treatment of target site number as a parameter requiring optimization. A log-likelihood function was formulated to allow optimization of the number of target sites, and an expectation-maximization (EM) algorithm was developed to optimize the log-likelihood function. Our results paint a different picture from those provided by previous ML methods. Results Test between the Coelomata and Ecdysozoa Hypotheses There are two hypotheses to explain the relationships of three eukaryotic groups: arthropods, nematodes, and deuterostomes (Figure 1) [9]. The coelomata hypothesis joins arthropods and deuterostomes, placing nematodes as the outgroup; the ecdysozoa hypothesis joins the arthropods and nematodes. Since our dataset includes representatives from all three groups (Drosophila melanogaster and Anopheles gambiae as arthropods, Caenorhabditis elegans as a nematode, and Homo sapiens as a deuterostome), we first tested both hypotheses using the pattern of intron position conservation. In our method, the maximum log-likelihood values for each tree configuration are computed, after which a χ2 test with one degree of freedom is performed for the value −2logΛ, where Λ is the likelihood ratio. The maximum log-likelihood value for the ecdysozoa hypothesis was −255.48 (Figures 2 and 3), whereas that for the coelomata hypothesis was −276.09 (Figure S1). Thus, −2logΛ = 41.22. Consequently, the coelomata hypothesis was rejected in favor of the ecdysozoa hypothesis (p < 10−9), a result consistent with that of another proposed method [9]. However, our method is simpler and appears to have a more sound mathematical basis. Hereafter, we will adopt the ecdysozoa hypothesis as the phylogenetic tree of the species studied. Figure 1 Two Hypotheses for the Relationship among Three Groups: Arthropods (A), Nematodes (N), and Deuterostomes (D) The coelomata hypothesis is shown on the left, and the ecdysozoa hypothesis is shown on the right. Figure 2 Relationship between θ and the Maximum Log-Likelihood Value The ecdysozoa phylogeny was used. The maximum log-likelihood value was calculated by treating all parameters other than θ as nuisance parameters and maximizing over them. The arrow shows the MLE of θ as 0.071. The horizontal line indicates a 95% CI of 0.055–0.096. The maximum log-likelihood value at the MLE of θ was −255.48. Figure 3 MLEs of the Number of Gains and Losses Using the Ecdysozoa Phylogeny Numbers of introns present in modern species (known) are in black. Numbers of introns present in ancestors (estimated) are in green. Numbers of gains and losses (estimated) are in red and blue, respectively. Branches that experienced >1.5 gains per loss are shown in red, and those that experienced >1.5 losses per gain are in blue. D.mel, D. melanogaster; A.gam, A. gambiae; C.ele, C. elegans; H.sap, H. sapiens; S.pom, S. pombe; A.tha, A. thaliana; P.fal, P. falciparum. Reconstruction of Intron Evolution The maximum likelihood estimator (MLE) of θ (see Materials and Methods for the definition of θ) was 0.071 (95% confidence interval [CI], 0.055–0.096; Figure 2). As a result, there was an average of one target site per 11.86 base pairs (bp) of sequence (95% CI, 9.27–14.39 bp), or 8.43 target sites per 100 bp (95% CI, 6.95–10.79). Figure 3 shows the MLEs of the numbers of intron gains and losses along each branch of the phylogenetic tree. All branches leading to H. sapiens, and the terminal branch leading to Arabidopsis thaliana, experienced many more gains than losses. In contrast, all other branches except that leading to C. elegans experienced many more losses than gains. Losses along the terminal branch leading to C. elegans slightly outnumbered gains. In total, there were 6,382 gains versus 5,099 losses. Figure 3 also shows the numbers of introns at internal nodes corresponding to the MLEs of the parameters. The intron density in the crown (plant–animal) ancestor was approximately one-third of that in H. sapiens, and increased until the bilateral ancestor appeared, after which it decreased in other lineages, while continuing to increase in H. sapiens. The intron density also increased in the terminal branch leading to A. thaliana, but decreased in the terminal branch leading to Schizosaccharomyces pombe. Although our method cannot infer the intron density at the root node of the phylogenetic tree, it inferred that 107 intron positions (the parameter Q 11 in Protocol S1) were shared between Plasmodium falciparum and the crown ancestor. This number will become the lower bound for intron density at the root node if we apply the MP principle to the two deepest branches. We also measured the frequencies of parallel intron gain and reacquisition following loss. Using our model of intron evolution, and incorporating the rates of intron gain and loss depicted in Figure 3, double parallel gains accounted for ~18.2% and triple parallel gains accounted for ~0.3% of the intron positions shared across two or more species (excluding P. falciparum). The probability that more than three introns would be gained in parallel at the same position was close to zero. Therefore, the total number of parallel intron gains accounted for ~18.5% of the shared intron positions. This result is slightly higher than the predicted range of 5% to 10%, but still less than the upper bound of 20% reported in a study using a different method on the same dataset [10]. The probability of reacquisition following loss was also small (~0.5% of the total intron positions). The observed and expected (using the MLEs of the parameters) numbers of cases for each external intron pattern are presented in Table 1. Since many expected numbers are <5, the χ2 test cannot be used directly. Therefore, we grouped 75 intron patterns showing values of <3 into five new equally sized categories, and the result of the χ2 test was not statistically significant (χ2 = 36.15 with 56 df, p = 0.98). Table 1 Observed and Expected Numbers of Cases for Each External Intron Pattern Verification of the Present Results Since the log-likelihood function is highly complex and no closed form is available for calculating the MLEs of the parameters, three numerical methods were applied. The first was based on an EM algorithm, the second on a downhill simplex (DS) method, and the third on a genetic algorithm (GA). See the Materials and Methods section for details on all of these methods. The present results were obtained using the EM algorithm, the advantage of which is its rapid convergence to a solution (Figure S2). However, its disadvantage is that its solution may not be the global maximum, but rather a local maximum, or even a stationary point. Therefore, the two other methods were used to verify whether or not the EM algorithm had found the global maximum. The DS method has the advantage that it does not require derivatives of the function to be optimized, and is less likely to be trapped in local maxima than other iterative methods such as EM. However, it is quite computationally expensive. The GA method is also unlikely to be trapped in local maxima, although its solution quality is not as high, and it is time-consuming. Over ten runs, neither DS nor GA methods showed an improvement in results; further, the best of each of these ten runs produced similar results to the EM algorithm (unpublished data). Therefore, it is highly likely that the EM algorithm found the global maximum. The fact that the EM algorithm, using θ = 100, produced almost the same results as a previous ML method [7] provides further support for this conclusion (Figure 4). Figure 4 Comparison of Results between the EM algorithm (θ = 100) and a Previous ML Method [6] The results of the EM algorithm when using θ = 100 are shown. Numbers in parentheses indicate the differences between the two results using our results as the benchmark. Numbers of introns present in modern species (known) are in black. Numbers of introns present in ancestors (estimated) are in green. Numbers of gains and losses (estimated) for each branch are in red and blue, respectively. Branches that experienced >1.5 gains per loss are shown in red, and those that experienced >1.5 losses per gain are in blue. D.mel, D. melanogaster; A.gam, A. gambiae; C.ele, C. elegans; H.sap, H. sapiens; S.pom, S. pombe; A.tha, A. thaliana; P.fal, P. falciparum. Discussion Implications of Target Site Frequency To the best of our knowledge, ours is the first estimate of frequency of target sites, and can be accounted for by two hypotheses. The first is the exon theory of genes, which postulates that present-day exons are remnants of primordial mini-genes (~45–60 bp long) from the time of the last common ancestor of prokaryotes and eukaryotes [11]. The second is the proto-splice–site hypothesis, which predicts that introns are not inserted randomly, but only into proto-splice sites. It is proposed that proto-splice sites contain the sequence MAG\|R [12] or MAG\|Gt [13], where M is A or C, and R is A or G. Based on our results, if the former hypothesis is true, then the average length of mini-genes should be ~12 bp. On the other hand, if the latter hypothesis holds true, there should be ~8.4 proto-splice sites per 100 bp. The average length of 12 bp, however, appears to be too short for mini-genes, which are supposed to be 45–60 bp long. Moreover, it is unlikely that introns mostly lost during the archaea period were then reacquired by H. sapiens and A. thaliana. Therefore, we believe that the proto-splice site hypothesis is the more likely explanation for our results. However, the estimated frequency of the pattern MAG\|R in the whole genomes of the six species we examined, excluding A. gambiae, is 2%–3% (unpublished data) and therefore too small to explain the optimal value of 8.4%. Consequently, the consensus pattern of proto-splice sites may need to be re-determined. Relative Importance of Intron Gains and Losses Our results show a slight excess (~25%) of intron gains over losses (6,382 versus 5,099). However, there was an excess of intron-lost branches over intron-gained branches (five versus four). Since these results may change upon addition of more species, the conclusions about the relative importance of intron gain and intron loss will depend upon the balance of future datasets. Assuming that the seven species studied here represent a balanced dataset, our conclusion is that intron gain has played a slightly larger role than intron loss during the course of eukaryotic evolution, but that the relative importance may vary with time and lineage. Note that the results shown here were determined using the intron patterns in conserved alignment regions. We predict that the relative importance of intron gain and loss would tip more toward intron gain if complete alignments were used, as suggested by the results of Rogozin et al. [4]. However, we do not know what the effect of misalignments would be in this case. Figure 3 shows that rates of intron gain and loss vary significantly among distantly related species, but only slightly among more closely related species (e.g., between D. melanogaster and A. gambiae). Moreover, except for the ecdysozoa–C. elegans branch in which intron gains and losses are approximately in balance, the other branches experienced either many more gains than losses, or losses than gains. There seems to be an inverse relationship between rates of gain and loss on each branch. One possible explanation for this pattern is that the rates of intron gain and loss may be controlled by selective pressure exerted through genome compaction and expansion. This pressure should vary significantly among distantly related species, but only slightly among those more closely related. Implications for the Introns-Early versus Introns-Late Debate Two theories, introns-early and introns-late, have been proposed for the origin of spliceosomal introns. The introns-early theory asserts that introns were required to facilitate the assembly of early genes and were already present in the last common ancestor of prokaryotes and eukaryotes. Thus, this theory suggests that intron loss is the main driving force for intron evolution [14–17]. In contrast, the introns-late theory suggests that introns were gained after the emergence of eukaryotes and that intron gain plays a major role in the modern distribution of introns [18–20]. Many attempts have been made to resolve the debate (e.g., [21,22]). Although our results do not provide a definitive answer, they appear to lend more support for the introns-late theory. In particular, our results indicate that eukaryotic evolution has not been characterized by a general decrease in intron density, as predicted by the introns-early theory. With regard to the lineage leading to D. melanogaster, intron gains were dominant during the period from the crown ancestor to the bilateral ancestor, and intron losses were dominant after this point. There are two scenarios for intron evolution during the period between the last common ancestor of eukaryotes and the crown ancestor: either an increase or a decrease in intron density. As there would be only one turning point in the former, but two in the latter, the parsimony principle favors the former scenario in which intron density increased during the earliest stage of eukaryotic evolution. Comparison with Previous ML Methods Our results differed greatly from those obtained by Roy and Gilbert [6,7], whose estimates for the numbers of introns at internal nodes were larger than ours. They determined an intron density in the crown (plant–animal) ancestor that was nearly twice that of our results. Their method also tended to infer fewer intron gains and more losses for each branch of the tree. On two branches, one from the crown ancestor to the opisthokont ancestor, and the other from the bilateral ancestor of H. sapiens, Roy and Gilbert's method predicted that gains and losses were approximately in balance. In contrast, our method predicted a notable excess of intron gains over losses. A possible explanation for these differences is that their model assumes that parallel intron gains do not occur [6,7]. This would lead to the assumption that the number of target sites must be extremely large, and would in turn bias the results. To test this hypothesis, we ran the EM algorithm with θ set to 100 (Figure 4), and the results generated were almost identical to those of Roy and Gilbert [6,7], thereby supporting our explanation for the differences. However, a valid value for θ should be in the range (0, 1), where θ = 0 indicates that the number of target sites is equal to the number of observed splice sites, and θ = 1 indicates that the number of target sites is equal to the total length of the alignments. Therefore, the value θ = 100 is unrealistic, since it implies that the number of target sites is approximately 100 times the total length of the alignments. Consequently, their results appear to underestimate intron gains and to overestimate losses. Our results also contradicted the findings of Qui et al. [8], who determined that most intron positions shared between kingdoms were the result of parallel gains. Although different datasets were used, the contradiction may have occurred because their approach includes the assumption that the number of target sites equals the number of observed intron positions (i.e., equivalent to θ = 0 in our method). This assumption may lead to an overestimation of the rate of intron gain, which in turn leads to an overestimation of parallel gains. When we used θ = 0, our method produced results in which the ratio of parallel gain reached ~98.4%, thus supporting our explanation of the differences between the two methods. Additionally, since their dataset contains an average of 68 sequences per gene family, the total number of intron patterns is 268. This is much larger than 49, which is the average number of intron positions per gene family. Thus, it is possible that their sample dataset was insufficiently large for a valid statistical inference. Note that assumptions about the target site frequency of both previous ML methods represented extreme versions of our method (θ = 100 and θ = 0) and were outside the 95% CI (0.055–0.096). Their log-likelihood values were −282.28 and −2199.16, respectively. Thus, they were both rejected in favor of the MLE of θ (p < 10−10 by χ2 tests, 1 df). Comparison with MP Methods Rogozin et al. [4] were the first group to apply MP to the problem of inferring intron gains and losses in the dataset that we used. However, they reported results only for the coelomata phylogeny; another group reported results for the ecdysozoa phylogeny [6,7]. When compared with our method, their method appears to underestimate the number of introns at all of the internal nodes, as well as the number of intron losses on all branches. Furthermore, the number of intron gains was underestimated on all internal branches, but overestimated on several terminal branches (e.g., those leading to H. sapiens, D. melanogaster, and A. gambiae). These variations are believed to be caused by the differences between ML and MP. In an MP method, the most parsimonious scenario is always chosen, even though the next-most parsimonious scenario has a close probability of occurrence. Let us consider an intron pattern in which the chances of the most and the next-most parsimonious scenarios occurring are 51% and 49%, respectively. In the MP method, 100% of cases will be assigned to the most parsimonious scenario, while none will be assigned to the next-most parsimonious. In the ML method, 51% and 49% of cases will be assigned to the most parsimonious and the next-most parsimonious scenarios, respectively. Although the results obtained by the method of Nielsen et al. [5] are not available for the dataset used in this study, we believe that they should fall somewhere between the results of Rogozin et al. [4] and ours. Nielsen et al. [5] used a probabilistic model to correct some intron patterns, for which the possibility of the next-most parsimonious scenario was high, thereby reducing the gap between the results using ML and MP. Although the debate between ML and MP is vigorous [23,24], much of it relates to the problem of phylogenetic tree reconstruction, and may not be relevant to the problem of intron evolution prediction. Errors in predicting the number of events can still lead to the correct tree. Therefore, we believe that ML will be a better predictor of intron evolution when there is a large amount of sample data, such the dataset used in this study. MP may be the method of choice when the sample dataset is insufficiently large for a valid statistical inference (such as the dataset of Qui et al. [8]). Investigating the performance of these two approaches in their prediction of intron evolution will be the subject of future work. Future Applications The method presented here may also be applied to at least two other problems: (i) determination of the phylogenetic tree of eukaryotic species based on the conservation of intron positions; and (ii) inference of rates of gene gain and loss on a genomic scale. For the former, we would consider the phylogenetic tree T as a parameter needing to be optimized, and find the MLE of T over all tree configurations; for the latter, we would need to construct a gene presence/absence matrix, rather than an intron presence/absence matrix. Materials and Methods Dataset. We used the 684 eukaryotic clusters of orthologous genes (KOGs), which are available at ftp://ftp.ncbi.nlm.nih.gov/pub/koonin/intron_evolution [4]. Each KOG contains genes from eight eukaryotes: D. melanogaster, A. gambiae, C. elegans, H. sapiens, S. pombe, Saccharomyces cerevisiae, A. thaliana, and P. falciparum. Multiple protein sequence alignments and intron presence/absence matrices of these KOGs were downloaded from the above site. Following Roy and Gilbert [6,7], we only extracted intron patterns in conserved alignment regions and excluded S. cerevisiae due to its sparse intron distribution. The total length of conserved alignment regions in the dataset was 488,157 bp. The number of intron positions in these conserved regions was 7,221; the number shared across two or more species was 1,787. When the outgroup P. falciparum was excluded, the number of intron positions was 7,049 and the number shared was 1,722. The log-likelihood function. Our model of intron evolution assumes that introns are inserted independently and uniformly at all available target sites. Similarly, intron loss occurs with equal probability at all sites possessing introns, regardless of when they were inserted. Rates of gain and loss are assumed to be constant along each branch of the phylogenetic tree, but they can vary between branches. As suggested by the results of Rogozin et al. [4] and Stoltzfus et al. [25], our model does not allow “intron sliding.” However, our model does differ from some of the others in that it allows for both parallel gains and reacquisition following loss. Given the model of intron evolution described above, our problem can be defined as follows: Let N be the number of species, G be the total length of all alignments in bp, and T be the phylogenetic tree of these species. Since the phylogenetic tree T is binary, it contains N external nodes, each corresponding to a species, M = N − 1 internal nodes, with each node corresponding to a divergent event, and B = 2M branches (including N external branches). An external intron pattern is a binary sequence s 1 s 2...sN of length N. When si = 0 or 1 (i = 1, 2, ..., N), this means that an intron is absent or present at the ith external node. The definition of an internal intron pattern is similar, but with length M instead of N. There are 2N (indexed from 0 to 2N − 1) possible external intron patterns, and 2M (indexed from 0 to 2M − 1) internal intron patterns. For the ith external/internal intron pattern, the binary code of i will give the intron states of the external/internal nodes. For example, for N = 4, the binary code of external intron pattern i = 3 will be “0011.” Thus, an intron will be absent at the first and second species, yet present at the third and fourth species in this pattern. We denote ni (i = 0, 1, …, 2N − 1) as the number of cases for the ith external pattern, and gk and lk (k = 1, 2, ..., B) as the number of gains and losses along the kth branch. We then have to estimate gk and lk for a given set of ni. Note that n 0 (i.e., the number of target sites without introns) cannot be observed, and must be treated as a parameter requiring optimization. Denoting S and P as the numbers of observed splice sites (intron positions) and target sites in the dataset, respectively, we have: It follows from Equation 2 that n 0 = P − S. Since P must ≤G (i.e., the total length of all alignments), n 0 must be constrained by: 0 ≤ n 0 ≤ G − S. If we define such that: , then from Equation 2 we get: . We denote and (k = 1, 2, …, B) as the probabilities of changing state from 0 to 1 (intron gain) and from 1 to 0 (intron loss) along the kth branch, respectively; bk and ek (k = 1, 2, …, B) as the node indexes of the beginning and end nodes of the kth branch, respectively; and sh (h = 1, 2, …, N + M) as the intron state of the hth node. The expected number of cases for the complete intron pattern ij (i = 0, 1, …, 2N − 1, j = 0, 1, …, 2M − 1), which is the combination of the ith external pattern and the jth internal pattern, is calculated by: where: Here is a variable denoting the fraction of target sites having introns at the root node and s 1 is the intron state of the root node. Then the probability pi (i = 0, 1, …, 2N − 1) for the ith external pattern is: Finally, the likelihood function is: and the log-likelihood function is: We have to find the MLEs that maximize the log-likelihood function. Note that with our definitions, the values of are all in the (0, 1) interval. We will denote as the probabilities of intron gain and loss along the two deepest branches, respectively. Then we have the following two propositions: Proposition 1: There are infinite sets of MLEs . Proof: See Protocol S1. Proposition 1 indicates that the number of introns at the root node, as well as the numbers of gains and losses along the two deepest branches, cannot be determined without additional information. Therefore, these values will not be reported in the Results section. Proposition 2: There are 2N −2 sets of MLEs (k = 3, 4, …, B). Proof: See Protocol S2. We define the most biologically meaningful solution in these 2N −2 sets of MLEs (k = 3, 4, …, B) as the one having the least sum of variances for intron gains and losses among all branches, i.e., having the smallest . The EM algorithm. An EM algorithm was proposed for calculating the MLEs of , given a fixed value of . The Brent algorithm was used for finding the MLE of based on the profile likelihood method, i.e., by treating all other parameters as nuisance parameters and maximizing over them. Implementation of the Brent algorithm is problem-independent and straightforward: we simply applied the code in the book Numerical Recipes in C [26]. However, implementation of the EM algorithm is problem-specific. In general, EM is an iterative algorithm comprising two steps: the E-step and the M-step [27]. Our implementation of these two steps is as follows: E-Step: In our problem, the complete dataset comprises the number of cases for all complete intron patterns (i.e., including external and internal nodes). The conditional expected number of cases for the complete intron pattern ij (i = 0, 1, …, 2N − 1, j = 0, 1, …, 2M − 1) given ni is calculated as follows: where mij is calculated based on Equation 3 using the current set of parameters . M-Step: In this step we must find a new set of parameters that maximize the likelihood of the complete data conditioned by the observed data. First, we calculate the conditional expected numbers of gains and losses for each branch k of the phylogenetic tree by: Next, the conditional expected number of introns at each node h is calculated by: Finally, the new parameters are calculated by the following formulas: where o 1\|ni is the conditional expected number of introns at the root node. First, we set the value of λ to S/2P, so that the introns are present in half of the observed intron positions at the root node; then set the values of to 0.01 and the values of to 0.1 (k = 1, 2, …, B). The two steps described above are then repeated until the difference between the new set and the current set of parameters, which is calculated based on Equation 12, is smaller than a predefined value (10−8 in our algorithm). The source code of the EM algorithm was written in C language and is available on request to the corresponding author. The DS method. To implement the DS method, we used the C code in [26] with only two minor modifications: the first limits all parameters to optimization within the (0, 1) range; and the second repeats the main procedure 50 times, in order to obtain a higher-quality solution. The GA method. Our GA is based on a multi-population steady-state GA [28] and can be run in parallel on a cluster of PCs to obtain results more quickly. In this GA, only one offspring solution is produced, either from two parental solutions by crossover, or from one parental solution by mutation in each generation. The offspring is immediately inserted into the population, and if it is fitter it replaces the worse parent. Linear ranking selection with a bias of 1.25 was used for selecting the parents. The population size was set to 400 individuals, which were divided equally into eight subpopulations. The mutation rate was set to 0.5. Supporting Information Figure S1 MLEs of the Numbers of Gains and Losses Using the Coelomata Phylogeny (11 KB PDF) Click here for additional data file. Figure S2 Convergence of the EM Algorithm (θ = 0.071) (17 KB PDF) Click here for additional data file. Figure S3 General Form of an Internal Node (9 KB PDF) Click here for additional data file. Protocol S1 Proof for Proposition 1 (80 KB PDF) Click here for additional data file. Protocol S2 Proof for Proposition 2 (90 KB PDF) Click here for additional data file. We thank Akihiro Nakao, Tamayo Uechi, and Sayomi Higa for useful discussion, and Dr. Ikuo Yoshihara for his encouragement. This study was supported by Grants-in-Aid for Scientific Research (14035103, 15310135, 178103, and 17770207) from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), and the Japan Society for the Promotion of Science (JSPS). HDN is a research fellow of the JSPS (17–05174). Competing interests. The authors have declared that no competing interests exist. Author contributions. HDN, MY, and NK conceived and designed the experiments. HDN performed the experiments. HDN, MY, and NK analyzed the data. HDN contributed reagents/materials/analysis tools. HDN and NK wrote the paper. Abbreviations bpbase pair CIconfidence interval DSdownhill simplex EMexpectation-maximization GAgenetic algorithm KOGeukaryotic cluster of orthologous genes MLmaximum likelihood MLEmaximum likelihood estimator MPmaximum parsimony ==== Refs References Sambrook J 1977 Adenovirus amazes at Cold Spring Harbor Nature 268 101 104 593301 Betts MJ Guigó R Agarwal P Russell RB 2001 Exon structure conservation despite low sequence similarity: A relic of dramatic events in evolution? EMBO J 20 5354 5360 11574467 Yoshihama M, Uechi T, Asakawa S, Kawasaki K, Kato S et al. 2002 The human ribosomal protein genes: Sequencing and comparative analysis of 73 genes Genome Res 12 379 390 11875025 Rogozin IB Wolf YI Sorokin AV Mirkin BG Koonin EV 2003 Remarkable interkingdom conservation of intron positions and massive, lineage-specific intron loss and gain in eukaryotic evolution Curr Biol 13 1512 1517 12956953 Nielsen CB Friedman B Birren B Burge CB Galagan JE 2004 Patterns of intron gain and loss in fungi PLoS Biol 2 e422. DOI: 10.1371/journal.pbio.0020422 . 15562318 Roy SW Gilbert W 2005 Complex early genes Proc Natl Acad Sci U S A 102 1986 1991 15687506 Roy SW Gilbert W 2005 Rates of intron loss and gain: Implications for early eukaryotic evolution Proc Natl Acad Sci U S A 102 5773 5778 15827119 Qiu WG Schisler N Stoltzfus A 2004 The evolutionary gain of spliceosomal introns: Sequence and phase preferences Mol Biol Evol 21 1252 1263 15014153 Roy SW Gilbert W 2005 Resolution of a deep animal divergence by the pattern of intron conservation Proc Natl Acad Sci U S A 102 4403 4408 15769859 Sverdlov AV Rogozin IB Babenko VN Koonin EV 2005 Conservation versus parallel gains in intron evolution Nucleic Acids Res 33 1741 1748 15788746 Gilbert W 1987 The exon theory of genes Cold Spring Harb Symp Quant Biol 52 901 905 2456887 Dibb NJ Newman AJ 1989 Evidence that introns arose at proto-splice sites EMBO J 8 2015 2021 2792080 Sverdlov AV Rogozin IB Babenko VN Koonin EV 2004 Reconstruction of ancestral protosplice sites Curr Biol 14 1505 1508 15324669 Gilbert W 1978 Why genes in pieces? Nature 271 501 622185 Doolittle WF 1978 Genes in pieces: Were they ever together? Nature 272 581 582 Blake CCF 1978 Do genes in pieces imply proteins in pieces? Nature 273 267 Roy SW 2003 Recent evidence for the exon theory of genes Genetica 118 251 266 12868614 Cavalier-Smith T 1991 Intron phylogeny: A new hypothesis Trends Genet 7 145 148 2068786 Palmer JD Logsdon JM Jr 1991 The recent origins of introns Curr Opin Genet Dev 1 470 477 1822279 Logsdon JM Jr 1998 The recent origins of spliceosomal introns revisited Curr Opin Genet Dev 8 637 648 9914210 Go M 1981 Correlation of DNA exonic regions with protein structural units in haemoglobin Nature 291 90 92 7231530 Logsdon JM Jr Stoltzfus A Doolittle WF 1998 Molecular evolution: Recent cases of spliceosomal intron gain? Curr Biol 8 R560 R563 9707398 Kolaczkowski B Thornton JW 2004 Performance of maximum parsimony and likelihood phylogenetics when evolution is heterogeneous Nature 431 980 984 15496922 Sober E 2004 The contest between parsimony and likelihood Syst Biol 53 644 653 15371252 Stoltzfus A Logsdon JM Jr Palmer JD Doolittle WF 1997 Intron “sliding” and the diversity of intron positions Proc Natl Acad Sci U S A 94 10739 10744 9380704 Press WH Teukolsky SA Vetterling WT Flannery BP 1992 Minimization or maximization of functions Numerical recipes in C: The art of scientific computing, 2nd edition New York Cambridge University Press 394 444 . Dempster AP Laird NM Rubin DB 1977 Maximum likelihood for incomplete data via the EM algorithm J R Stat Soc B 39 1 38 Whitley D 1989 The GENITOR algorithm and selective pressure: Why rank-based allocation of reproductive trials is best Schaffer JD Proceedings of the Third International Conference on Genetic Algorithms San Francisco Morgan Kaufmann 116 121 .	16389300	PMC1323467	CC BY	2021-01-05 09:18:23	no	PLoS Comput Biol. 2005 Dec 30; 1(7):e79	utf-8	PLoS Comput Biol	2,005	10.1371/journal.pcbi.0010079	oa_comm
==== Front PLoS PathogPLoS PathogppatplpaplospathPLoS Pathogens1553-73661553-7374Public Library of Science San Francisco, USA 1638930110.1371/journal.ppat.001003605-PLPA-RA-0130R2plpa-01-04-05Research ArticleBiotechnologyCell BiologyEpidemiology - Public HealthImmunologyInfectious DiseasesMicrobiologyPathologyVirologyHomo (Human)Mus (Mouse)VirusesUsing the TAP Component of the Antigen-Processing Machinery as a Molecular Adjuvant A New Paradigm in VaccinationVitalis Timothy Z 12Zhang Qian-Jin 12¤aAlimonti Judie 123¤bChen Susan S 124Basha Genc 12Moise Alex 124¤cTiong Jacqueline 123Tian Mei Mei 123Choi Kyung Bok 12Waterfield Douglas 5Jeffries Andy 12Jefferies Wilfred A 12346* 1 The Michael Smith Laboratories, University of British Columbia, Vancouver, British Columbia, Canada 2 Biomedical Research Centre, University of British Columbia, Vancouver, British Columbia, Canada 3 Department of Microbiology and Immunology, University of British Columbia, Vancouver, British Columbia, Canada 4 Department of Zoology, University of British Columbia, Vancouver, British Columbia, Canada 5 Department of Oral Biology, Faculty of Dentistry, University of British Columbia, Vancouver, British Columbia, Canada 6 Department of Medical Genetics, University of British Columbia, Vancouver, British Columbia, Canada Manchester Marianne EditorScripps Research Institute, United States of America* To whom correspondence should be addressed. E-mail: [email protected]¤a Current address: Cellular Biology and Anatomy, LSU Health Sciences Center, Shreveport, Louisiana, United States of America ¤b Current address: Manitoba Institute of Cell Biology, Winnipeg, Manitoba, Canada ¤c Current address: Case Western Reserve University, Cleveland, Ohio, United States of America 12 2005 30 12 2005 1 4 e3615 8 2005 25 10 2005 Copyright: © 2005 Vitalis 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.We hypothesize that over-expression of transporters associated with antigen processing (TAP1 and TAP2), components of the major histocompatibility complex (MHC) class I antigen-processing pathway, enhances antigen-specific cytotoxic activity in response to viral infection. An expression system using recombinant vaccinia virus (VV) was used to over-express human TAP1 and TAP2 (VV-hTAP1,2) in normal mice. Mice coinfected with either vesicular stomatitis virus plus VV-hTAP1,2 or Sendai virus plus VV-hTAP1,2 increased cytotoxic lymphocyte (CTL) activity by at least 4-fold when compared to coinfections with a control vector, VV encoding the plasmid PJS-5. Coinfections with VV-hTAP1,2 increased virus-specific CTL precursors compared to control infections without VV-hTAP1,2. In an animal model of lethal viral challenge after vaccination, VV-hTAP1,2 provided protection against a lethal challenge of VV at doses 100-fold lower than control vector alone. Mechanistically, the total MHC class I antigen surface expression and the cross-presentation mechanism in spleen-derived dendritic cells was augmented by over-expression of TAP. Furthermore, VV-hTAP1,2 increases splenic TAP transport activity and endogenous antigen processing, thus rendering infected targets more susceptible to CTL recognition and subsequent killing. This is the first demonstration that over-expression of a component of the antigen-processing machinery increases endogenous antigen presentation and dendritic cell cross-presentation of exogenous antigens and may provide a novel and general approach for increasing immune responses against pathogens at low doses of vaccine inocula. Synopsis The development of protective vaccines against infectious diseases such as AIDS, SARS, and West Nile virus has become a societal priority but remains a scientific challenge. In recent years, the threat of bioterrorism agents such as anthrax and smallpox has heightened the need for the rapid development of effective new vaccines. One of the major stumbling blocks to the implementation of any vaccine is the toxic side effects on the vaccine candidate. For example, a significant number of doses of a new vaccine against smallpox have been commissioned, but approximately 20% of the individuals targeted to be inoculated will suffer toxicity due to vaccination. Furthermore, an additional difficulty in the production of vaccines is the creation of sufficient doses to vaccinate a large population. The authors have identified a novel approach that appears to address these issues. They demonstrate that the inclusion, in low doses of vaccines, of a normal component of the antigen-processing pathway, the transporter associated with antigen processing (TAP), confers protective immunity against lethal viral loads during viral challenges. This new paradigm is shown to be applicable to many viruses, including poxviruses, and could significantly advance the creation of new vaccines and improve those that already exist. Citation:Vitalis TZ, Zhang QJ, Alimonti J, Chen SS, Basha G, et al. (2005) Using the TAP component of the antigen-processing machinery as a molecular adjuvant. PLoS Pathog 1(4): e36. ==== Body Introduction Major histocompatibility complex (MHC) class I molecules are highly polymorphic cell-surface glycoproteins, which function to bind peptides for presentation to cytotoxic lymphocytes (CTLs) following microbial infection or cell transformation [1–3]. In humans, a number of genes located mainly in the MHC class II region of Chromosome 6 are responsible for the generation, assembly, and transport of MHC class I molecules, referred to as the antigen-processing pathway. These genes include (1) the proteasome components, low molecular-weight polypeptides (LMPs): LMP2, LMP7, and LMP10; (2) transporters associated with antigen processing (TAP): TAP1 and TAP2; (3) the chaperone proteins: calnexin, calreticulin, and tapasin; and (4) MHC class I heavy chain and β2-microglobulin [4–8]. Peptide antigens are transported into the endoplasmic reticulum (ER) by TAP and are loaded onto the MHC complex with the aid of the chaperone proteins. The functional MHC class I complexes are, in turn, transported to the cell surface and presented to T lymphocytes. Precursors of CTLs, through the T cell receptor, recognize foreign peptides derived from pathogens, and begin a cascade of activities leading to stimulation of specific immune responses against pathogen-infected cells. Overall, the expression of stable MHC class I molecules on the cell surface is regulated by the peptides supplied by TAP [1,9]. Antigen-presenting cells (APCs) derived from TAP1−/− mice cannot transport antigenic peptides from the cytoplasm into the lumen of the ER for MHC class I binding and eventual presentation on the cell surface. Therefore, these cells lack the capacity to direct the priming of antigen-specific T cells [8]. The development of vaccine adjuvants that promote immunity at low doses of inocula is one approach to generate protection in individuals who would otherwise respond adversely to the administration of standard doses of inocula. Adverse responses to standard doses of inocula are a problem encountered in vaccination against viruses such as smallpox and, as a result, conventional vaccines cannot be administered to a significant fraction of the population who are either immune suppressed or who would otherwise react adversely to the established vaccine protocols [10]. As vaccination against a variety of pathogens becomes more widespread, there will be a greater need to increase the efficiency of the inocula while reducing the sizes of the batches of vaccine required for vaccination of an entire population. This would be particularly important during times of acute need, when rapid responses are required during an emergent epidemic. To increase vaccine potency and efficiency, a variety of adjuvants have been developed. These either suffer from substantial toxicity or cannot be implemented because their mode of action is obscure [11]. Here we report that a profound increase in T-cell-mediated immune responses to several infectious viruses was achieved using an immunization strategy involving a combination of both the infectious agents and a recombinant vaccinia virus (VV) containing a TAP-gene construct. This unexpected observation suggests that recombinant TAP may be used as a novel adjuvant to increase vaccine efficacy and potency. Results TAP Expression Is Required for Antigen-Specific H-2 (Mouse Major Histocompatibility Complex) Kb Surface Expression In Vitro T2-Kb cells express H-2Kb but lack both TAP1 and TAP2 [12,13] and have a very low expression of MHC class I on the cell surface owing to inefficient antigen processing. In a CTL assay, vesicular stomatitis virus (VSV)–specific effectors were able to lyse T2-Kb cells coinfected with VV containing minigene for VSV-NP52–59 (VV-NP-VSV) and recombinant vaccinia virus carrying human TAP1 and TAP2 (VV-hTAP1,2) in a dose-dependent manner. This indicated that a functional TAP complex was formed by the VV-hTAP1,2 infection, leading to high levels of H-2Kb–vesicular stomatitis virus nucleoprotein (VSV-NP) surface expression (Figure 1A). In contrast, T2-Kb targets coinfected with VV-NP-VSV and VV encoding the plasmid PJS-5 (VV-PJS-5, negative control vector), or with VV-NP-VSV alone, were strongly resistant to lysis, and the level of lysis was similar to that in uninfected targets. These targets did not respond to increasing doses of effectors in the assay, indicating that the surface expression of H-2Kb–VSV-NP52–59 is below the threshold for CTL recognition. The levels of lysis in the targets coinfected with VV-NP-VSV and VV-PJS-5, or with VV-NP-VSV alone are small (3%–5%) relative to the targets infected with VV-hTAP1,2 which are associated with up to 45% lysis. Figure 1 VV-hTAP1,2 Restores Antigen Processing in the TAP-Deficient Cell Line T2-Kb and Increases Immune Responses to VSV (A) A standard chromium-release assay was performed to establish the ability of VV-hTAP1,2 to restore antigen processing in the TAP-deficient cell line T2-Kb. T2-Kb cells coinfected with VV-hTAP1,2 and VV-NP-VSV were used as targets, and splenocytes from VSV-infected mice were used as effectors. Targets coinfected with both VV-PJS-5 and VV-NP-VSV or infected with VV-NP-VSV alone, or uninfected cells, were used as negative controls for VV-hTAP1,2. (B) A standard chromium-release assay was performed to measure the ability of VV-hTAP1,2 to increase specific CTL activity in immunized mice. RMA cells pulsed with VSV-NP55–59 peptide were used as targets, and effectors were obtained from mice coinfected with VV-hTAP1,2 and low-dose VSV. Effectors from mice coinfected with VSV and VV-PJS-5 or a low dose of VSV alone were used as negative controls for the presence of VV-hTAP1,2 in the coinfections. Effectors from mice infected with a high dose of VSV demonstrated maximal CTL activity and were used as a positive control. (C) A standard chromium-release assay was used to confirm that the increase in immune responses was due to TAP-dependent transport of NP-VSV rather than to nonspecific effects of VV infection on antigen processing. RMA cells pulsed with VSV-NP55–59 peptide were used as targets, and effectors were obtained from mice coinfected with VV-hTAP1,2 and VV-NP-VSV. Effectors from mice infected with a high dose of VSV were used as positive controls for maximal CTL activity. Effectors from mice coinfected with VV-PJS-5 and VV-NP-VSV or from mice infected with VV-NP-VSV alone were negative controls for the presence of VV-hTAP1,2. Values represent the mean of triplicate measurements ± standard error of the mean. Increased TAP Expression Increases Immune Responses to VSV and Sendai Virus Infection In Vivo Splenocytes from mice coinfected with the low dose of VSV and VV-hTAP1,2 showed dramatic increases in CTL activity against RMA cell targets (Figure 1B). The response was specific to the expression of TAP rather than to the VV vector alone, since coinfection with VV-PJS-5 did not increase CTL responses to VSV. To exclude the possibility that the effect seen with VV-hTAP1,2 was the result of VV-dependent alteration of proteasome function interacting with TAP over-expression, VV-NP-VSV was used in the coinfections as the source of antigen instead of VSV [14]. The epitope generated by VV-NP-VSV does not require degradation by the proteasome, and therefore any increase in CTL activity can be attributed to TAP over-expression. Mice coinfected with VV-NP-VSV and VV-hTAP1,2 exhibited dramatic increases in vaccinia virus carrying the vesicular stomatitis virus nucleocapsid protein minigene (VSV-NP52–59)–specific CTL responses when compared to coinfection with VV-NP-VSV and VV-PJS-5 (Figure 1C). Another well-characterized CTL epitope, from Sendai virus nucleoprotein (SV-NP), was investigated to provide further evidence that TAP over-expression can augment antiviral responses. The infectious dose of Sendai virus (SV) required to achieve minimal and maximal CTL responses was also first determined by titration (data not shown). As was the case for VSV, mice coinfected with SV and VV-hTAP1,2 exhibited an increased specific immune response against SV-NP epitope when compared to responses elicited by the vector controls (Figure 2A). Figure 2 VV-hTAP1,2 Increases Antigen Presentation and Immune Responses to SV and VV in Mice (A) A standard chromium-release assay was used to determine the ability of VV-hTAP1,2 to increase immune responses to SV. RMA cells pulsed with SV-NP peptides were used as targets, and effectors were obtained from the mice coinfected with a low dose of SV and VV-hTAP1,2. The mice coinfected with a low dose of SV and VV-PJS-5 or with a low dose of SV alone were used as negative controls. Effectors from the mice infected with a high dose of SV were used as positive controls for maximal SV-specific CTL activity. (B) A standard chromium-release assay was used to determine the ability of VV-hTAP1,2 to stimulate VV-specific CTL responses. RMA cells infected with VV-PJS-5 were used as targets, and effectors were obtained from the mice vaccinated with a low dose of VV-hTAP1,2. Effectors from the mice vaccinated with an equivalent low dose of VV-PJS-5 were used as negative controls, and effectors from the mice vaccinated with a high dose of VV-PJS-5 were used as positive controls for maximal CTL activity. (C) A standard chromium-release assay was used to measure the ability of human TAP expression to increase antigen presentation in normal mouse splenocytes. Naïve splenocytes, which had been stimulated overnight with LPS (LPS blasts) and infected with VV-hTAP1,2, were used as targets for VV-specific effectors; VV-specific effectors were obtained from mice infected with VV-PJS-5. LPS blasts infected with VV-PJS-5 were used as negative controls. Values represent mean of triplicate measurements ± standard error of the mean. Increased TAP Expression Increases Immune Responses to VV Infection In Vivo The augmentation of a specific immune response against a virus by increasing TAP expression in APCs may require that viral infection and the over-expression of TAP occur in the same cell. To address this, we generated CTLs directed against antigen(s) derived from VV. VV-specific CTL activity in mice infected with a low dose of VV-hTAP1,2 was much greater than that from mice infected with an equivalent low dose of the control, VV-PJS-5 (Figure 2B). Increased TAP Expression Increases Endogenous Antigen Processing We examined whether TAP over-expression could increase endogenous antigen processing. A VV-specific CTL assay was used to compare H-2Kb and H-2Db VV-specific antigen processing in naïve splenocytes infected with VV-hTAP1,2 or VV-PJS-5. Naïve splenocyte targets infected with VV-hTAP1,2 were more susceptible to killing by VV-specific effectors than naïve splenocytes infected with VV-PJS-5 (Figure 2C). Increased TAP Expression Increases the Frequency of VSV-Specific CD8+ VSV-NP/Kb-specific tetramer analysis compared the proportion of splenocytes specific for H-2Kb VSV-NP52–59 among CD8+ splenocytes between VSV-infected mice and mice coinfected with VV-hTAP1,2 and VSV (Figure 3). Mice coinfected with VV-hTAP1,2 and a low dose of VSV elicited a higher frequency of VSV-NP52–59–specific CD8+ splenocytes (17.5% of CD8+ splenocytes) than mice coinfected with VV-PJS-5 and low-dose VSV (12.8%), or with a low dose of VSV alone (8.3%). The differences between a combined VV-PJS-5 and VSV low-dose vaccination, on the one hand, and a combined VV-hTAP1,2 and VSV low-dose vaccination, on the other hand, were highly significant (z-statistic = 4.701, p < 0.0001). The maximum VSV-specific CD8+ frequency (27.6%) was observed with the highest dose of VSV infection. Figure 3 Antigen-Specific Tetramer Staining Was Used to Determine T-Cell Responses in Coinfections with VV-hTAP1,2 and VSV The percentage of CD8+ splenocytes specific for H-2Kb–VSV-NP52–59 was determined by flow cytometry using double labeling with an anti-CD8+ antibody and a VSV-NP–specific tetramer. The value in the upper-right quadrant of the scatter-plots represents the percentage of CD8+ cells specific for H-2Kb–VSV-NP52–59 for mice infected with a low dose of VSV and VV-hTAP1,2. The mice coinfected with both VSV and VV-PJS-5 or with a low dose of VSV, or uninfected mice, were used as negative controls for VV-hTAP1,2. The mice infected with a high dose of VSV alone were used as a positive control. Peptide-Transport Activity and Human TAP Expression Human TAP1 protein was detected in splenocytes by immunoblotting (Figure 4A), and RT-PCR analysis showed that both human TAP1 and human TAP2 mRNA were present (Figure 4B). Human TAP1 mRNA levels were quantified with real-time RT-PCR in splenocytes from mice 4, 8, and 24 h after intraperitoneal (i.p.) infection. Calculations based on the threshold cycle number indicated that the abundance of human TAP1 mRNA was equal to mouse TAP1 mRNA 4 and 8 h after infection, but decreased to 2% of endogenous mouse TAP1 mRNA by 24 h after infection. Immunocytochemistry showed that 7% of total splenocytes were positive for human TAP1. Double staining for cell-specific antigens and human TAP1 showed that 3.0% of B cells, 2.4% of macrophages, and 1.9% of dendritic cells (DCs) were positive for human TAP1 expression (Figure 4C). A peptide-transport assay clearly showed that splenocytes from VV-hTAP1,2–infected mice had a higher capacity to transport a 125I-labeled peptide library into the lumen of the ER than splenocytes from normal mice or from the mice infected with VV-PJS-5 (Figure 4D). Figure 4 Human TAP Expression and Activity Was Determined in Splenocytes 24 h after the Mice Were Infected with VV-hTAP1,2 (A) Human TAP1 protein expression in mouse splenocytes was determined by Western blot. The mice infected with VV-PJS-5 were used as negative controls for human TAP1 expression. (B) The expression of human TAP1 and human TAP2 was detected by RT-PCR 24 h after the mice were infected with VV-hTAP1,2. The mice infected with VV-PJS-5 were negative for human TAP1 and TAP2. (C) Immunofluorescence visualized with confocal microscopy identified human TAP1 expression in antigen-presenting splenocytes isolated from mice 24 h after infection with VV-hTAP1,2. The mice infected with VV-PJS-5 were used as negative controls for human TAP1 expression (green fluorescence) (I). Cell-surface markers (red fluorescence) identified cell types. Representative images show the following cell types: (I) B cell from a mouse infected with VV-PJS-5 (negative control); (II) B cell that is positive for human TAP1; (III) macrophage that is positive for human TAP1; and (IV) DC that is positive for human TAP1. (D) ATP-dependent TAP activity was measured in splenocytes taken 24 h after the mice were infected with VV-hTAP1,2 or VV-PJS-5 (negative control). Active transport activity was measured in the presence or absence of ATP by a peptide-transport assay that determined the translocation of radioactive peptides from the cytosol into the ER. Normal uninfected mice, uninfected TAP−/− mice, and mice infected with VV-PJS-5 were used as negative controls when assessing the effect of VV-hTAP1,2 infections on peptide-transport activity. The bars represent the mean value ± standard error of the mean of triplicate measurements. The data are representative of the experiment performed in duplicate. Increased TAP Expression in DCs Increases MHC Class I–Restricted Presentation of Exogenous Antigens A subpopulation of splenocytes corresponding to the splenocyte DCs showed a 7.4% increase in total H-2Kb following infection with VV-hTAP1,2 when compared to V-PJS-5 (data not shown). This effect on DCs led us to investigate the cross-presentation of ovalbumin (OVA), an exogenously derived antigen. Treatment of DCs with VV-hTAP1,2 or recombinant vaccinia virus carrying mouse TAP1 (VV-mTAP1) significantly increased the amount of H-2Kb–SIINFEKL expression when compared to DCs infected with VV-PJS-5 (p < 0.01). The assay was repeated three times, with each assay showing statistically significant increases in H-2Kb–SIINFEKL and total H-2Kb expression by DCs infected with VV-hTAP1,2 or VV-mTAP1 (p < 0.01) (Figure 5A and 5B). DCs infected with VV-hTAP1,2 or VV-mTAP1 and incubated with OVA expressed significantly higher numbers of total H-2Kb complexes compared to VV-PJS-5–infected DCs (40% and 30% increase, respectively; p < 0.01) (Figure 5C and 5D). Figure 5 The Effect of VV-hTAP1,2 and VV-mTAP1 Infection on the Cross-Presentation Activity of OVA/SIINFEKL by Normal Spleen-Derived DCs DCs infected with VV-hTAP1,2 expressed greater (A) H-2Kb–SIINFEKL and (B) total H-2Kb than DCs infected with VV-PJS-5. DCs infected with VV-mTAP1 also expressed greater (C) H-2Kb–SIINFEKL and (D) total H-2Kb than DCs infected with VV-PJS-5. DCs infected with VV-PJS-5, but not incubated with OVA, served as negative controls for cross-presentation. The data are representative of the experiment performed in duplicate. VV-hTAP1,2–Vaccinated Mice Resist Lethal Viral Challenge A viral-challenge experiment determined whether TAP-dependent increases in immune function are significant in generating protective immune responses in vivo. Weight change was monitored in groups of mice, following vaccination with escalating doses of VV-hTAP1,2, VV-PJS-5, or PBS, and then administration of a lethal vaccinia virus Western Reserve strain (VV-WR) challenge [15–18]. Five out of six mice receiving the lowest VV-hTAP1,2 vaccine doses survived the challenge with minimal weight loss (less than 5%) and returned to normal, pre-challenge weight within 6 d. All the mice vaccinated with the intermediate and high doses of VV-hTAP1,2 survived challenge without weight loss. In contrast, the mice vaccinated with the lowest dose of VV-PJS-5 suffered significant morbidity (approximately 20% weight loss) and high mortality (four out of six mice died). The mice vaccinated with the intermediate dose of VV-PJS-5 also experienced high morbidity and one death. These mice were unable to regain weight to the pre-challenge level until 14 d after viral challenge. Mice receiving the highest dose of VV-PJS-5 were completely protected. All sham-vaccinated mice (PBS) were dead by 7 d post-challenge, consistent with a 1e5 plaque-forming unit (PFU) intranasal challenge (Figure 6A and 6B). VV-hTAP1,2 vaccination provided significantly greater protection than vaccination with VV-PJS-5 (p < 0.05) in a dose-dependent manner (p < 0.01). Mice vaccinated with VV-hTAP1,2 at the lowest doses were able to resist a lethal challenge equivalent to the highest vaccine dose of VV-PJS-5. This represents a 100-fold increase in the efficacy of VV-hTAP1,2 to generate a protective immune response compared with VV-PJS-5. Figure 6 A Viral-Challenge Experiment Was Used to Measure the Protection Provided by Low-Dose Vaccination with VV-hTAP1,2 (A) Three groups of mice were vaccinated with escalating doses of VV-hTAP1,2 (3e3 PFU, 3e4 PFU, 3e5 PFU) and were challenged 14 d later with a lethal dose of VV-WR (1e5 PFU). Percentage weight change was measured as an indication of death and morbidity. Three doses of low-dose VV were administered. (B) Three Groups of mice were vaccinated with escalating doses of VV-PJS-5 (3e3 PFU, 3e4 PFU, 3e5 PFU), and were challenged 14 d later with a lethal dose of VV-WR (1e5 PFU). These groups served as negative controls for the effect of VV-hTAP1,2 on protection from lethal viral challenge. Mice vaccinated with PBS served as positive controls for lethal viral challenge. Data points represent mean weight changes ± standard error of the mean (n = 6) recorded daily. Discussion Contrary to expectation, we observed that increasing TAP expression in mice augments the cellular immune response to a variety of viral pathogens including VSV, SV, and VV at infectious doses that normally do not generate significant CTL activity. These responses do not appear to be the result of a reversal of viral immuno-evasion mechanisms with respect to antigen presentation since VSV and SV are not known to down-regulate MHC class I surface expression. The correlation of immune responses with increasing levels of viral infection in mice reflects the fact that CTL priming requires a threshold level of relevant viral peptides to be expressed on the surface of APCs [19–22]. The increase in CTL activity in mice coinfected with VV-hTAP1,2 and with low infectious doses of VSV, SV, and VV-NP-VSV shows that this increase is dependent on the expression of TAP activity alone. It is unlikely that this augmentation is due to the more efficient translocation of VSV-derived peptides by human TAP and/or interspecies (human/mouse) TAP heterodimers compared to mouse TAP complexes. It has been reported that human TAP preferentially transports peptides containing either hydrophobic or positively charged amino-acids at their C-terminus, while mouse TAP is slightly more restrictive and favors peptides with hydrophobic amino-acids at their C-terminus [23]. The VSV-NP and SV-NP are two murine Kb-restricted epitopes that contain the same hydrophobic residue (leucine) at the C-terminus. The transport of these peptides by human TAP would compete with an additional peptide pool containing positively charged C-terminal residues. This might lead to a reduced amount of VSV-NP or SV-NP entering the ER lumen for surface presentation through human TAP heterodimers. In addition, the SV-NP epitope has an aromatic residue (phenylalanine) at the peptide position 1 (N-terminus), and this has a strong deleterious effect for human TAP binding [24] and for transport. Therefore, we conclude that the transport of SV-NP by human TAP would be no better than by mouse TAP. For interspecies TAPs, the transport of peptides is restricted to those with hydrophobic C-terminal residues, similar to mouse TAP [25]. This would imply that when mouse TAP1 or TAP2 associates with its human TAP counterpart, they play a dominant role in selecting the peptides for transport. Furthermore, once the transport-permissive peptide, for example SV-NP, binds to interspecies TAPs, the phenylalanine residue at the N-terminus that is in contact with the human TAP counterpart may limit its binding capacity and, therefore, its transport. For these reasons, we conclude that the augmentation of the CTL responses against viruses in our experiments is justified by TAP over-expression rather than by the increased efficiencies of interspecies TAP heterodimers. The priming of T cells requires cell-to-cell contact, and therefore APCs adjacent to T cells play a critical role [2,9]. The increase in VSV-specific CD8+ cells observed with VV-hTAP1,2 coinfections indicates a TAP-dependent increase in APC cross-priming and cross-presentation activity and is explained by an increase in TAP expression and peptide-transport activity in the APCs of the spleen. To reconcile the increase in peptide-transport activity observed in the translocation assays with the frequency of human TAP-positive splenocytes, we estimate that the human TAP-positive splenocytes need to express 12 to 17 times more human TAP1 mRNA than endogenous mouse TAP1. This was confirmed by the high expression of the human TAP gene early in the infection. DCs generate virus-specific CTLs via cross-presentation of exogenously acquired viral antigens in the context of MHC class I molecules [26]. This is achieved by a TAP-dependent process although additional TAP-independent pathways have been described recently [27–29]. Increased expression of mouse TAP1 appeared to be as effective as increased expression of human TAP1 and TAP2 in raising the level of MHC class I complexes (H-2Kb–SIINFEKL) on the cell surface. Expression of TAP1 alone has been shown to stabilize TAP2 protein and TAP2 mRNA in TAP-deficient cells and could explain the effectiveness of TAP1 alone [30]. We conclude that supra-normal expression of TAP increases endogenous antigen processing (see Figure 2C) and the levels of both total and cross-presented MHC class I antigens on the surface of DCs, thereby leading to greater CTL responses in vivo. Under normal conditions, it is estimated that only one-third of all TAP molecules translocate peptides actively. During an acute viral infection, however, TAP activity increases significantly owing to the rapid increase in the available intracellular peptide pool [31]. Therefore, the supply of peptides to MHC class I molecules is theoretically the limiting factor in antigen presentation. It is known that increasing the delivery of peptides into the ER by the artificial creation of a signal-sequence peptide fusion increases antigen presentation and immune responses [32]. However, these experiments bypass TAP altogether, and the relative amounts of peptide generated in the two pathways are difficult to equilibrate and, therefore, to compare. Our results indicate that during a viral infection there is competition between self and viral peptides in the MHC class I binding-peptide pool, limiting the amount of viral epitopes reaching the lumen of the ER and subsequently the cell surface. Increased TAP activity leads to more viral epitopes presented on the cell surface by MHC class I, resulting in increased immune responses. In DCs, over-expression of TAP has the additional effect of increasing the cross-presentation pathway for MHC class I, a pathway believed to be unique to these cells. A framework for how cross-presentation may operate has been published recently [26,27,33,34]. Endocytosed or phagocytosed exogenous antigens gain access to one or more types of vesicles where loading of antigenic peptide onto nascent MHC class I molecules is thought to occur. MHC class I molecules gain access to the endolysosomal compartment by virtue of a tyrosine-based sorting signal in their cytoplasmic domains [26]. It has been suggested that TAP molecules reside in this compartment and that a fusion event with the ER may make antigen processing more efficient for both exogenous and endogenous antigens. Over-expression of TAPs in the endogenous and the exogenous antigen-processing compartments of DCs could increase the efficiency of both pathways, leading to enhanced specific immune responses without stimulation of detectable autoimmune responses (D. Waterfield, unpublished data). The presence of TAP genes in VV provides protection against lethal viral challenges at 100-fold lower amounts of inocula than VV without TAP genes. The inclusion of TAP in vaccination regimens acts as a gene-based adjuvant to boost immune responses against viral antigens, thereby allowing for reduced vaccine doses. Increased immune responses in response to low-dose vaccinations are desirable in the elderly and the very young, in whom immune systems may be compromised [35]. An additional advantage of including TAP as an adjuvant is its ability to increase peptide transport of a number of immunogenic peptides simultaneously, thereby aiding in the delivery of diverse peptides for binding to most HLA (human major histocompatibility complex) class I alleles expressed in the immunized population. TAP could be used as an adjuvant in peptide vaccines, but its use does not have to be restricted to viral vectors. For example, it could also be injected in other forms, such as in DNA plasmids attached to gold particles, or in any other system that inserts the TAP complex directly into the cell's protein-processing pathway [36]. Finally, the use of TAP as an adjuvant has the advantage that we have a solid intellectual understanding of its mechanism of action. This appears to be lacking in the case of many other generalized adjuvants [37]. We have shown here, in an animal model, that TAP over-expression can augment cell-mediated immunity against the cowpox virus (vaccinia), a close relative of smallpox (variola). It is conceivable that this approach could have applications in augmenting responses against smallpox in humans. Recently, clinical trials for some promising HIV-vaccine candidates have been suspended because of poor cellular immune responses [38]. Inclusion of TAP in such vaccines may be able to improve their efficacy. Future clinical experiments will help to further establish whether the inclusion of TAP in vaccine regimens has advantages over existing protocols and whether other components of the intracellular antigen-processing pathway(s) are also limiting in healthy individuals. Nonetheless, the approach that has been discovered is novel and may have tremendous potential for vaccination of humans and animals against a variety of infectious diseases. Materials and Methods Animals, cells, and viruses. The mouse strain C57BL/6 (H-2Kb) was obtained from Jackson Laboratories (Bar Harbor, Maine, United States) and housed at the Biotechnology Breeding Facility (University of British Columbia) according to the guidelines of the Canadian Council on Animal Care. Mice (5–12 wk of age) were screened for pathogens with the Murine ImmunoComb Test (Charles River Laboratories, Wilmington, Massachusetts, United States). VSV, Indiana Strain (a gift from Frank Tufaro, University of British Columbia) was cultured on vero cells (American Type Culture Collection [ATCC], Manassas, Virginia, United States). VV-hTAP1,2 and VV-mTAP1 (both gifts from John Yewdell, NIH/NIAID, Bethesda, Maryland, United States), VSV-NP52–59, VV-NP-VSV, VV-PJS-5 (PJS-5 a gift from Bernard Moss, NIH/NIAID), and VV Western Reserve strain (a gift from Shirley Gillam, University of British Columbia) were cultured on CV-1 cells [5,39,40]. SV was purchased from ATCC. CV-1 cells were cultured in DMEM/10% FBS and vero cells, and RMA cells were cultured in RPMI/10% FBS. VSV and VV infectious units were determined by tissue-culture infectious-dose assay (TCID50) or by standard plaque assay. SV infectious units were determined by chicken-egg infectious-dose (CEID50) as indicated on the label. T2 cells negative for TAP1 and TAP2 and transfected with mouse H-2Kb were a gift from Peter Cresswell, (Yale University, New Haven, Connecticut, United States). Generation of VSV, SV, and VV-specific effectors. H-2Kb–restricted virus-specific CTLs were generated by i.p. injection of mice with VSV (low-dose [3.6e4 TCID50] or high-dose [1.5e7 TCID50]), SV (low-dose [1.6e5 CEID50] or high-dose [1.6e7 CEID50]), VV-PJS-5 (low-dose [3e4 PFU] or high-dose [5e6 PFU]), VV-hTAP1,2 (3e4 PFU), and VV-NP-VSV (3e6 PFU). Titrations determined the infectious doses of VSV and SV that would generate a minimum (low dose) and maximum (high dose) CTL response against peptide-pulsed RMA targets. Splenocytes were harvested 6 d after infection and cultured (RPMI-1640/10% FBS complete media) for 2–3 d to eliminate nonspecific natural killer cell killing. For VSV infection alone or via coinfection with VV-hTAP1,2 or VV-PJS-5, splenocytes were incubated without peptide. For infections with SV or VV-NP-VSV, splenocytes were stimulated with SV peptide (1 μM SV-NP324–332 peptide, FAPGNYPAL) or VSV-NP peptide (1 μM VSV-NP52–59 peptide, RGYVYQGL). VV-hTAP1,2 or VV-PJS-5–infected mouse splenocytes were restimulated in vitro for 3 d with irradiated syngeneic splenocytes infected with VV-PJS-5 (3.4e7 PFU/1e8 syngeneic splenocytes). Cytotoxicity assays. Cytotoxic activity was measured in standard 4-h 51Cr-release assays using T2-Kb cells, RMA cells, or naïve splenocytes as targets. T2-Kb targets were infected with VV-NP-VSV alone or in combination with VV-hTAP1,2 or VV-PJS-5 (multiplicity of infection [MOI] = 10) for 6 h. The RMA target cells were pulsed with VSV-NP52–59 peptide (5–25 μM) or SV-NP324–332 peptide (5–25 μM) for the relevant CTL assay. For VV antigen-specific killing, the RMA targets were infected overnight with VV-PJS-5 (MOI = 0.34). For the measurement of endogenous antigen processing, targets were generated by the in vitro stimulation of naïve splenocytes (2e7 cells) for 2 d with lipopolysaccharide (LPS) (Escherichia coli J5 LPS [1 μg/ml], Calbiochem, San Diego, California, United States), followed by overnight infection with either VV-PJS-5 or VV-hTAP1,2 (5e6 PFU). Detection of CD8+/VSV-NP52–59–specific H-2Kb–restricted splenocytes. The frequency of CD8+/VSV-NP52–59–specific splenocytes was measured in two separate experiments using a H-2Kb–VSV-NP52–59 epitope-specific tetramer conjugated to phycoerythrin (Custom iTAg MHC Tetramer, Beckman Coulter Canada, Mississauga, Ontario, Canada) and a fluorescein-conjugated rat anti-mouse CD8a (Ly-2) monoclonal antibody (clone 53–6.7) after Fc receptor block (clone 2.4G2, Mouse Fc Block), (BD Biosciences Pharmingen, San Diego, California, United States). Mice were coinfected with VV-PJS-5 (3e4 PFU) plus low-dose VSV, or VV-hTAP1,2 (3e4 PFU) plus low-dose VSV, or infected with low-dose VSV, or high-dose VSV. Six days after infection, CD8+/VSV-NP52–59–specific fluorescence as a percentage of total CD8+ fluorescence was determined by flow cytometry. Detection of human TAP expression in splenocytes. Human TAP1 expression in splenocytes from the mice infected with VV-hTAP1,2 was determined by SDS-PAGE and Western blot. The blots were probed for human TAP1 with rabbit anti-human TAP1 antiserum (Stressgen Biotechnologies, Victoria, British Colombia, Canada) and visualized by enhanced chemiluminescence (Amersham Biosciences, Little Chalfont, United Kingdom). Quantitative RT-PCR for human and mouse TAP1 in splenocytes. RT-PCR was used to detect human TAP1 and TAP2, and mouse TAP1 and TAP2, in spleens 24 h after infection. In addition, a time-course quantification of human TAP1 and mouse TAP1 was performed using quantitative real-time PCR (QRT-PCR). Total RNA was extracted (RNeasy Midi Kit, Qiagen, Valencia, California, United States) from mouse spleens 4 h, 8 h, and 24 h after i.p. infection with VV-hTAP1,2 (3e4 PFU). QRT-PCR reactions (Sigma-Genosys Canada, Oakville, Ontario, Canada) were performed in duplicate using a Light Cycler (Roche Diagnostics, Mannheim, Germany) for mouse TAP1 and for human TAP1, plus a ribosomal small subunit S15. The sequences of the primer pairs used in the RT-PCR and the QRT-PCR are listed in Table 1. Table 1 Sequences of the Primer Pairs Used in the RT-PCR and the QRT-PCR The threshold cycle (CT) above background was determined for mouse TAP1 and human TAP1 and normalized to the lowest S15 CT value among the reactions. The differences in CT were used to calculate the abundance of human TAP1 relative to mouse TAP1. Relative abundance = 2 (CTmTAP1–CThTAP1). The CT values represent the average of three mice per time point. Visualization of human TAP expression in splenocytes. Visualization of human TAP expression in spleen-derived antigen presentation cells from mice infected with VV-hTAP1,2 or VV-PJS-5 was performed using confocal fluorescence microscopy. Splenocytes were double labeled with rabbit anti-human TAP1 antiserum (Stressgen Biotechnologies) and one of the following cell-surface markers: rat anti-mouse B220 (B cell marker, BD Biosciences Pharmingen), rat anti-mouse MAC-1 (macrophage marker, BD Biosciences Pharmingen), or rat anti-mouse NLDC-145 antibodies (DC marker) (gift from Ralph Steinman, the Rockefeller University, New York, New York, United States). The presence of human TAP1 was determined in approximately 300 cells per surface marker in 20 randomly chosen fields. Splenocytes from VV-PJS-5–infected mice were used as negative controls. Transport activity of human TAP in mouse splenocytes. TAP heterodimer activity, in the presence or absence of adenosine triphosphate (ATP), was detected by a streptolysin-O–mediated peptide-transport assays in splenocytes harvested 24 h after mice were infected with VV-hTAP1,2 or VV-PJS-5 using a radio-iodinated peptide library containing a glycosylation site (NXT) (125I, specific activity, 10 Ci/mmol) [41,42]. Splenocytes from uninfected normal mice and TAP−/− mice were used as controls. Cross-presentation of OVA peptides (SIINFEKL) by DCs. Splenic DCs were isolated using CD11c+ magnetic beads run through an AutoMacs (Miltenyi Biotec, Auburn, California, United States) after enrichment by centrifugation (Ficoll Paque Plus, Amersham Biosciences). DCs were infected with VV-PJS-5, VV-mTAP1, or VV-hTAP1,2 (3e6 cells/1e4 PFU) for 2 h, followed by incubation with GM-CSF (10 ng/ml complete RPMI 1640) and OVA (10 mg/ml) (Worthington Biochemical, Lakewood, New Jersey, United States) or PBS (18 h at 37 °C). DCs stained with phycoerythrin–conjugated rat anti-mouse CD11c antibody, Fc blocker (Fcγ III/II receptor), and either FITC-conjugated anti-H-2Kb antibody (clone AF6–88.5) (BD Biosciences Pharmingen) or 25.D1.16 antibody specific for Kb/SIINFEKL (a gift from R. Germain, NIH/NIAID) [43]. Flow cytometry was used to quantify H-2Kb and H-2Kb–SIINFEKL complexes. VV-WR–challenge experiments. VV-hTAP1,2 and VV-PJS-5 viruses were demonstrated to replicate equally in cell culture. Forty-two mice were randomized into seven groups (six mice per cage) and were vaccinated with three different doses of VV-hTAP1,2 or VV-PJS-5 (3e3, 3e4, and 3e5 PFU in 300 μl PBS i.p.) or PBS. Fourteen days later, mice were weighed and challenged with a lethal dose of VV-WR (1e5 PFU in 20 μl of clarified cell lysate delivered intranasally) under isoflurane anesthesia. Weight was recorded daily for 14 d, and any mouse falling below 25% of pre-challenge weight was euthanized. Mean weight going forward was calculated from the remaining survivors. Statistical analysis. The z-statistic was calculated to determine the statistical significance of the differences in the proportions of H-2Kb, VSV-NP, and CD8+ cells generated by the vaccination protocols. A two-way ANOVA, after square-root transformation of the data, was used to analyze the main effects of dose and the recombinant VV on mouse weight 5 d after VV-WR challenge. Bonferroni procedure corrected p-values for multiple comparisons. A chi-square test (univariate comparison, using FlowJo 3.7.1 [Treestar, http://www.treestar.com]) compared flow-cytometry histograms for differences in total H-2Kb or H-2Kb–SIINFEKL complexes. p-Value of < 0.01 (99% confidence interval) was considered significant, and T(X) > 10 was empirically determined as a cut-off value. Supporting Information Accession Numbers The Swiss-Prot (http://www.ebi.ac.uk/swissprot) accession numbers for the proteins discussed in this paper are calnexin (BC040244), calreticulin (NM 007591), LMP2 (U22919), LMP7 (U22031), LMP10 (U77784), MHC class I heavy chain (V00747) and β2-microglobulin (NM 009735), TAP1 (NM 013683), TAP2 (NM 011530), and tapasin (A7316613). The GenBank (http://www.ncbi.nlm.nih.gov/Genbank) accession numbers for the strains discussed in this paper are mouse strain C57BL/6 (V00746), recombinant VV carrying human TAP1 (NM 000593) and TAP2 (NM 018833) genes, SV (X00087), VSV, Indiana Strain (J02428), and VV Western Reserve strain (AY243312). The authors thank Terry Pearson (University of Victoria, Victoria, British Columbia, Canada) and Karl Erik Hellstrom (University of Washington, Seattle, Washington, United States) for their comments on the manuscript. The authors acknowledge the Canadian Institutes of Health Research, the National Cancer Institute of Canada, the CANVAC Network of Excellence, and GeneMax (Vancouver, British Columbia, Canada) for providing financial support. Competing interests. The University of British Columbia and its scientists (TZV, QZ, JA, and WAJ) received equity in GeneMax as part of a technology license agreement between the University of British Columbia and GeneMax. WAJ is a director of GeneMax. GeneMax provided partial financial support for this study. Author contributions. WAJ conceived and designed the experiments. TZV, QZ, JA, SSC, GB, AM, JT, MMT, KBC, DW, and AJ performed the experiments. TZV and WAJ analyzed the data and wrote the paper. Abbreviations APCantigen-presenting cell ATPadenosine triphosphate CEID50chicken-egg infectious-dose CTthreshold cycle CTLcytotoxic lymphocyte DCdendritic cell ERendoplasmic reticulum H-2mouse major histocompatibility complex HLAhuman major histocompatibility complex i.p.intraperitoneal LMPlow molecular-weight polypeptide LPSlipopolysaccharide MHCmajor histocompatibility complex MOImultiplicity of infection OVAovalbumin PFUplaque-forming unit QRT-PCRquantitative real-time PCR SVSendai virus SV-NPSendai virus nucleoprotein TAPtransporter associated with antigen processing TCID50tissue-culture infectious-dose assay VSVvesicular stomatitis virus VSV-NPvesicular stomatitis virus nucleoprotein VSV-NP52–59vesicular stomatitis virus nucleocapsid protein minigene VVvaccinia virus VV-hTAP1,2recombinant vaccinia virus carrying human TAP1 and TAP2 VV-mTAP1recombinant vaccinia virus carrying mouse TAP1 VV-NP-VSVvaccinia virus containing minigene for VSV-NP55–59 VV-PJS-5vaccinia virus encoding the plasmid PJS-5 VV-WRvaccinia virus Western Reserve strain ==== Refs References Zinkernagel RM Doherty PC 1979 MHC-restricted cytotoxic T cells: Studies on the biological role of polymorphic major transplantation antigens determining T-cell restriction-specificity, function, and responsiveness Adv Immunol 27 51 177 92183 Townsend AR Rothbard J Gotch FM Bahadur G Wraith D 1986 The epitopes of influenza nucleoprotein recognized by cytotoxic T lymphocytes can be defined with short synthetic peptides Cell 44 959 968 2420472 Suh WK Cohen-Doyle MF Fruh K Wang K Peterson PA 1994 Interaction of MHC class I molecules with the transporter associated with antigen processing Science 264 1322 1326 8191286 Frisan T Zhang QJ Levitskaya J Coram M Kurilla MG 1996 Defective presentation of MHC class I-restricted cytotoxic T-cell epitopes in Burkitt's lymphoma cells Int J Cancer 68 251 258 8900437 Alimonti J Zhang QJ Gabathuler R Reid G Chen SS 2000 TAP expression provides a general method for improving the recognition of malignant cells in vivo Nat Biotechnol 18 515 520 10802618 Deverson EV Gow IR Coadwell WJ Monaco JJ Butcher GW 1990 MHC class II region encoding proteins related to the multidrug resistance family of transmembrane transporters Nature 348 738 741 1979660 Kelly A Powis SH Glynne R Radley E Beck S 1991 Second proteasome-related gene in the human MHC class II region Nature 353 667 668 1922385 Norbury CC Princiotta MF Bacik I Brutkiewicz RR Wood P 2001 Multiple antigen-specific processing pathways for activating naive CD8+ T cells in vivo J Immunol 166 4355 4362 11254689 Ortmann B Androlewicz MJ Cresswell P 1994 MHC class I/beta 2-microglobulin complexes associate with TAP transporters before peptide binding Nature 368 864 867 8159247 Belyakov IM Earl P Dzutsev A Kuznetsov VA Lemon M 2003 Shared modes of protection against poxvirus infection by attenuated and conventional smallpox vaccine viruses Proc Natl Acad Sci U S A 100 9458 9463 12869693 Takahashi H Takeshita T Morein B Putney S Germain RN 1990 Induction of CD8+ cytotoxic T cells by immunization with purified HIV-1 envelope protein in ISCOMs Nature 344 873 875 2184369 Salter RD Cresswell P 1986 Impaired assembly and transport of HLA-A and -B antigens in a mutant TxB cell hybrid EMBO J 5 943 949 3522223 Spies T DeMars R 1991 Restored expression of major histocompatibility class I molecules by gene transfer of a putative peptide transporter Nature 351 323 324 2034277 Smith GL Symons JA Khanna A Vanderplasschen A Alcami A 1997 Vaccinia virus immune evasion Immunol Rev 159 137 154 9416508 Hooper JW Custer DM Schmaljohn CS Schmaljohn AL 2000 DNA vaccination with vaccinia virus L1R and A33R genes protects mice against a lethal poxvirus challenge Virology 266 329 339 10639319 Brandt TA Jacobs BL 2001 Both carboxy- and amino-terminal domains of the vaccinia virus interferon resistance gene, E3L, are required for pathogenesis in a mouse model J Virol 75 850 856 11134298 Smee DF Bailey KW Sidwell RW 2001 Treatment of lethal vaccinia virus respiratory infections in mice with cidofovir Antivir Chem Chemother 12 71 76 11437324 Reading PC Smith GL 2003 Vaccinia virus interleukin-18-binding protein promotes virulence by reducing gamma interferon production and natural killer and T-cell activity J Virol 77 9960 9968 12941906 Bellone M Iezzi G Manfredi AA Protti MP Dellabona P 1994 In vitro priming of cytotoxic T lymphocytes against poorly immunogenic epitopes by engineered antigen-presenting cells Eur J Immunol 24 2691 2698 7525300 Momburg F Neefjes JJ Hammerling GJ 1994 Peptide selection by MHC-encoded TAP transporters Curr Opin Immunol 6 32 37 8172678 Uebel S Meyer TH Kraas W Kienle S Jung G 1995 Requirements for peptide binding to the human transporter associated with antigen processing revealed by peptide scans and complex peptide libraries J Biol Chem 270 18512 18516 7543103 Motta I Lone YC Kourilsky P 1998 In vitro induction of naive cytotoxic T lymphocytes with complexes of peptide and recombinant MHC class I molecules coated onto beads: Role of TCR/ligand density Eur J Immunol 28 3685 3695 9842911 Armandola EA Momburg F Nijenhuis M Bulbuc N Fruh K 1996 A point mutation in the human transporter associated with antigen processing (TAP2) alters the peptide transport specificity Eur J Immunol 26 1748 1755 8765016 van Endert PM Riganelli D Greco G Fleischhauer K Sidney J 1995 The peptide-binding motif for the human transporter associated with antigen processing J Exp Med 182 1883 1895 7500034 Momburg F Roelse J Howard JC Butcher GW Hammerling GJ 1994 Selectivity of MHC-encoded peptide transporters from human, mouse and rat Nature 367 648 651 8107849 Lizee G Basha G Tiong J Julien JP Tian M 2003 Control of dendritic cell cross-presentation by the major histocompatibility complex class I cytoplasmic domain Nat Immunol 4 1065 1073 14566337 Guermonprez P Saveanu L Kleijmeer M Davoust J Van Endert P 2003 ER-phagosome fusion defines an MHC class I cross-presentation compartment in dendritic cells Nature 425 397 402 14508489 Chen L Jondal M 2004 Endolysosomal processing of exogenous antigen into major histocompatibility complex class I-binding peptides Scand J Immunol 59 545 552 15182249 Chen L Jondal M 2004 Alternative processing for MHC class I presentation by immature and CpG-activated dendritic cells Eur J Immunol 34 952 960 15048705 Seliger B Wollscheid U Momburg F Blankenstein T Huber C 2001 Characterization of the major histocompatibility complex class I deficiencies in B16 melanoma cells Cancer Res 61 1095 1099 11221838 Reits EA Vos JC Gromme M Neefjes J 2000 The major substrates for TAP in vivo are derived from newly synthesized proteins Nature 404 774 778 10783892 Bacik I Cox JH Anderson R Yewdell JW Bennink JR 1994 TAP (transporter associated with antigen processing)-independent presentation of endogenously synthesized peptides is enhanced by endoplasmic reticulum insertion sequences located at the amino- but not carboxyl-terminus of the peptide J Immunol 152 381 387 8283027 Ackerman AL Kyritsis C Tampe R Cresswell P 2003 Early phagosomes in dendritic cells form a cellular compartment sufficient for cross presentation of exogenous antigens Proc Natl Acad Sci U S A 100 12889 12894 14561893 Houde M Bertholet S Gagnon E Brunet S Goyette G 2003 Phagosomes are competent organelles for antigen cross-presentation Nature 425 402 406 14508490 Melnick JL 1989 Viral vaccines: Achievements and challenges Acta Virol 33 482 493 2576588 Dertzbaugh MT 1998 Genetically engineered vaccines: An overview Plasmid 39 100 113 9514708 Singh M O'Hagan D 1999 Advances in vaccine adjuvants Nat Biotechnol 17 1075 1081 10545912 Kahn P 2002 NIH drops plans for phase III trials IAVI Rep 6(1) 1 16 Russ G Esquivel F Yewdell JW Cresswell P Spies T 1995 Assembly, intracellular localization, and nucleotide binding properties of the human peptide transporters TAP1 and TAP2 expressed by recombinant vaccinia viruses J Biol Chem 270 21312 21318 7673167 Norbury CC Princiotta MF Bacik I Brutkiewicz RR Wood P 2001 Multiple antigen-specific processing pathways for activating naive CD8+ T cells in vivo J Immunol 166 4355 4362 11254689 Androlewicz MJ Anderson KS Cresswell P 1993 Evidence that transporters associated with antigen processing translocate a major histocompatibility complex class I-binding peptide into the endoplasmic reticulum in an ATP-dependent manner Proc Natl Acad Sci U S A 90 9130 9134 8415666 Heemels MT Schumacher TN Wonigeit K Ploegh HL 1993 Peptide translocation by variants of the transporter associated with antigen processing Science 262 2059 2063 8266106 Porgador A Yewdell JW Deng Y Bennink JR Germain RN 1997 Localization, quantitation, and in situ detection of specific peptide-MHC class I complexes using a monoclonal antibody Immunity 6 715 726 9208844	16389301	PMC1323471	CC BY	2021-01-05 12:11:25	no	PLoS Pathog. 2005 Dec 30; 1(4):e36	utf-8	PLoS Pathog	2,005	10.1371/journal.ppat.0010036	oa_comm
==== Front PLoS PathogPLoS PathogppatplpaplospathPLoS Pathogens1553-73661553-7374Public Library of Science San Francisco, USA 1673870510.1371/journal.ppat.001004005-PLPA-OP-0197plpa-01-04-08OpinionsInfectious DiseasesMicrobiologyStatisticsNoneYeast and FungiCoping with Multiple Virulence Factors: Which Is Most Important? OpinionMcClelland Erin E Bernhardt Paul Casadevall Arturo To whom correspondence should be addressed. E-mail: [email protected] E. McClelland and Arturo Casadevall are in the Department of Medicine, Albert Einstein College of Medicine, Bronx, New York, United States of America. Paul Bernhardt is in the Department of Psychology, Westminster College, Salt Lake City, Utah, United States of America. 12 2005 30 12 2005 1 4 e40Copyright: © 2005 McClelland 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. Citation:McClelland EE, Bernhardt P, Casadevall A (2005) Coping with multiple virulence factors: Which is most important? PLoS Pathog 1(4): e40. ==== Body Despite being a time when genomics and proteomics are becoming popular modes of scientific inquiry, most microbe-centric researchers continue to use reductionism to study virulence by identifying the microbial characteristics associated with virulence and characterizing each independently. Reductionism is a critical and remarkably powerful tool for determining the contribution of a single virulence factor to the overall virulence phenotype. However, the approach has the following limitation: most pathogenic microbes possess numerous attributes that contribute to virulence, and this virulence is a microbial trait that is expressed only in a susceptible host. Unfortunately, it is often very difficult, if not impossible with current technology, to study combinations of virulence factors simultaneously, especially at the molecular level. Hence, new tools are needed to study the larger picture of virulence in a pathogenic organism and to better understand how the host and microbe interact. And while computer modeling, genomics, and proteomics contribute much to the larger picture of virulence, a more accessible method is available via multivariate statistics. Consider two microbes: Cryptococcus neoformans and Bacillus anthracis. For C. neoformans, at least a half-dozen cellular attributes have been associated with virulence, including the polysaccharide capsule, melanin production, phospholipase secretion, mating factor, laccase, and urease [1]. Similarly, for B. anthracis the poly-D-glutamic acid capsule, lethal toxin, edema factor, and anthrolysin are each associated with the virulence phenotype [2,3]. Since immune responses to virulence factors often negate the virulence phenotype, vaccines often target virulence factors, as is the case for both C. neoformans and B. anthracis. In addition, the highly successful vaccines to Streptococcus pneumoniae and Haemophilus influenzae type B elicit antibodies to their polysaccharide capsules. In the same way, toxoid vaccines protect against tetanus and diphtheria by eliciting neutralizing antibody responses. Consequently, an investigator may want to prioritize the relative importance of virulence factors that contribute to the overall virulence phenotype, especially when designing new vaccines or antimicrobial drugs. However, to our knowledge, no methods have been developed to accomplish this. Fortunately, the goal of prioritizing virulence factors for an individual microbe is similar to problems that have been encountered and solved in other disciplines where investigators have had to confront phenomena caused by multiple components. Sophisticated statistical methods have been developed to approach these problems and are applicable to the problem of microbial virulence. A commonly used statistical tool is multivariate linear regression analysis (MLRA), which has found a myriad of uses in the social sciences, biology, and medicine. For example, MLRA has been used to study aspects of school readiness such as the prediction of learning-related skills in children [4], factors that contribute to smoking in youths [5], the degree and location of white matter changes in patients with Alzheimer disease [6], and whether hand lead contamination is associated with blood lead contamination [7]. MLRA simultaneously assesses the relationships of many independent variables to one dependent variable, and can easily be used to examine the relative contribution of microbial virulence factors to virulence. In fact, MLRA can be used to rank virulence factors in disease importance. There are three different kinds of MLRA: standard multiple regression, sequential (hierarchical) regression, and statistical regression [8]. In standard multiple regression, all independent variables are entered together so that the relative contribution of each independent variable to the dependent variable is assessed at the same time. Thus, standard regression illustrates how much of the dependent variable is explained by each of the independent variables at once. In hierarchical regression, the independent variables are entered in different steps in a specific order, with the order of entry resulting from theoretical or logical importance. Thus, hierarchical regression allows the investigator to examine the unique contribution of each independent predictor variable to the variance in the dependent variable, while taking into account the contribution of previously identified independent variables. In statistical regression, the independent variables are entered or removed in different steps, in an order that is specified by statistical criteria. Thus, statistical regression is useful for selecting which independent variables best predict the dependent variable when there is no theoretical rationale for a priori prioritization [8]. Depending on how much information is available on certain microbial virulence factors, and the focus of the research question, investigators can use any or all of the different kinds of MLRA. One limitation of MLRA is the need for an adequate sample size. A general rule of thumb is that ten data points (or microbial strains) are needed for every independent variable. So if the microbe of choice has three virulence factors, a sample size of at least 30 strains demonstrating some variation in virulence and virulence factor expression are needed to ascertain the relative contribution of each virulence factor to the overall virulence phenotype. However, Maxwell estimates that this rule of thumb is overly optimistic [9]. A more conservative estimate is 20–30 data points for every independent variable, based on the correlation between the independent variables (EEM, PB, and AC, unpublished data). More data points are needed if the independent variables are more correlated to each other. In addition, the sample size required is dependent on the question asked. A smaller sample size is needed if the goal is to establish which virulence factor contributes most to virulence (a large effect, such as which virulence factor is the most important candidate for drug and vaccine development research). However, a much larger sample size is needed if the investigator is looking for smaller effects, such as how much each microbial virulence factor contributes to virulence. Another consideration is the measure of virulence that will be used as the dependent variable. Such measures could include survival, microbial numbers in tissue, as well as different measures of host damage and the immune response [10]. Alternatively, strains could be tested for virulence in alternate hosts such as C. elegans and amoeba, since many microbial virulence factors can induce damage in multiple hosts and may be evolutionarily conserved [11]. The use of this method requires knowledge of statistics or collaborations with statisticians. As the average microbiologist may not be statistically trained, and statisticians that are familiar with experimental biological methods are often few and far between, it is our opinion that this could be addressed by requiring a statistics course for any graduate program, including a microbiology PhD program or medical school. Not only will every scientist benefit from statistical training with regard to experimental design and planning for statistical power, but because statistics is not often stressed in certain scientific disciplines, many journals are now requiring that statistical analysis be part of the methods section of submitted manuscripts. Thus, statistical training is an increasingly important requirement for every scientist. We have proposed that for microbes with multiple virulence factors, MLRA can be used to explore their contribution to virulence, with the caveat that investigators need to use an appropriate sample size for confident predictions. Information gleaned from estimates of the relative contribution of virulence factors to the larger picture of virulence could be exploited for selecting targets for drug design and vaccines. India ink of C. neoformans Abbreviation MLRAmultivariate linear regression analysis ==== Refs References Casadevall A Perfect JR 1998 Cryptococcus neoformans Washington (DC) ASM Press 541 p. Oncu S Sakarya S 2003 Anthrax—An overview Med Sci Monit 9 RA276 RA283 14586293 Shannon JG Ross CL Koehler TM Rest RF 2003 Characterization of anthrolysin O, the Bacillus anthracis cholesterol-dependent cytolysin Infect Immun 71 3183 3189 12761097 McClelland MM Morrison FJ Holmes DL 2000 Children at-risk for early academic problems: The role of learning-related social skills Early Child Res Q 15 307 329 Dunn CL Pirie PL 2005 Empowering youth for tobacco control Am J Health Promot 20 7 10 16171155 Bracco L Piccini C Moretti M Mascalchi M Sforza A 2005 Alzheimer's Disease: Role of size and location of white matter changes in determining cognitive deficits Dement Geriatr Cogn Disord 20 358 366 16192726 Sato M Yano E 2005 The association between lead contamination on the hand and blood lead concentration: A workplace application of the sodium sulphide (Na(2)S) test Sci Total Environ Epub ahead of print. Tabachnick BG Fidell LS 2001 Using multivariate statistics. 4th ed Boston Allyn and Bacon 932 p. Maxwell SE 2000 Sample size and multiple regression analysis Psychol Methods 5 434 458 11194207 Casadevall A Pirofski LA 1999 Host-pathogen interactions: Redefining the basic concepts of virulence and pathogenicity Infect Immun 67 3703 3713 10417127 Casadevall A 2005 Host as the variable: Model hosts approach the immunological asymptote Infect Immun 73 3829 3832 15972467	16738705	PMC1323474	CC BY	2021-01-05 12:11:25	no	PLoS Pathog. 2005 Dec 30; 1(4):e40	utf-8	PLoS Pathog	2,005	10.1371/journal.ppat.0010040	oa_comm
==== Front PLoS PathogPLoS PathogppatplpaplospathPLoS Pathogens1553-73661553-7374Public Library of Science San Francisco, USA 1638930210.1371/journal.ppat.001004505-PLPA-RA-0135R1plpa-01-04-09Research ArticleEpidemiology - Public HealthEvolutionInfectious DiseasesMicrobiologyEubacteriaNone Bordetella pertussis, the Causative Agent of Whooping Cough, Evolved from a Distinct, Human-Associated Lineage of B. bronchiseptica Evolution of Human-Associated Bordetella Species Diavatopoulos Dimitri A 12Cummings Craig A 34Schouls Leo M 1Brinig Mary M 34Relman David A 345Mooi Frits R 12* 1 Laboratory for Vaccine-Preventable Diseases, National Institute of Public Health and the Environment, Bilthoven, Netherlands 2 Eijkman Winkler Institute, University Medical Center, Utrecht, Netherlands 3 Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, California, United States of America 4 VA Palo Alto Health Care System, Palo Alto, California, United States of America 5 Department of Medicine, Stanford University School of Medicine, Stanford, California, United States of America Hultgren Scott EditorWashington University in St. Louis, United States of America* To whom correspondence should be addressed. E-mail: [email protected] 2005 30 12 2005 1 4 e4523 8 2005 28 11 2005 Copyright: © 2005 Diavatopoulos 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. Bordetella pertussis, B. bronchiseptica, B. parapertussishu, and B. parapertussisov are closely related respiratory pathogens that infect mammalian species. B. pertussis and B. parapertussishu are exclusively human pathogens and cause whooping cough, or pertussis, a disease that has resurged despite vaccination. Although it most often infects animals, infrequently B. bronchiseptica is isolated from humans, and these infections are thought to be zoonotic. B. pertussis and B. parapertussishu are assumed to have evolved from a B. bronchiseptica–like ancestor independently. To determine the phylogenetic relationships among these species, housekeeping and virulence genes were sequenced, comparative genomic hybridizations were performed using DNA microarrays, and the distribution of insertion sequence elements was determined, using a collection of 132 strains. This multifaceted approach distinguished four complexes, representing B. pertussis, B. parapertussishu, and two distinct B. bronchiseptica subpopulations, designated complexes I and IV. Of the two B. bronchiseptica complexes, complex IV was more closely related to B. pertussis. Of interest, while only 32% of the complex I strains were isolated from humans, 80% of the complex IV strains were human isolates. Comparative genomic hybridization analysis identified the absence of the pertussis toxin locus and dermonecrotic toxin gene, as well as a polymorphic lipopolysaccharide biosynthesis locus, as associated with adaptation of complex IV strains to the human host. Lipopolysaccharide structural diversity among these strains was confirmed by gel electrophoresis. Thus, complex IV strains may comprise a human-associated lineage of B. bronchiseptica from which B. pertussis evolved. These findings will facilitate the study of pathogen host-adaptation. Our results shed light on the origins of the disease pertussis and suggest that the association of B. pertussis with humans may be more ancient than previously assumed. Synopsis Bordetella pertussis causes whooping cough, which kills 300,000 persons annually, and is reemerging despite vaccination. This human-restricted species is closely related to the respiratory pathogens B. parapertussishu, which is also human restricted, and B. bronchiseptica, which infects a broad range of mammals. Based on its limited genetic diversity and lack of historical descriptions, it has been suggested that the association between B. pertussis and humans is recent. In this study, the authors examined the genetic diversity and evolutionary relationships of these three Bordetella species. Their results suggest that B. parapertussis evolved from an animal-associated lineage of B. bronchiseptica, while B. pertussis evolved from a distinct B. bronchiseptica lineage that may already have had a preference for hominids up to 2.5 million years ago. Extant members of this newly identified B. bronchiseptica lineage were found to circulate in human populations. Comparisons of gene content revealed genomic features that are shared by and specific to B. pertussis and the B. bronchiseptica human-associated lineage and that may be important for association with the human host. These two lineages also have differences in key virulence genes that may reflect immune competition in the human host. By elucidating the evolutionary origins of human-adapted Bordetella, this study sets the stage for identification of key molecular events in host adaptation. Citation:Diavatopoulos DA, Cummings CA, Schouls LM, Brinig MM, Relman DA, et al. (2005) Bordetella pertussis, the causative agent of whooping cough, evolved from a distinct, human-associated lineage of B. bronchiseptica. PLoS Pathog 1(4): e45. ==== Body Introduction The genus Bordetella is composed of several species, of which three are exclusively respiratory pathogens of mammalian hosts: B. bronchiseptica, B. pertussis, and B. parapertussis (henceforth referred to as the mammalian bordetellae). B. bronchiseptica causes chronic and often asymptomatic respiratory tract infections in a wide variety of mammals. It is only sporadically isolated from humans [1,2], particularly from immunocompromised individuals, and human infections have been considered to be zoonotic [3]. B. parapertussis consists of two distinct lineages: one found in humans and the other found in sheep (B. parapertussishu and B. parapertussisov respectively) [4]. B. pertussis and B. parapertussishu have been isolated only from humans and cause acute, transient infections and disease, designated whooping cough or pertussis. Pertussis is especially severe in young, unvaccinated children and has reemerged in recent years in vaccinated populations [5–7]. Previous research indicated that B. pertussis and B. parapertussishu independently evolved from a B. bronchiseptica–like ancestor [8.9], and comparison of the genomes of the three isolates chosen for sequencing suggested that the time to the last common ancestor (LCA) for B. pertussis and B. parapertussishu and for B. bronchiseptica was 0.7 to 3.5 and 0.27 to 1.4 million years, respectively [10]. Despite their different host tropisms, the mammalian bordetellae are very closely related [8,9]. Analysis of their genome sequences revealed that the adaptation of B. pertussis and B. parapertussishu to the human host was accompanied by extensive genome decay [10]. Their differences in host tropism in contrast to their close genetic relationships make the mammalian bordetellae attractive candidates for the study of host-adaptation. Such studies are facilitated by the availability of genome sequences of B. bronchiseptica, B. pertussis, and B. parapertussishu [10]. So far, only a single representative of each species has been sequenced, and it is important to determine their relationships to the Bordetella population as a whole. To that end and to identify genetic events that may be associated with host adaptation, we used a combination of multilocus sequence typing (MLST) [11], comparative genomic hybridization (CGH) with whole-genome microarrays [12], and the distribution of several insertion sequence elements (ISEs) to characterize 132 mammalian Bordetella strains with diverse host associations. This work identified two B. bronchiseptica lineages, the first of which is composed of mainly strains of animal origin and includes the B. bronchiseptica strain from which the genome sequence has been determined. The second lineage, comprising strains mainly of human origin, is more closely related to B. pertussis than the first lineage. Comparison of the two B. bronchiseptica lineages to B. pertussis revealed genetic differences that may be associated with adaptation to the human host. Results Population Structure of the Mammalian Bordetellae, Based on Multilocus Sequence Typing To determine the relationships between the mammalian bordetellae, we determined the partial sequences of seven housekeeping genes from 132 strains (Table S1 and http://pubmlst.org/bordetella). We observed 32 sequence types (STs) among the 132 Bordetella isolates. Allele segments were divided into five equally sized subloci, and a minimum spanning tree (MST) algorithm was used to cluster the subloci [13]. Complexes were defined as groups of strains differing at fewer than five of 35 subloci with a minimum of two STs per complex. Using this criterion, strains could be assigned to one of four complexes, designated complexes I through IV (Figure 1). Figure 1 Minimum Spanning Tree of B. bronchiseptica, B. pertussis, and B. parapertussis The tree was based on the sequence of seven housekeeping genes. Individual genes were split into five subloci, and a categorical clustering was performed. In the minimum spanning tree, sequence types sharing the highest number of single locus variants were connected first. Each circle represents a sequence type (ST) the size of which is related to the number of isolates within that particular ST. Colors within circles indicate host distribution. The numbers between connected STs represent the number of different subloci between those STs. The clonal complexes (I, II, III, and IV) are indicated by colored strips between connected STs. ST16 (B. bronchiseptica complex I) harbors the B. parapertussisov strains. STs containing strains of which the genome has been sequenced (B. pertussis Tohama, B. parapertussis 12822 or B. bronchiseptica RB50) are indicated by a thickset, dashed line. The distribution of the insertion sequence elements IS481, IS1001, IS1002, and IS1663 is shown in boxes (see also Table S1); numbers between parentheses indicate the percentage of strains that contained the ISE as determined by PCR amplification. The divergence times between B. bronchiseptica complexes I and IV and B. pertussis complex II are shown. Complexes II and III contained the B. pertussis and B. parapertussishu isolates, respectively. Both of these complexes showed very limited genetic diversity (H = 0.65 and 0.35, respectively), as described previously [8,9]. B. bronchiseptica was divided into two distinct populations, designated complexes I and IV, respectively. The genetic diversity of these two complexes (H = 2.16 and 2.45, respectively) was much higher than that of complexes II and III. Complex I contained the majority of the B. bronchiseptica strains (76 of 91 strains), including the sequenced RB50 strain. In addition, it contained the B. parapertussisov isolates in the study population (ST16). B. bronchiseptica complex IV was more closely related to B. pertussis than was B. bronchiseptica complex I. Furthermore, the host species associations of the two complexes were quite distinct. Of the B. bronchiseptica complex IV isolates, 80% were isolated from humans, while this was the case for only 32% of the complex I isolates. It should be noted that human B. bronchiseptica isolates were overrepresented in our strain collection, in comparison with their occurrence in natural populations. However, the human complex IV isolates originated from different continents, comprising North America, South America, and Europe. Previously, phylogenetic analysis based on CGH suggested the existence of a distinct B. bronchiseptica lineage that was closely related to B. pertussis [12]. The relationships among the mammalian bordetellae inferred from housekeeping gene sequences were confirmed by an analysis based on the pertactin gene (prn), which codes for a surface-associated virulence factor involved in adherence [14,15]. A UPGMA tree was constructed from the aligned prn sequence data, and the topology of this tree was very similar to the MLST tree (Figure 2). B. bronchiseptica strains are grouped into two lineages, corresponding to complex I and IV in the MLST tree. Also, B. bronchiseptica complex IV and B. pertussis strains clustered together in one branch, which was supported by bootstrapping. B. parapertussishu comprised a separate branch within a larger cluster that also contained the B. bronchiseptica complex I strains. The B. parapertussisov strains were indistinguishable from B. bronchiseptica complex I strains, as was observed in the MLST tree. Figure 2 UPGMA Tree Based on the Analysis of the Pertactin Gene of Bordetella Isolates Used in the MLST Analysis The DNA segment coding for the extracellular domain of pertactin (P.69) was used for analysis, with the exclusion of the repeat regions 1 and 2. Bootstrap values are shown for the nodes separating the complexes and are based on 500 bootstrap replicates. The scale indicates the genetic distance along the branches. Colors of the branches indicate the four complexes defined by MLST. The number of strains of each branch is shown in boxes, as well as the host distribution. Distribution of Insertion Sequence Elements The distribution of ISEs has been used to reveal evolutionary relationships among the Bordetella population [9]. Toward this end, we screened our strain collection for the presence of IS481, IS1001, IS1002, and IS1663 using PCR (Table S1). The distribution of the ISEs was mapped onto the MST (see Figure 1). IS481 was detected in all B. pertussis strains but not in any other species, with the exception of two B. bronchiseptica isolates, both from a horse (B1975, B0230, ST6), consistent with previous observations [9,16]. IS1001 was detected in all B. parapertussishu and B. parapertussisov strains. Additionally, IS1001 was detected in most (21 of 25) B. bronchiseptica strains belonging to ST7 in complex I, but not in other STs, including STs in complex II and IV. IS1002 was detected only in B. pertussis and in B. parapertussishu strains, confirming previous observations [4,9]. IS1663 [10] was detected in all B. pertussis isolates but also in ten of 13 B. bronchiseptica complex IV strains. The three complex IV strains in which IS1663 was not detected belonged to STs 18 and 21. Divergence Times of Complexes Under the assumption that the mutation rate in prokaryotes is relatively constant, the time since descent from the LCA can be estimated using pairwise mean allele distances (K S) [17,18]. The clock rates described by Whittam [19] and by Guttman and Dykhuizen [20] were used to estimate a range of divergence times between complexes. Calculations indicated that B. pertussis and B. bronchiseptica complex IV separated approximately 0.3 to 2.5 million years ago (Mya), which suggests a more recent divergence time than B. pertussis and B. bronchiseptica complex I, estimated at 1.1 to 5.6 Mya. B. parapertussishu and B. bronchiseptica complex I diverged between 0.7 and 3.5 Mya according to our calculations. The divergence times of combinations of complexes is shown in Figure 1. Gene Content of Strains from B. bronchiseptica Complexes I and IV The high percentage of complex IV strains of human origin compared to the percentage of those in complex I suggested a preference for the human host in complex IV strains. To identify genetic events that may have played a role in host adaptation or host restriction, we used CGH with Bordetella DNA microarrays [17,18]. Genomic DNA from 26 B. bronchiseptica complex I and 13 complex IV strains was hybridized to microarrays (CGH data files have been deposited in ArrayExpress, accession E-TABM-32). Significance Analysis of Microarrays (SAM) was used to identify probes with statistically significant log intensity ratio differences between the two complexes (Table S2). This approach detected sequences that have been deleted more often in one of the two complexes, as well as DNA sequences diverging from the reference sequences in one of the two complexes. Thirty-one probes, representing 29 genes and an ISE, hybridized more strongly or more frequently to the genomes of complex IV strains than to those of complex I. Two of these probes represented the B. pertussis homologs of the virulence genes prn and tcfA, encoding pertactin and tracheal colonization factor, respectively [15,21] (Figure 3). Sequence analysis confirmed that the B. bronchiseptica complex IV prn sequences were more similar to those of B. pertussis than to those of the B. bronchiseptica complex I (see Figure 2). Figure 3 Gene Content of the Differentially Hybridizing Virulence Loci between B. bronchiseptica Complex I and IV, as Determined by CGH Each column represents one strain. Strain numbers and STs are indicated above the columns. Each row represents one ORF (in B. bronchiseptica RB50 gene order), ORF designations are shown to the right of the rows. In the case of tcfA and prn, the origins of the probes are indicated between parentheses. The BP probe of tcfA was 100% similar to B. pertussis Tohama and 85.1% similar to B. bronchiseptica RB50. The BP prn probe was 100% similar to B. pertussis Tohama and 86% similar to B. parapertussis 12822 and B. bronchiseptica RB50. The BB/BPP prn probes were both 100% similar to B. parapertussis 12822 (BPP) and B. bronchiseptica RB50 (BB) and 86% similar to B. pertussis Tohama. The yellow-black-blue color scale indicates the hybridization value relative to the reference; references are B. bronchiseptica RB50, B. parapertussis 12822, and B. pertussis Tohama. For B. bronchiseptica RB50 and B. pertussis Tohama, the data in the figure are based on the genomic sequences. Yellow indicates decreased hybridization, black indicates hybridization values comparative to the references, and blue indicates gene duplications. Intermediate values indicate partial deletions or sequence divergence. Missing data are represented in gray. Sixteen of these 31 probes had been identified previously as B. pertussis–specific [12]. In most cases, these probes hybridized to nine or more complex IV strains but not to any complex I or III strain. The genes represented by these probes encode diverse functions involved in metabolism, transport, regulation, and transposition (Table S2). With the possible exception of BP0703, which encodes a TonB-dependent iron receptor, no obvious virulence genes were observed among them. Most of the 16 genes were located in small clusters along the B. pertussis Tohama chromosome. The presence of these genes in B. bronchiseptica complex IV and B. pertussis but not in B. bronchiseptica complex I or B. parapertussishu suggests that they were acquired by the common ancestor of B. pertussis and B. bronchiseptica complex IV. We also identified 248 probes, representing 237 genes, that exhibited significantly stronger hybridization to complex I genomes than to complex IV genomes (Table S2). Sixty-eight (27%) corresponded to genes associated with mobile elements such as prophages, while many of the other probes represented genes involved in metabolic, transport, and regulatory functions. Surprisingly, several virulence-associated genes were found to be missing or divergent in the complex IV strains, as compared to complex I strains. These included the B. bronchiseptica homologs of tcfA and prn, Bvg-regulated intermediate phase gene A (bipA), the alcaligin biosynthesis locus (alcA/E), the pertussis toxin synthesis and transport locus (ptx/ptl), the dermonecrotic toxin gene (dnt), and the lipopolysaccharide (LPS) biosynthetic locus (Figures 3 and 4). Figure 4 Expression of LPS by B. bronchiseptica Complex I and Complex IV Strains and Gene Content Variation at the LPS Biosynthesis Locus (A) Top: Electrophoretic LPS profiles obtained by tricine-SDS-PAGE and silver staining. Middle: Western blot of the same samples with mAb 36G3, which detects band A. Bottom: Western blot of the same samples with mAb BL8, which detects band B. (B) Gene content of the LPS biosynthesis locus as determined by CGH. See Figure 3 for details. For B. bronchiseptica RB50 and B. pertussis Tohama, the data in the figure are based on the genomic sequences. The genes wbmPQRSTU represent an alternative LPS O-antigen biosynthesis sublocus that is orthologous to the genes found in B. parapertussis 12822 [10] and B. bronchiseptica C7635E [26]. LPS genetic profiles as described in the text are indicated at the top of the columns. Color scale as in Figure 3. Missing data are represented in gray. Interestingly, ten of 13 complex IV strains harbored deletions in the ptx and ptl loci, which encode pertussis toxin (Ptx) and its secretion machinery, respectively (Figure 3) [22,23]. While conditions under which Ptx is expressed by either B. parapertussis or B. bronchiseptica have not been found, the structural genes are generally conserved among these species [24], suggesting selective pressure to retain the ability to produce functional Ptx under certain circumstances. The complex IV strains in which ptx/ptl was still present (ST3/17/29) were tested for expression of Ptx by immunoblotting, and these strains were found not to express Ptx under the growth conditions used (unpublished data). Another distinguishing characteristic of the complex IV strains was the deletion of the dermonecrotic toxin gene (dnt) in 8 of 13 strains. In contrast, this gene was detected in all B. pertussis, B. parapertussishu, B. parapertussisov, and B. bronchiseptica complex I strains. LPS Genetic and Structural Diversity in B. bronchiseptica The genetic structure of the LPS biosynthesis locus differed between complex I and IV strains. The LPS molecules of Gram-negative bacteria usually consist of three, covalently linked, major domains: the lipid A, the branched chain oligosaccharide core, and the hydrophilic O-antigen. A number of genetic loci have been implicated in the synthesis of these domains in Bordetella, such as the lpx locus (lipid A), the waa locus (inner core), the wlb locus (outer core), and the wbm locus (O-antigen) [25–29]. B. pertussis LPS usually consists of lipid A and an inner core, to which the outer core (a trisaccharide) is attached; this is also referred to as band A. Certain B. pertussis strains produce only band B, which is identical to band A except that it lacks the trisaccharide [30]. The O-antigen, which is added to the trisaccharide, is found only in B. bronchiseptica and B. parapertussis. This structure is missing in B. pertussis due to the deletion of the genes wbmA-U [26]. In Figure 4, the gene content of the LPS locus is shown for 13 complex I and 13 complex IV strains and for B. pertussis Tohama and B. bronchiseptica RB50. Four LPS gene content profiles, designated LPS 1–4, could be distinguished among the B. bronchiseptica strains. Strains with the LPS 1 profile had an LPS gene composition similar to RB50, characterized by the absence of wbmPQRSTU. The LPS 2 profile was characterized by the absence of the genes wbmORS and BB0124–BB0127 and the presence of an alternative O-antigen locus, comprising wbmPQRSTU, orthologous to the B. parapertussis 12822 genes [10]. Both of these genotypes appear competent for the production of a full length LPS. Strains with the LPS 3 profile lacked wbmD-U and BB0124-BB0127, suggesting that they may not produce O-antigen. The LPS 4 profile was similar to the LPS 3 profile but additionally lacked wbmABC and wlbD-L, suggesting that strains of this genotype may be deficient for the production of trisaccharide as well as O-antigen. The deletion in the O-antigen genes of LPS 4 strains was similar to that observed in the O-antigen genes of B. pertussis Tohama. All complex I strains displayed either an LPS 1 or an LPS 2 profile. Nine of 13 complex IV strains had either an LPS 1 or an LPS 2 profile, while four complex IV strains showed more extensive deletions, resulting in LPS 3 and LPS 4 profiles. To study the effect of these deletions on LPS production, proteinase K–treated cell lysates were analyzed by Tricine-SDS-PAGE, followed by silver staining or immunoblotting with monoclonal antibodies (mAbs) directed against either band A or band B (mAbs 36G3 and BL-8, respectively [31,32]) (Figure 4). Silver-staining showed that all complex I strains produced band A, except for B1985, which produced band B, and B2112, which produced a band migrating at a position between bands A and B. These strains showed no obvious deletions at their wlb or wbm loci, and the fact that they did not produce band A and O-antigen may be attributed to point mutations, e.g., in their wlb locus, or to regulatory effects. These results were confirmed by immunoblotting with mAb 36G3. The epitope in the trisaccharide that is recognized by this mAb was also present in the O-antigenic repeats, as was described previously [33]. In general, most strains that produced band A also produced an additional band just above band A. This extra band was also recognized by 36G3 and therefore is likely derived from band A. The nine complex IV strains with LPS 1 or 2 profiles all produced band A. The two strains with the LPS 4 profile, B2490 and B2506, produced a band smaller than band A, which failed to be recognized by mAb 36G3. Of the two strains with an LPS 3 profile, one strain (B0243) produced only band B and no O-antigen. Unexpectedly, the other strain, B2494, produced band A and O-antigen as detected by immunoblotting, indicating that this strain contains as yet uncharacterized O-antigen biosynthesis genes. Silver staining also suggested that the LPS 4 strains may produce O-antigen, although the O-antigen failed to be recognized by mAb 36G3. Discussion Although it has long been speculated that B. pertussis evolved from a B. bronchiseptica strain [8,9], a specific lineage has not been identified. Here we identify and characterize such a B. bronchiseptica lineage. Analysis of MLST data from the mammalian bordetellae identified four distinct complexes. Complex I and IV comprised B. bronchiseptica strains, while complex II and III comprised the human pathogens B. pertussis and B. parapertussishu, respectively. Our results suggest that B. pertussis and B. parapertussishu evolved from complexes I and IV, respectively, indicating that adaptation to humans occurred as two independent events, consistent with previous data [9,10]. The population structure of the mammalian bordetellae inferred from MLST data largely corresponded with a maximum parsimony phylogeny derived from a previous CGH study [12], with the exception of the relationship of B. parapertussisov and B. parapertussishu. In the current study, B. parapertussishu and B. parapertussisov are clearly derived from different STs in complex I. Further, in contrast to B. parapertussishu, B. parapertussisov is actually part of B. bronchiseptica complex I. The closer relationship between the sheep- and human-derived B. parapertussis lineages that was inferred from CGH data may be an artifact of long-branch attraction in the maximum parsimony tree [34]. Consistent with previous studies [8,9,35], B. pertussis and B. parapertussishu showed a relatively low degree of genetic diversity, suggesting that they evolved recently or encountered a recent evolutionary bottleneck. Of the three B. pertussis STs observed, two were found exclusively before 1960, whereas all modern strains belong to ST2. The temporal shift in B. pertussis STs is consistent with our previous studies on antigenic shifts that show major changes in the B. pertussis population after the introduction of mass vaccination against pertussis in the 1950s and 1960s [35,36]. Most human disease is by far caused by B. pertussis, and we therefore focused on the relationship of B. bronchiseptica complex IV with B. pertussis. B. bronchiseptica complex IV strains were found to be more closely related to B. pertussis than to the complex I B. bronchiseptica strains. A tree based on prn nucleotide sequences also suggested a closer relationship of complex IV strains to B. pertussis than to complex I strains. A number of other features of complex IV strains were consistent with their close relationship with B. pertussis. Most complex IV strains were isolated from humans (80%), while the majority of complex I strains were of animal origin (68%). Almost all B. bronchiseptica complex IV strains were isolated from patients with whooping cough symptoms. Further, complex IV strains and B. pertussis shared an IS element, IS1663, that was not found outside these two lineages. The sharing of an IS element may be explained by either vertical or horizontal transfer. The former suggests a common ancestry, while the latter would point to niche sharing of B. pertussis and B. bronchiseptica complex IV. It seems unlikely that the association of complex IV strains with humans is due to a sampling artifact, as the strains analyzed were from widely separated geographic regions, including North America, South America, and Europe. Thus, these strains were not epidemiologically related. Three STs (ST12, ST23, and ST27) found in complex I also contained a high percentage of human strains (55%, 43%, and 77%, respectively). All other, nonhuman isolates of these STs were collected from domesticated animals. Thus, both complex I and IV contain B. bronchiseptica strains that are well adapted to the human host. However, the particular relevance of the human-associated lineage in complex IV appears to be its evolutionary relationship with B. pertussis. The high frequency of human isolates observed in complex IV may be due to the close interaction of humans with the animal hosts in which these strains reside or to the fact that complex IV strains are better adapted to a human environment than B. bronchiseptica complex I strains. In either case, the B. bronchiseptica complex IV infections of humans would be zoonotic. Another intriguing possibility is that B. bronchiseptica complex IV strains are to a large extent adapted to the human host and primarily transmitted between humans. Microarray-based CGH revealed 29 genes and IS1663 to be more frequently present in, or more similar to, B. pertussis orthologs in B. bronchiseptica complex IV than to complex I strains. Of these genes, 16 were unique to B. pertussis and B. bronchiseptica complex IV, suggesting they were acquired after the common ancestor of complex IV and B. pertussis diverged from complex I. With the exception of prn and tcfA, which hybridized more strongly to the B. pertussis–derived probe, no known virulence genes were identified among these 30 genes. Conversely, 237 genes were absent or divergent in complex IV compared to complex I strains, suggesting that the B. bronchiseptica complex IV genome is decaying, as has been assumed for B. pertussis and B. parapertussishu. Genome decay has been associated with host restriction or niche change in a number of pathogens such as Yersinia pestis [37,38] and Burkholderia mallei [39] and has been suggested to be a driving force of host restriction for B. pertussis and B. parapertussis as well [10]. Likewise, the apparent preference of complex IV strains for humans may also be associated with genome decay. Because putative complex IV–specific sequences were not represented on the microarray used here, we were unable to address the possibility that complex IV strains have acquired, through lateral transfer, genetic loci that may have promoted host preference. However, gene acquisition appears to have been a rare event in the evolution of B. pertussis and B. parapertussis from B. bronchiseptica complex I [10]. Differences between complex IV strains and B. pertussis were observed with respect to three major virulence factors, Ptx, Dnt, and LPS. All B. bronchiseptica complex I strains examined contained intact Ptx genes. Although conditions under which the Ptx genes are expressed in B. bronchiseptica have not been identified, their conservation suggests that they may confer a selective advantage in the ecology of complex I strains. In contrast to complex I strains, the genes encoding Ptx and its secretion apparatus were deleted from most complex IV strains (10 of 13). In the strains that did retain the Ptx locus, no in vitro expression of Ptx was observed, even though these strains were closely related to B. pertussis. Another characteristic that sets complex IV strains apart from all other mammalian bordetellae is that in eight of 13 strains, dnt was deleted. Dnt is an intracellular toxin that activates the small GTPase Rho through deamidation or polyamination [40]. It has been shown that Dnt is important for turbinate atrophy and the colonization of the upper respiratory tract by B. bronchiseptica in pigs [41]. Based on CGH, four LPS genetic profiles were distinguished. The LPS genetic locus was generally more polymorphic in complex IV strains than in complex I strains, and deletions were observed in the O-antigen and trisaccharide biosynthesis genes in some complex IV strains. In complex IV strains with the LPS 4 profile, the extent of deletion in the O-antigen genes was very similar to that seen in B. pertussis Tohama. B. pertussis does not produce repetitive O-antigen as it lacks the wbm genes but makes a lipo-oligosaccharide that consists of lipid A to which a single trisaccharide is attached [26]. Like B. pertussis, four complex IV strains lacked the O-antigen genes known to be present in the sequenced genomes of B. bronchiseptica and B. parapertussishu. Despite the absence of these genes, in at least one of these strains O-antigen was detected by immunoblotting, suggesting that this strain carries LPS genes distinct from those in RB50 or 12822. The two LPS 4 strains, both isolated from humans, also lacked the genes required for biosynthesis of the trisaccharide and failed to produce trisaccharide detectable by immunoblotting. The absence of the trisaccharide is intriguing in view of the fact that it was found to be otherwise conserved in all Bordetella strains analyzed. It seems likely that in addition to gene loss and acquisition, differences in gene regulation have significantly contributed to host adaptation [42]. The differences observed between complex IV strains and B. pertussis, particularly with respect to Ptx, Dnt, and LPS, may be due to differences in niches occupied. Another possibility is that these differences have arisen in response to immune competition between B. bronchiseptica complex IV strains and B. pertussis. Gupta and co-workers [43] provided evidence that immunodominant surface antigens are organized into nonoverlapping combinations as a result of selection by the host immune system. This process could also have driven the inactivation and conservation of virulence factors in mammalian bordetellae infecting the same host. In a similar vein, Bjørnstad and Harvill [44] hypothesized that, since B. pertussis and B. parapertussishu both infect humans, they may have evolved to evade cross-immunity by the other pathogen. The authors propose that immune competition provides an explanation for differences observed between B. pertussis and B. parapertussishu. For example, B. pertussis but not B. parapertussishu expresses Ptx, although both contain the required genes. Conversely, B. parapertussishu expresses O-antigen, while the corresponding genes have been deleted from B. pertussis. Similarly, the deletion of the genes for Ptx, Dnt, and genes involved in trisaccharide syntheses by complex IV strains may have been driven by immune competition with B. pertussis and possibly also with B. parapertussishu . The origin of the disease whooping cough is still a mystery. Although the disease has very typical symptoms in children and was one of the major causes of child mortality before the introduction of vaccination, the first written reference to the disease in Europe is found in 1540 [45]. The first description of an epidemic, which occurred in Paris, was given by Baillon in 1578 [46]. Particularly interesting are the observations made by Nils Rosen von Rosenstein in 1766, who wrote [47], “The hooping cough never appeared in Europe originally, but was transported thither from other parts of the world by means of merchandise, seamen and animals. Its first appearance in Sweden cannot be determined with any certainty; but in France it began in the year 1414.” In contrast, 16th- and 17th-century descriptions of the disease and epidemics in Europe are documented frequently in the literature [46]. The absence of references to pertussis-like symptoms in the ancient literature has been taken as evidence that the association of B. pertussis with humans is of recent origin. We propose that the association of B. pertussis with humans is, in fact, ancient but that the introduction of B. pertussis into Europe may be more recent. Complex IV strains showed a degree of diversity that was comparable to complex I strains (2.16 and 2.45, respectively), and thus, assuming that complex IV strains are primarily adapted to the human host, this association must be ancient. Parkhill et al. [10] previously estimated the time to the LCA of a B. bronchiseptica complex I strain (RB50) and B. pertussis to be 0.7 to 3.5 Mya, based on the mean number of synonymous substitutions per synonymous site of orthologous gene pairs. Our data indicate that current B. pertussis strains expanded clonally from the B. pertussis–B. bronchiseptica complex IV LCA 0.32 to 2.53 Mya, further supporting an ancient association of B. pertussis with humans. However, we cannot rule out the possibility that more recent human-associated ancestors of B. pertussis are extinct or undiscovered. Such recent ancestors would indicate a more recent origin of B. pertussis. Although it is tempting to speculate that the LCA of B. pertussis and B. bronchiseptica complex IV was associated with humans, the possibility remains that this association emerged after the split with B. pertussis. A possible evolutionary scenario (Figure 5) represents the adaptation of an ancestral B. bronchiseptica complex I strain to humans or their hominid ancestors. From this lineage, the LCA of B. bronchiseptica complex IV and B. pertussis evolved, subsequently giving rise to B. bronchiseptica complex IV and B. pertussis. Figure 5 Model of the Evolution of the Mammalian Bordetellae The bar on the left indicates increasing degrees of adaptation to the human host. Arrows indicate descent; double arrows between complexes indicate possible within-host immune competition. In boxes, genetic events are shown that may have played a role in speciation and niche adaptation. Numbers between parentheses refer to references. See text for details. Recent emergence of a pathogenic clone from a more ancient human-associated progenitor species has been proposed as the mechanism for the origin of Mycobacterium tuberculosis [48]. Although previous genetic analysis had suggested that M. tuberculosis emerged as little as 20,000 years ago, phylogenetic analysis of M. tuberculosis and a closely related but more diverse group of smooth tubercle bacilli indicated that this more broadly defined species has been associated with hominids for up to 3 million years. Yersinia pestis, the causative agent of plague, is a clone that evolved from Y. pseudotuberculosis 1,500 to 20,000 years ago, shortly before the first known pandemics of human plague, and its recent origin is further suggested by the complete lack of polymorphism in housekeeping genes [18]. Similarly, B. pertussis also shows limited diversity. However, in contrast to Y. pestis, which reveals absolutely no polymorphisms in housekeeping genes, we observed three STs in B. pertussis. This may suggest an older origin of B. pertussis compared to Y. pestis, although other factors, such as population size and bottlenecks, could also explain these differences. The most plausible explanation from our data is that the association of B. pertussis with humans originated in the LCA of B. pertussis and B. bronchiseptica complex IV. Based on that assumption, the apparent emergence of pertussis in Europe within the last 500 years may be attributable to import via travel or migration or to the recent acquisition by B. pertussis of the ability to cause more severe, whooping cough–like symptoms. Although most of the B. bronchiseptica complex IV strains in our collection were isolated from patients suspected to have pertussis, we know little of the severity of the symptoms caused by these strains. It is conceivable that B. bronchiseptica preceded B. pertussis in Europe and that its disease was not documented because of its relatively mild and nonspecific course. The work presented here places the three sequenced mammalian Bordetella strains within a phylogenetic context, thereby facilitating rational selection of strains for further genomic sequencing. In particular, sequencing of one or more members of complex IV may shed more light on processes involved in host adaptation and immune competition. Further, the identification of a B. bronchiseptica lineage which circulates in human populations may be important for public health. In recent years, whole cell vaccines have been replaced by acellular vaccines comprised of one to five antigens derived from B. pertussis [49]. The acellular vaccines induce a less cross-reactive immune response compared to whole cell vaccines [50] and may therefore result in an increase in B. parapertussis and B. bronchiseptica infections in vaccinated human populations. Materials and Methods Bacterial strains. A total of 132 Bordetella isolates were used in this study: 91 B. bronchiseptica, 9 B. parapertussishu, 3 B. parapertussisov, and 29 B. pertussis isolates (see Table S1). The three strains from which the genome sequence has been determined, B. bronchiseptica RB50, B. pertussis Tohama and B. parapertussis 12822 [10], were included. The collection included clinical isolates from humans and a broad range of animal species. Strains were grown on Bordet Gengou (BD, Franklin Lakes, New Jersey, United States) agar supplemented with 15% sheep blood at 37 °C for 2 to 5 days. Chromosomal DNA was isolated using the Wizard Genomic DNA Purification Kit (Promega, Madison, Wisconsin, United States), according to the manufacturers' protocol for Gram-negative bacteria. DNA sequencing. The nucleotide sequences were determined for internal regions of seven housekeeping genes for all strains (http://pubmlst.org/bordetella [51]). The nucleotide sequence of the prn region encoding the extracellular domain of the surface-associated autotransporter pertactin, P.69, was determined for 116 strains, with the exclusion of the repeat regions 1 and 2 [35]. These regions are comprised of amino acids repeats and are highly polymorphic due to insertion or deletion of the repeat unit. Primer information is listed in Table S3. Detection of ISEs. The distribution of IS481, IS1001, IS1002, and IS1663 was determined for all strains using PCR amplification (see Table S1). For PCR amplification of IS481, IS1001, and IS1002, primers were used that have been described previously [4,52]. Primer characteristics are listed in Table S3. LPS SDS-PAGE and Western blotting. BG-agar grown bacteria were harvested, boiled in 1× sample buffer (7.5% glycerol, 0.125 M Tris-HCl [pH 6.8], 1.5% SDS), and treated with proteinase K [33]. Tricine-SDS-PAGE was then performed in 4% stacking and 16% separating gels, as previously described by Lesse et al. [53]. Silver staining was performed as described by Tsai and Frasch [54]. LPS was transferred to PVDF membranes (Amersham Biosciences, Buckinghamshire, United Kingdom) and blocked with 0.5% (w/v) Protifar nonfat dried milk, 0.5% bovine serum albumin (w/v), and 0.1% Tween 20 in PBS. Immunoblotting was performed with monoclonal antibodies 36G3 and BL-8, directed against band A and band B LPS, respectively [31,32]. Sequence data analysis. Analysis of nucleotide sequence data was performed using Bionumerics software package version 4.0 beta 4 (Applied Maths, Sint-Martens-Latem, Belgium). The Bordetella MLST database can be accessed at http://pubmlst.org/bordetella [51]. For each locus in the MLST analysis, the allele sequences for all strains were trimmed to a uniform length, and an allele number was assigned to each unique allele sequence. The combination of the allele numbers at the seven loci defines the ST or allelic profile of each strain. Construction of trees based on allelic profiles may not accurately reflect the true genetic distance because both single and multiple nucleotide polymorphisms are given equal weight. Consequently, the degree of sequence difference between two alleles is not quantitatively reflected in the MLST profile. Conversely, tree construction based on concatenated allele sequences does not take into account the introduction of clustered multiple base substitutions due to a single recombinational event. As a result, trees based on MLST sequences often contain long branches, incorrectly suggesting a large genetic distance. Therefore, we used a method designated as split-MLST, in which each locus is split into a user-defined number of equally sized subloci (D. A. Diavatopoulos, P. Vauterin, L. Vauterin, F. R. Mooi, and L. M. Schouls, unpublished data). Using this method, the sensitivity of categorical clustering could be increased, without the perturbing effect of recombination. The topology of the tree appeared to vary if the number of subloci per MLST locus was lower than five. However, above the value four, increasing the number of subloci had no significant effect on the topology of the tree, and we therefore selected the lowest possible split value, five, resulting in a total of 35 subloci. The genetic diversity for each complex was calculated using the Shannon-Weiner index of diversity (H) using the following formula: where Pi is the frequency of the ith type [55]. For estimation of divergence times between complexes, we calculated the pairwise mean distance (Ks) between alleles using DNASP 4.00 [56]. The divergence time was calculated using the following formula: where Ks is the number of synonymous substitutions per synonymous site and r is the molecular clock rate of Escherichia coli as determined by Whittam [19] or by Guttman and Dykhuizen [20]. We used these two rates to calculate a range of divergence times. The divergence time was first calculated for each combination of STs between complexes, and from these the averaged age between complexes was calculated. Comparative genomic hybridization. The preparation of PCR product-based microarrays and the comparative genomic hybridization was performed essentially as described by Cummings et al. (12). This study employed a new array design that contained all of the probes from the first array plus 1,417 additional probes that brought the theoretical ORF coverage of these arrays up to 97.4% for B. pertussis Tohama, 98.5% for B. bronchiseptica RB50, and 97.9% for B. parapertussis 12822. Like the previously used array probes, these additional probes were PCR products with a size of less than 300 base pairs and amplified from the sequenced reference genomes with ORF-specific oligonucleotides (Illumina, San Diego, California, United States) designed with Microarray Architect (C. A. Cummings and D. A. Relman, unpublished data). A total of 26 complex I and 13 complex IV strains were hybridized to the arrays. The genomic DNA of B. bronchiseptica was labeled with Cy5 and hybridized to the array in conjunction with a Cy3-labeled genomic DNA reference comprising the three sequenced mammalian Bordetella genomes (B. pertussis Tohama, B. parapertussis 12822, and B. bronchiseptica RB50). For the list of strains analyzed by CGH, see Table S1. Labeled probes were purified using the Cyscribe GFX Purification Kit (Amersham Biosciences, Freiburg, Germany) following the manufacturer's protocol for probes produced by the CyScribe First-Strand cDNA Labelling Kit. After purification, the test and reference-labeled DNA samples were concentrated to less than 8.5 μl using a Savant SpeedVac SVC-100H. The test and reference samples were combined and 150 μg of yeast tRNA (Invitrogen Life Technologies, San Diego, California, United States) was added to block nonspecific binding. The probe volume was adjusted to 24 μl with water and then 5.1 μl of 20× SSC (1× SSC = 0.15 M NaCl plus 0.015 M sodium citrate) and 0.9 μl of 10% sodium dodecyl sulfate (SDS) were added. Thirty microliters of the probe was added to the array and covered with a 25 × 40 mm No. 1 glass coverslip. Hybridization was performed in GeneMachines Hybchambers (Genomic Solutions, Ann Arbor, Michigan, United States) with 2× 30 μl of 3× SSC to maintain humidity and incubated at 65 °C overnight. Arrays were washed in 0.5× SSC, 0.03% SDS for 30 s, 0.1× SSC, 0.01% SDS for 30 s, 0.05× SSC, 0.005% SDS for 1 min, and 0.025× SSC for 1 min. The first wash was performed at 65 °C, and the remaining washes were performed at room temperature. Slides were dried using a Quick-Dry Filtered Air Gun (Matrix Technologies Corporation, Hudson, New Hampshire, United States). Images were acquired on a PerkinElmer ScanArray 4000XL scanner using Scanarray Express software (PerkinElmer Life and Analytical Sciences, Inc., Boston, Massachusetts, United States). Images were analyzed with GenePix Pro software (Axon Instruments, Union City, California, United States). Processed two-color array image data were submitted to an in-house microarray database. Data were extracted using filters to eliminate automatically and manually flagged spots and spots with very low background subtracted signal intensity (<150) in the reference channel. B. bronchiseptica complex I and complex IV enriched sequences were identified using the Significance Analysis for Microarrays software (SAM) [57]. The probes were analyzed using 26 complex I and 13 complex IV strains that were hybridized to the arrays. SAM analysis was run using the two-class option with KNN missing value imputation. In addition to a statistically significant difference, a 2-fold difference in mean signal intensity ratio for each probe was also required. Supporting Information Dataset S1 CGH Data of the Mammalian Bordetellae (5.0 MB XLS) Click here for additional data file. Dataset S2 CGH Data of the Differentially Hybridizing Probes between Complexes I and IV as Identified by SAM (207 KB XLS) Click here for additional data file. Table S1 Characteristics of the Strains Used in the MLST Analysis (16 KB PDF) Click here for additional data file. Table S2 Probes That Hybridized Differentially to B. bronchiseptica Complex I and IV Genomes as Determined by SAM Analysis (19 KB PDF) Click here for additional data file. Table S3 Primer Characteristics for the Genes Used in Multilocus Sequence Typing, Pertactin Sequencing, and the Detection of the Insertion Sequence Elements (41 KB PDF) Click here for additional data file. Accession Numbers The nucleotide sequences of pertactin have been deposited in GenBank (http://www.ncbi.nlm.nih.gov/Genbank) under accession numbers DQ141700 through DQ141711 and DQ141713 through DQ141816. We are grateful to Dr. Geoffrey Foster (SAC Veterinary Science Division, Inverness) and to Dr. Gary Sanders (Centers for Disease Control and Prevention, Atlanta, Georgia, United States) for providing B. bronchiseptica strains. We thank Dr. Eric Harvill (Penn State University, Pennsylvania, United States) for sharing unpublished data and discussions and Ing. Marjolein van Gent, Ing. Betsy Kuipers, and Ing. Hendrik-Jan Hamstra for assistance and introduction to LPS work. We also thank Dr. Martin Maiden and Dr. Keith Jolley for assistance with setting up the Bordetella MLST database. This work was supported by a travel grant from Netherlands Organization for Scientific Research (NWO). CAC was supported by an American Lung Association Research Training Fellowship. DAR received grant support from the National Institutes of Health (grants AI54970 and AI057188). Competing interests. The authors have declared that no competing interests exist. Author contributions. DAD and FRM conceived and designed the experiments. DAD performed the experiments. DAD, CAC, LMS, and MMB analyzed the data. CAC and LMS contributed reagents/materials/analysis tools. DAD, CAC, LMS, DAR, and FRM wrote the paper. Abbreviations CGHcomparative genomic hybridization ISEinsertion sequence element LCAlast common ancestor LPSlipopolysaccharide mAbmonoclonal antibody Myamillion years ago MLSTmultilocus sequence typing MSTminimum spanning tree STsequence type ==== Refs References Gueirard P Weber C Le Coustumier A Guiso N 1995 Human Bordetella bronchiseptica infection related to contact with infected animals: Persistence of bacteria in host J Clin Microbiol 33 2002 2006 7559937 Bauwens JE Spach DH Schacker TW Mustafa MM Bowden RA 1992 Bordetella bronchiseptica pneumonia and bacteremia following bone marrow transplantation J Clin Microbiol 30 2474 2475 1401019 Woolfrey BF Moody JA 1991 Human infections associated with Bordetella bronchiseptica Clin Microbiol Rev 4 243 255 1889042 van der Zee A Groenendijk H Peeters M Mooi FR 1996 The differentiation of Bordetella parapertussis and Bordetella bronchiseptica from humans and animals as determined by DNA polymorphism mediated by two different insertion sequence elements suggests their phylogenetic relationship Int J Syst Bacteriol 46 640 647 8782670 Mooi FR van Loo IH King AJ 2001 Adaptation of Bordetella pertussis to vaccination: A cause for its reemergence? Emerg Infect Dis 7 (3 Suppl) 526 528 11485646 Crowcroft NS Britto J 2002 Whooping cough—A continuing problem BMJ 324 1537 1538 12089076 Orenstein WA 1999 Pertussis in adults: Epidemiology, signs, symptoms, and implications for vaccination Clin Infect Dis 28 (Suppl 2) S147 S150 10447034 Musser JM Hewlett EL Peppler MS Selander RK 1986 Genetic diversity and relationships in populations of Bordetella spp J Bacteriol 166 230 237 3957867 van der Zee A Mooi F van Embden J Musser J 1997 Molecular evolution and host adaptation of Bordetella spp.: Phylogenetic analysis using multilocus enzyme electrophoresis and typing with three insertion sequences J Bacteriol 179 6609 6617 9352907 Parkhill J Sebaihia M Preston A Murphy LD Thomson N Harris DE 2003 Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica Nat Genet 35 32 40 12910271 Maiden MC Bygraves JA Feil E Morelli G Russell JE Urwin R 1998 Multilocus sequence typing: A portable approach to the identification of clones within populations of pathogenic microorganisms Proc Natl Acad Sci U S A 95 3140 3145 9501229 Cummings CA Brinig MM Lepp PW Van De PS Relman DA 2004 Bordetella species are distinguished by patterns of substantial gene loss and host adaptation J Bacteriol 186 1484 1492 14973121 Schouls LM van der Heide HGJ Vauterin L Vauterin P Mooi FR 2004 Multiple-locus variable-number tandem repeat analysis of Dutch Bordetella pertussis strains reveals rapid genetic changes with clonal expansion during the late 1990s J Bacteriol 186 5496 5505 15292152 Leininger E Roberts M Kenimer JG Charles IG Fairweather N Novotny P 1991 Pertactin, an Arg-Gly-Asp-containing Bordetella pertussis surface protein that promotes adherence of mammalian cells Proc Natl Acad Sci U S A 88 345 349 1988935 Charles IG Dougan G Pickard D Chatfield S Smith M Novotny P 1989 Molecular cloning and characterization of protective outer membrane protein P.69 from Bordetella pertussis Proc Natl Acad Sci U S A 86 3554 3558 2542937 McLafferty MA Harcus DR Hewlett EL 1988 Nucleotide sequence and characterization of a repetitive DNA element from the genome of Bordetella pertussis with characteristics of an insertion sequence J Gen Microbiol 134 (Pt 8) 2297 2306 2908119 Jukes TH Cantor CR 1969 Mammalian protein metabolism New York Academic 132 p. Achtman M Zurth K Morelli G Torrea G Guiyoule A Carniel E 1999 Yersinia pestis, the cause of plague, is a recently emerged clone of Yersinia pseudotuberculosis Proc Natl Acad Sci U S A 96 14043 14048 10570195 Whittam TS 1996 Escherichia and Salmonella cellular and molecular biology Washington (DC) American Society of Microbiologists 2,720 p. Guttman DS Dykhuizen DE 1994 Clonal divergence in Escherichia coli as a result of recombination, not mutation Science 266 1380 1383 7973728 Finn TM Stevens LA 1995 Tracheal colonization factor: A Bordetella pertussis secreted virulence determinant Mol Microbiol 16 625 634 7476158 Kotob SI Hausman SZ Burns DL 1995 Localization of the promoter for the ptl genes of Bordetella pertussis, which encode proteins essential for secretion of pertussis toxin Infect Immun 63 3227 3230 7622254 Weiss AA Johnson FD Burns DL 1993 Molecular characterization of an operon required for pertussis toxin secretion Proc Natl Acad Sci U S A 90 2970 2974 8464913 Arico B Rappuoli R 1987 Bordetella parapertussis and Bordetella bronchiseptica contain transcriptionally silent pertussis toxin genes J Bacteriol 169 2847 2853 3584073 Isobe T White KA Allen AG Peacock M Raetz CR Maskell DJ 1999 Bordetella pertussis waaA encodes a monofunctional 2-keto-3-deoxy-D-manno-octulosonic acid transferase that can complement an Escherichia coli waaA mutation J Bacteriol 181 2648 2651 10198035 Preston A Allen AG Cadisch J Thomas R Stevens K Churcher CM 1999 Genetic basis for lipopolysaccharide O-antigen biosynthesis in Bordetellae Infect Immun 67 3763 3767 10417135 Preston A Maskell D 2001 The molecular genetics and role in infection of lipopolysaccharide biosynthesis in the Bordetellae J Endotoxin Res 7 251 261 11717578 Allen AG Isobe T Maskell DJ 1998 Identification and cloning of waaF (rfaF) from Bordetella pertussis and use to generate mutants of Bordetella spp. with deep rough lipopolysaccharide J Bacteriol 180 35 40 9422589 Allen A Maskell D 1996 The identification, cloning and mutagenesis of a genetic locus required for lipopolysaccharide biosynthesis in Bordetella pertussis Mol Microbiol 19 37 52 8821935 Peppler MS 1984 Two physically and serologically distinct lipopolysaccharide profiles in strains of Bordetella pertussis and their phenotype variants Infect Immun 43 224 232 6317567 Martin D Peppler MS Brodeur BR 1992 Immunological characterization of the lipooligosaccharide B band of Bordetella pertussis Infect Immun 60 2718 2725 1377181 Poolman JT Kuipers B Vogel ML Hamstra HJ Nagel J 1990 Description of a hybridoma bank towards Bordetella pertussis toxin and surface antigens Microb Pathog 8 377 382 1980000 van den Akker WM 1998 Lipopolysaccharide expression within the genus Bordetella : Influence of temperature and phase variation Microbiology 144 (Pt 6) 1527 1535 9639923 Felsenstein J 1978 Cases in which parsimony or compatibility methods will be positively misleading Systematic Biol 27 401 410 van Loo IH van der Heide HG Nagelkerke NJ Verhoef J Mooi FR 1999 Temporal trends in the population structure of Bordetella pertussis during 1949–1996 in a highly vaccinated population J Infect Dis 179 915 923 10068587 Mooi FR van Oirschot H Heuvelman K van der Heide HG Gaastra W Willems RJ 1998 Polymorphism in the Bordetella pertussis virulence factors P.69/pertactin and pertussis toxin in The Netherlands: Temporal trends and evidence for vaccine-driven evolution Infect Immun 66 670 675 9453625 Parkhill J Wren BW Thomson NR Titball RW Holden MT Prentice MB 2001 Genome sequence of Yersinia pestis, the causative agent of plague Nature 413 523 527 11586360 Chain PS Carniel E Larimer FW Lamerdin J Stoutland PO Regala WM 2004 Insights into the evolution of Yersinia pestis through whole-genome comparison with Yersinia pseudotuberculosis Proc Natl Acad Sci U S A 101 13826 13831 15358858 Nierman WC DeShazer D Kim HS Tettelin H Nelson KE Feldblyum T 2004 Structural flexibility in the Burkholderia mallei genome Proc Natl Acad Sci U S A 101 14246 14250 15377793 Kashimoto T Katahira J Cornejo WR Masuda M Fukuoh A Matsuzawa T 1999 Identification of functional domains of Bordetella dermonecrotizing toxin Infect Immun 67 3727 3732 10417130 Brockmeier SL Register KB Magyar T Lax AJ Pullinger GD Kunkle RA 2002 Role of the dermonecrotic toxin of Bordetella bronchiseptica in the pathogenesis of respiratory disease in swine Infect Immun 70 481 490 11796573 Mattoo S Yuk MH Huang LL Miller JF 2004 Regulation of type III secretion in Bordetella Mol Microbiol 52 1201 1214 15130135 Gupta S Maiden MC Feavers IM Nee S May RM Anderson RM 1996 The maintenance of strain structure in populations of recombining infectious agents Nat Med 2 437 442 8597954 Bjornstad ON Harvill ET 2005 Evolution and emergence of Bordetella in humans Trends Microbiol 13 355 359 15990312 Major RH 1945 Classic descriptions of disease Springfield (Illinois) C. C. Thomas Lapin JH 1943 Whooping cough Springfield (Illinois) C. C. Thomas Still GF 1931 The history of paediatrics London Oxford University Press 526 p. Gutierrez MC Brisse S Brosch R Fabre M Omais B Marmiesse M 2005 Ancient origin and gene mosaicism of the progenitor of Mycobacterium tuberculosis PLoS Pathog 1 e5. DOI: 10.1371/journal.ppat.0010005 . 16201017 Edwards KM Decker MD 2004 Pertussis vaccine Plotkin SA Orenstein WA Vaccines Philadelphia WB Saunders 2708 2720 David S van Furth R Mooi FR 2004 Efficacies of whole cell and acellular pertussis vaccines against Bordetella parapertussis in a mouse model Vaccine 22 1892 1898 15121300 Jolley KA Chan MS Maiden MC 2004 mlstdbNet—Distributed multi-locus sequence typing (MLST) databases BMC Bioinformatics 5 86 15230973 van der Zee A Agterberg C Peeters M Schellekens J Mooi FR 1993 Polymerase chain reaction assay for pertussis: Simultaneous detection and discrimination of Bordetella pertussis and Bordetella parapertussis J Clin Microbiol 31 2134 2140 8370741 Lesse AJ Campagnari AA Bittner WE Apicella MA 1990 Increased resolution of lipopolysaccharides and lipooligosaccharides utilizing tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis J Immunol Methods 126 109 117 2106001 Tsai CM Frasch CE 1982 A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels Anal Biochem 119 115 119 6176137 Margalef R 1958 Information theory in ecology General Systems 3 36 71 Rozas J Sanchez-DelBarrio JC Messeguer X Rozas R 2003 DnaSP, DNA polymorphism analyses by the coalescent and other methods Bioinformatics 19 2496 2497 14668244 Tusher VG Tibshirani R Chu G 2001 Significance analysis of microarrays applied to the ionizing radiation response Proc Natl Acad Sci U S A 98 5116 5121 11309499	16389302	PMC1323478	CC BY	2021-01-05 12:11:28	no	PLoS Pathog. 2005 Dec 30; 1(4):e45	utf-8	PLoS Pathog	2,005	10.1371/journal.ppat.0010045	oa_comm
==== Front PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1638334910.1371/journal.pmed.0030036EssayEpidemiology/Public HealthHealth PolicyOncologyScreeningCancer: gastroenterologicalGastroenterologyPublic HealthHealth PolicyThe Science and Politics of Colorectal Cancer Screening EssayHoff Geir Bretthauer Michael Geir Hoff and Michael Bretthauer are at the Institute of Population-Based Cancer Research, Oslo, Norway. Geir Hoff is also at the Telemark Hospital, Skien, Norway. Michael Bretthauer is also at the Department of Medicine, Rikshospitalet University Hospital, Oslo, Norway. To whom correspondence should be addressed. E-mail: [email protected] Competing Interests: The authors declare that they have no competing interests. 1 2006 3 1 2006 3 1 e36Copyright: © 2006 Hoff and Bretthauer.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.Hoff and Bretthauer discuss the scientific and political controversies surrounding the introduction of a national screening programme for colorectal cancer. ==== Body Politicians in developed countries have secured a certain standard of health care for the sick as well as vaccination programmes to reduce the risk of illness. For some of these countries, cancer screening is next on the agenda, with breast and cervical cancer screening programmes at the top of the list. When it comes to colorectal cancer (CRC), randomised trials of screening with the faecal occult blood test (FOBT) have shown that such screening reduces CRC incidence [1] and mortality [1–3]. Indirect evidence suggests that endoscopic screening (flexible sigmoidoscopy and colonoscopy) may be far more effective, but results from large-scale randomised trials are still awaited [4–7]. In the United States, CRC screening, using any of these screening modalities, has been recommended for many years [8–10]. The European Union now encourages its member states to initiate CRC screening, provided that it is done in an organised fashion and linked to quality assurance programmes [11]. Although the EU guidelines point out that only FOBT screening has been proven in randomised trials to have an effect on CRC, some member states (Poland, Germany, and Italy) have gone straight for “gold standard” colonoscopy screening without evidence from randomised trials [12,13]. The ideal screening method has not yet been found, and the future may see more use of promising screening modalities such as stool-based DNA tests [14] and virtual colonoscopy [15]. At present, however, the fact that these newer methods are still at an early stage of development may only serve as an excuse for politicians to “wait and see”, and thus further postpone a decision on CRC screening. In this essay, we discuss the scientific and political controversies surrounding the introduction of a national CRC screening programme. We argue that it is in the best interest of governments to fund well-conducted randomised trials that can guide their decision making. Public Awareness of CRC Enthusiastic health professionals often find politicians and health authorities too slow to respond to the considerable evidence in favour of CRC screening [13], and they may take on a political role themselves in promoting their good cause. In addition, a number of celebrities—including Ronald Reagan, Katie Couric (an American television personality), and Lynn Faulds Wood (a British television presenter)—have helped to raise public awareness of CRC ([16]; http://www.bowelcancer.tv/cgi-bin/page.pl). His Holiness, the late Pope Paul John II, promoted CRC screening by being the first patron of the International Digestive Cancer Alliance, with strong support from United States Senator Hillary Clinton [17]. But while there is growing public awareness of CRC, there has been disturbingly little public discussion about the scientific pros and cons of CRC screening compared with what some countries have seen for breast cancer screening. One of the main reasons for nonparticipation in CRC screening has been that people are not convinced that they will gain anything by participating [18]. Similarly, the failure of some physicians to recommend screening may reflect a general need for more convincing evidence and a convincing presentation of such evidence. Women being offered mammographic screening for breast cancer have expressed a particular need for information about logistics as well as about false positive and negative tests, and they want to play an active part in decision making on screening [19]. This need for balanced information on screening applies to other types of cancer screening and should be the “standard of care” for future cancer screening awareness campaigns. By failing to recommend screening, physicians representing the medical profession may appear divided in their view to screen or not to screen, and undecided on which method to use, although the health problem of CRC is recognised by all. In many countries, CRC is one of the two most frequently encountered cancers for men and women collectively. Five-year survival is only about 50%, increasing to 80% if diagnosed at an early, asymptomatic stage [20]. Thus, early diagnosis and treatment is the obvious key to improved cure and survival. Some enthusiasts initially recommending national screening programmes with continuous quality assurance now accept opportunistic screening (i.e., screening done outside of a national screening programme), since “public health policy has not yet included population CRC screening” [13]. On the other hand, many doctors feel that there are still important questions to be answered before implementing national screening programmes. Besides, some are concerned that not enough is being done for patients who need basic health care rather than preventive measures. How Good Is the Evidence for Screening? So what kind of evidence can we offer to increase professional enthusiasm for CRC screening? Do we only have the support of celebrities and little else? FOBT is the only screening method submitted to adequately designed randomised trials, which showed 15%–33% mortality reduction from CRC after 8–13 years' follow-up [1–3]. The relative mortality reduction was significant in all three trials, proving without doubt that “FOBT screening does work”. But how good is this in absolute, not relative, terms? With 5% lifetime risk of CRC and 50% mortality from the disease, the risk of dying from CRC is 2.5% without FOBT screening. With biennial FOBT screening, this may be reduced by 23% for those attending, i.e., reduced by 0.6% to 1.9% if attendance is 100%. About 98% of us will die from something else. How do you sell such figures both to doctors and to the public? In the US, the CDC (Centers for Disease Control and Prevention) is promoting CRC screening for people over 50 years old In contrast to FOBT, we only have case-control studies to support the usefulness of endoscopy screening, and such studies are particularly prone to overestimate the effectiveness of screening tests when compared with randomised trials [21]. The first reports from the four ongoing randomised trials on flexible sigmoidoscopy screening are expected over the course of the next several years [4–7], while similar randomised trials on colonoscopy screening have not, so far, been launched. Mortality reduction by colonoscopy screening is expected, but not proven, to be much higher than the 23% for those attending FOBT screening, in addition to an incidence reduction expected through polypectomies. There has been some concern about a possible excess number of non-CRC deaths in the screening group in CRC screening trials [22,23]. In addition, a recent 17-year follow-up of the Danish FOBT study showed that the significant reduction in mortality seen after about ten years could no longer be observed, probably due to poorer attendance with an increasing number of screening rounds [24]. Since high attendance rates are crucial for the success of screening with the FOBT, the best choice among the present CRC screening options may differ between countries and cultures, thus indicating a need for research on CRC screening referable to the target population. Furthermore, there is only limited knowledge about what impact an increasing menu of screening services might have on the responsibility people take for their own health. For example, patients may abandon healthy lifestyles (exercising, eating plenty of fruit and vegetables, giving up smoking) if they believe that screening will pick up cancers at an early stage. How Far Should We Compromise on Scientific Proof? The crucial question is the following: how much evidence (and at what level in the well-known hierarchy of evidence [http://www.shef.ac.uk/scharr/ir/units/systrev/hierarchy.htm]) is required before implementing national screening programmes? The World Health Organization guidelines for screening request evidence for mortality and morbidity reduction from high-quality randomised controlled trials before starting any population screening programme [25]. In other fields of medicine, such as pharmacotherapy, randomised trials have been a standard requirement for introducing any new drug on the market. Many health services have admittedly been introduced based on much weaker evidence than we now have for CRC screening. But historic substandards should be no argument for turning a blind eye to the need for the best-quality evidence or for allowing health policy decisions to be based more on ideology and convictions than science. Also, the concept of pushing a screening service onto presumably healthy people is quite different from establishing a health service that people seek when they feel sick. The level of evidence should therefore be more foolproof for screening services than for other services, and the evidence should be able to withstand debates such as the one surrounding mammography. High-quality evidence on CRC screening would also help to unite the somewhat divergent professional views on this intervention. The Finnish Model We need to recognise that some countries feel that the introduction of national CRC screening programmes is an urgent matter. Ideally, these should be rolled out in a stepwise, randomised fashion, just as Finland has done (see Box 1), allowing evaluation of its FOBT screening programme after five years before deciding what their next step of action should be [26]. The people behind the Finnish strategy deserve credit for persuading their politicians to choose this cautious, stepwise model, and the politicians and health authorities deserve credit for listening. Box 1. Introducing a CRC Screening Programme: The Finnish Approach In 2004, Finland launched a randomised, stepwise introduction of CRC screening using FOBT (biennial, unrehydrated Hemoccult-II). For the first six years of introduction, each age cohort is randomised to screening or “care as usual” (no screening) at age 60–64 years. In six years, 50% of the entire target population will be covered. This provides an opportunity to evaluate the programme after five years and to further adjust the screening strategy or to implement FOBT screening to all Finnish citizens aged 60–69 within the ensuing five years. Presently, representatives from the five Nordic countries are collaborating to develop the Nordic Initiative on Colorectal Cancer, an investigator-initiated protocol with an aim to further the scientific knowledge of CRC screening. In the Finnish model, half of each age cohort is randomised to screening or no screening. The Finnish model must have required a lot of explanation to authorities that this approach was clearly the best way to proceed. It was, of course, risky for politicians to voluntarily throw away half (or more) of their target candidate supporters by declaring, in essence, “We believe in CRC screening, but aren't sure about it, and half of you will be offered screening while the other half will not”. In contrast to the Finnish approach, there have been examples of political decisions to simply screen everyone or none at all, i.e., to reject the idea of large-scale randomised trials. This rejection may leave health-care providers and doctors with a very difficult choice: either to go straight for national screening or to reject the political offer and wait for a new set of politicians who may understand the need for further randomised trials. Short-Term Gain, Long-Term Pain? For those who have experienced CRC or have been close to someone suffering from it (a considerable number in Westernised societies), any proven benefit from screening will do. Furthermore, CRC screening may appear politically attractive and tempting to politicians when campaigning for a general election. And cost–benefit estimates have even suggested that such screening may be economically sound and at least as cost-effective as established national screening programmes (for breast and cervical cancer) [27]. But politicians should be concerned about whether CRC screening is the right way to spend available resources and taxpayers' money for the benefit of individuals, families, employers, and the community. There are still few data to guide decisions about CRC screening. It is hard to understand why politicians are reluctant to invest in randomised clinical trials, since the results of these will equip them to make better political decisions in the end. The alternative to taking a firm grip on development toward national programmes through the funding of large, well-designed randomised controlled trials seems to be uncontrolled out-of-pocket expenditure on sporadic screening, which cannot generate any real evidence about the effectiveness of screening. Any country considering national CRC screening can surely afford to do its own randomised trials, but time is running out. Could it be that none of the providers really want to know, but prefer to make decisions by convictions and “gut feelings” for short-term gain? For a screening programme to survive over time, it has to deliver according to expectations. If expectations are not based on trials referable to the target population, any programme will be vulnerable to devastating debates, such as those surrounding mammographic screening seen in some countries. Health policymakers must also remain sceptical of the role of celebrity endorsements. Communication on complex decisions such as cancer screening, with an aim to inform rather than persuade, is not an obvious task for celebrities [28]. Likewise, it is not an obvious task for doctors to take on the role of politicians. Instead, politicians must accept the need for more science in their decision making. We need more objective facts, relevant to the target population, to be communicated—and not personal convictions that doctors, doctor-politicians, and politicians only present as “facts”. To paraphrase former US President Franklin Delano Roosevelt: “Look to Finland”. Citation: Hoff G, Bretthauer M (2006) The science and politics of colorectal cancer. PLoS Med 3(1): e36. Abbreviations CRCcolorectal cancer FOBTfaecal occult blood test ==== Refs References Mandel JS Church TR Bond JH Ederer F Geisser MS The effect of fecal occult blood screening on the incidence of colorectal cancer N Engl J Med 2000 343 1603 1607 11096167 Hardcastle JD Chamberlain JO Robinson MH Ross SM Amar SS Randomised controlled trial of faecal-occult-blood screening for colorectal cancer Lancet 1996 348 1472 1477 8942775 Kronborg O Fenger C Olsen J Jorgensen OD Sondergaard O Randomised study of screening for colorectal cancer with faecal-occult-blood test Lancet 1996 348 1467 1471 8942774 United Kingdom Flexible Sigmoidoscopy Screening Trial Investigators Single flexible sigmoidoscopy screening to prevent colorectal cancer: Baseline findings of a UK multicentre randomised trial Lancet 2002 359 1291 1300 11965274 Segnan N Senore C Andreoni B Aste H Bonelli L Baseline findings of the Italian multicenter randomized controlled trial of “once-only sigmoidoscopy”—SCORE J Natl Cancer Inst 2002 94 1763 1772 12464648 Gondal G Grotmol T Hofstad B Bretthauer M Eide TJ The Norwegian Colorectal Cancer Prevention (NORCCAP) screening study: Baseline findings and implementations for clinical work-up in age groups 50–64 years Scand J Gastroenterol 2003 38 635 642 12825872 Palitz AM Selby JV Grossman S Finkler LJ Bevc M The colon cancer prevention program (PLCO): Rationale, implementation and preliminary results HMO Pract 1997 11 5 12 10165556 Winawer S Fletcher R Rex D Bond J Burt R Colorectal cancer screening and surveillance: Clinical guidelines and rationale—Update based on new evidence Gastroenterology 2003 124 544 560 12557158 Rex DK Johnson DA Lieberman DA Burt RW Sonnenberg A Colorectal cancer prevention 2000: Screening recommendations of the American College of Gastroenterology Am J Gastroenterol 2000 95 868 877 10763931 American Cancer Society American Cancer Society guidelines 2004 Oakland (California) American Cancer Society Available: http://www.cancer.org/docroot/PED/content/PED_2_3X_ACS_Cancer_Detection_Guidelines_36.asp . Accessed 15 November 2005 Boyle P Autier P Bartelink H Baselga J Boffetta P European code against cancer and scientific justification: Third version Ann Oncol 2003 14 973 1005 12853336 Butruk E Regula J Polkowski M Rupinski M Przytulski K National colorectal cancer screening programme in Poland Endoscopy 2002 34 939 940 Rozen P Winawer SJ Report of the OMED colorectal cancer screening committee meeting, New Orleans, 2004—In collaboration with the IDCA Eur J Cancer Prev 2004 13 461 464 15452461 Osborn NK Ahlquist DA Stool screening for colorectal cancer: Molecular approaches Gastroenterology 2005 128 192 206 15633136 Pickhardt PJ Choi JR Hwang I Butler JA Puckett ML Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults N Engl J Med 2003 349 2191 2199 14657426 Gorman C Everything you need to know about colon cancer and how to prevent it Time 2000 March 40-46 Clinton HR Senator Clinton's letter of support 2002 Edinburgh World Gastroenterology Organisation Available: http://www.worldgastroenterology.org/idca/clintonletter.html . Accessed 16 November 2005 Harewood GC Wiersema MJ Melton LJ A prospective, controlled assessment of factors influencing acceptance of screening colonoscopy Am J Gastroenterol 2002 97 3186 3194 12492209 Nekhlyudov L Li R Fletcher SW Information and involvement preferences of women in their 40s before their first screening mammogram Arch Intern Med 2005 165 1370 1374 15983285 Di Gregorio C Benatti P Losi L Roncucci L Rossi G Incidence and survival of patients with Dukes' A (stages T1 and T2) colorectal carcinoma: A 15-year population-based study Int J Colorectal Dis 2005 20 147 154 15592853 Mandel JS Sigmoidoscopy screening probably works, but how well is still unknown J Natl Cancer Inst 2003 95 571 573 12697843 Hoff G CRC screening: Review of the evidence and suggestions on when and how to move on from randomised trials to screening programmes Scand J Gastroenterol 2004 38 99 103 Hoff G In response to Ederer, Church and Mandel Scand J Gastroenterol 2004 39 1029 1032 Kronborg O Jørgensen O Fenger C Rasmussen M Randomised study of biennial screening with a faecal occult blood test: Results after nine screening rounds Scand J Gastroenterol 2004 39 846 851 15513382 World Health Organization National Cancer Control Programmes Policies and managerial guidelines. 2nd Edition 2002 Geneva World Health Organization Available: http://www.who.int/entity/cancer/media/en/408.pdf . Accessed 15 November 2005 Malila N Anttila A Hakama M Colorectal cancer screening in Finland: Details of the national screening programme implemented in autumn 2004 J Med Screen 2005 12 28 32 15814016 Geul KW Bosman FT van Blankenstein M Grobbee DE Wilson JH Prevention of colorectal cancer. Costs and effectiveness of sigmoidoscopy Scand J Gastroenterol Suppl 1997 223 79 87 9200311 Larson RJ Woloshin S Schwartz LM Welch HG Celebrity endorsements of cancer screening J Natl Cancer Inst 2005 97 693 695 15870440	16383349	PMC1323498	CC BY	2021-01-05 10:40:33	no	PLoS Med. 2006 Jan 3; 3(1):e36	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030036	oa_comm
==== Front PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1638335010.1371/journal.pmed.0030057PerspectivesImmunologyInfectious DiseasesVirologyAllergy/ImmunologyHIV/AIDSHIV Infection/AIDSInfectious DiseasesImmunology and allergyMicrobiologyNew Approaches to Vaccine Adjuvants: Inhibiting the Inhibitor PerspectiveGraham Barney S Barney S. Graham is a Senior Investigator at the Vaccine Research Center, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland, United States of America. E-mail: [email protected] Competing Interests: The author declares that no competing interests exist. 1 2006 3 1 2006 3 1 e57Copyright: 2006This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration, which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. An Alternative and Effective HIV Vaccination Approach Based on Inhibition of Antigen Presentation Attenuators in Dendritic Cells Graham discusses a study that explores new approaches to enhancing vaccine-induced immune responses. ==== Body Development of an effective HIV-1 vaccine has been an elusive goal for over 20 years despite being an urgent global priority. The agonizingly slow progress is not from lack of effort, but is a consequence of the insidious biology of the virus. HIV-1 has many features that make vaccine development challenging, if not impossible [1]. Obstacles to Vaccine Development Pessimism is based in part on the empirical observation that there has never been a confirmed case of viral clearance and recovery from HIV-1 infection, and from the mounting evidence that HIV-1 superinfection (see Glossary) is common (in other words, if natural infection does not protect against infection with other HIV strains, why would we expect vaccination to offer protection?) [2]. Glossary gp120: The extracellular portion of the HIV-1 Env glycoprotein responsible for binding to CD4+ and co-receptors. gp120-pulsed bone marrow–derived dendritic cells: Dendritic cells are key antigen-presenting cells that in this case were derived from mouse bone marrow and expanded with interleukin-4 and granulocyte/monocyte colony stimulating factor in vitro before treatment with gp120. gp160: Full length HIV-1 Env glycoprotein of 160 kD molecular weight responsible for attachment to and entry into target cells. IRF-3: Transcription factor, interferon regulatory factor-3. JAKs: JAKs associate with cytokine receptors, and are important for tyrosine phosphorylation of the receptor, of each other, and of signal transduction and activator-of-transcription molecules that participate in the signaling cascade from the cytokine receptors to the nucleus. LPS: Lipopolysaccharide is analogous to endotoxin derived from Gram-negative bacteria. MyD88: Myeloid differentiation factor 88. NF-κB: Transcription factor, nuclear factor-κB. PAMP: Pathogen-associated molecular pattern. PolyI:C: Polyinosinic:polycytidylic acid is a synthetic molecule that mimics double-stranded RNA. R837: (1-(2-methyl propyl)-1H-imidazo[4,5-c]quinolin 4-amine} is a TLR-7 and TLR-8 ligand. siRNA inhibition: Small interfering RNAs are approximately 22 nucleotide-long RNA molecules that efficiently inhibit translation of their complementary mRNA. STAT: Signal transduction and activator-of-transcription molecules associate in dimers after phosphorylation by JAKs, translocate to the nucleus, and promote transcription of selected genes. Superinfection: Infection with a strain of HIV-1 that is genetically distinct from the HIV-1 present in a person with a stable immune response to the original infection. Th1 response: A T helper 1 response implies a polarized CD4+ T cell response with dominant expression of interferon-γ. In mice, this is associated with a predominant IgG2a antibody isotype response. Th2 response: A T helper 2 response implies a polarized CD4+ T cell response with dominant expression of interleukin-4, interleukin-5, interleukin-9, and interleukin-13. In mice, this is associated with a predominant IgG1 antibody isotype response. TIR: A cytoplasmic signaling domain on TLRs, Toll/interleukin-1 receptor–resistance domain. TLR-3: Toll-like receptor 3 recognizes double-stranded RNA including the synthetic molecule polyI:C. TLR-4: Toll-like receptor 4 recognizes LPS. TLR-7/8: Toll-like receptors 7 and 8 recognize single-stranded RNA and some other ligands such as imiquimod and resiquimod. Toll-like receptor: Family of proteins homologous to the Drosophila Toll receptor, found to have the capacity to recognize molecular patterns associated with pathogens. TRIF: TIR domain-containing adapter-inducing interferon, also called TICAM-1 (Toll-IL-1 receptor homology domain-containing adapter molecule). The high replication error rate of HIV-1 is the basis for two of the greatest challenges in vaccine design. First, extreme genetic diversity makes vaccine antigen selection difficult. Second, the high mutation rate provides many opportunities for the virus to escape vaccine-induced immune responses. In addition, HIV-1 glycoprotein (gp)160 can directly cause dysfunction of antigen-presenting cells, and HIV-1 can infect CD4+ T cells, which cripples the key elements required to initiate the adaptive immune response to viral pathogens. After cells are infected, there are virus-specific mechanisms that disrupt the normal regulation of immune activation, and HIV-1 can also become latent or infect immunoprivileged sites and remain hidden from the immune response. The list of barriers to vaccine development continues with the multitude of structural features of gp160 that evade antibody neutralization [3]. Therefore, an effective HIV-1 vaccine will need to induce immune responses that can swiftly respond to infection and efficiently clear or control infection at very low levels of replication. To do this may require better vaccine adjuvants or delivery vehicles than are currently available. A Novel Approach to Vaccine Development In a new study in PLoS Medicine, Song et al. explore new approaches to enhancing vaccine-induced immune responses [4]. The field of vaccine adjuvants has been rapidly evolving since the discovery of the Toll-like receptor (TLR) family of pattern recognition molecules [5]. Traditionally, adjuvants have been developed empirically from natural substances found to cause inflammation. Since cytokines were found to be the effector molecules for many adjuvant effects [6], there has been an effort to build the optimal vaccine adjuvant effect one cytokine at a time. However, this approach underestimated the complexity and timing of events necessary to augment immune responses, and has given way to using specific TLR ligands as vaccine adjuvants [7]. Activating immune responses at this level is analogous to using the original empirically derived adjuvants, but the molecular mechanisms are better defined. In the current study, Song and colleagues have achieved an even broader effect by inhibiting a natural inhibitor of TLR and cytokine signaling. The family of suppressor of cytokine signaling (SOCS) molecules targets Janus kinases (JAKs) and nuclear factor-κB (NF-κ B) pathways involved in transmitting signals from cytokine receptors and TLRs to the nucleus to control genes encoding mediators of inflammation (Figure 1) [8]. Figure 1 SOCS “Silencing” Enhances the Adjuvant Effect of TLR Ligands Substances common in viral pathogens (substances such as single-stranded RNA [ssRNA] and double-stranded RNA [dsRNA]) or in bacteria (substances such as endotoxin) are recognized by TLRs as pathogen-associated molecular patterns. When the TLRs are triggered, a series of signaling events occur that is simplified and schematized in this figure. These events lead to inflammation and activation of innate and adaptive immune responses. SOCS family members are also activated, and serve as an internal control to diminish the intensity and duration of inflammation. In the left-hand panel, R837 mimics ssRNA as a ligand for TLR-7, and initiates signaling through the MyD88 pathway, eventually resulting in the release of NF-κB and in the upregulation of SOCS1 and many genes involved in inflammation. PolyI:C is a synthetic mimic of dsRNA and triggers TLR-3-associated JAK/signal transduction and activator-of-transcription (STAT) signaling through TIR domain-containing adapter-inducing interferon (TRIF), activation of interferon regulatory factor (IRF)-3, and increased production of type 1 interferon. LPS can activate both pathways. SOCS1 specifically interferes with JAK2 and may also inhibit a step between MyD88 and NF-κB release into the nucleus, although this is controversial. This balanced internal feedback mechanism results in a controlled inflammatory process with adjuvant and antiviral effects. When SOCS1 production is blocked by siRNA (right-hand panel), the control of the inflammatory process is temporarily lost, leading to a greater adjuvant and antiviral response that appears to improve vaccine-induced immune responses. (Illustration: Giovanni Maki) The authors found that small interfering RNA (siRNA) inhibition of SOCS1 in HIV-1 gp120-pulsed bone marrow–derived dendritic cells (DCs) improved the immune response to those DCs delivered as a carrier of vaccine antigen. In addition, SOCS1 inhibition improved the adjuvant effect of polyinosinic:polycytidylic acid (polyI:C), R837, or lipopolysaccharide (LPS) on the DC vaccine. These adjuvants are recognized by TLR-3, TLR-7/8, or TLR-4, respectively. SOCS1 inhibition increased the magnitude of the antibody response and the magnitude and lytic activity of the CD8+ T cell response. It also appeared to increase the duration of antibody and T cell responses. SOCS1 inhibition increased cytokine production from in vitro–stimulated DCs and also led to increased cytokine production from CD4+ T cells. The pattern of increased cytokine production was broad and included proinflammatory cytokines, as well as cytokines traditionally associated with polarized T helper cell (Th)1 or Th2 responses. Importantly, in vivo delivery of plasmid DNA expressing the SOCS1 siRNA, together with a plasmid DNA vaccine encoding a modified HIV-1 envelope (Env) protein, improved Env-specific immune responses in mice, particularly when the polyI:C or the R837 adjuvant was administered following immunization. Implications of the Approach We maintain a tenuous balance between adequate control of pathogens or neoplasms and excessive chronic inflammation (e.g., autoimmune diseases) and uncontrolled proliferation (e.g., lymphoma) of our immune system. There are more pathways and receptors devoted to controlling the immune response than to inducing immune responses. It is critical for health maintenance to have a highly evolved process for turning off immune responses when there are so many inflammatory challenges. Therefore, it is not surprising that enhancing immune responses by removing inhibition may be more potent than the typical adjuvant concept of actively stimulating immune responses. This has been noted before when evaluating the adjuvant effects of cytokines [9]. Song and colleagues' study has shown the potential to broadly augment immune activation triggered by vaccine antigens and a variety of adjuvants. Their approach allows the antigen-presenting cell to determine the composition and kinetics of effector molecules and the co-stimulation needed for the development of potent adaptive immune responses. Before considering this approach for clinical use, it will be important to determine whether the immune response enhancement seen in mice can be achieved in higher-order animals such as nonhuman primates. The safety of this approach will need to be more fully defined in species other than mice. It will be important to show that SOCS inhibition is temporary, and that normal regulation of immune responses is maintained following the vaccination. This may require long-term evaluation in multiple species to assure there is not excessive or prolonged inflammation. Identifying pharmacological inhibitors of SOCS family members may provide a more temporally controlled approach for releasing the inhibition of TLR and cytokine signaling, and provide an additional safeguard against the theoretical risk of chronic inflammation. As the authors suggest, waking up the immune system by inhibiting the inhibitors may have therapeutic benefits in some settings. The struggle to develop a preventive vaccine for HIV-1 has taught us much of what we know about immunology, and will continue to have benefits at the conceptual level and for antiviral vaccine development in general. HIV-1 vaccine development will be facilitated by innovative approaches to vaccine formulation and delivery such as SOCS “silencing” that improve the kinetics, the magnitude, and the composition of immune responses. However, ultimate success will depend on the design of vaccine antigens that can elicit the right antibody and T cell specificity to achieve virus neutralization and clearance. I thank Kathryn L. Bonaparte for her critical review and comments. Citation: Graham BS (2006) New approaches to vaccine adjuvants: Inhibiting the inhibitor. PLoS Med 3(1): e57. Abbreviations DCdendritic cell Envenvelope gpglycoprotein JAKJanus kinase LPSlipopolysaccharide NF-κ,Bnuclear factor-κB polyI: Cpolyinosinic:polycytidylic acid siRNAsmall interfering RNA SOCSsuppressor of cytokine signaling ThT helper cell TLRToll-like receptor ==== Refs References Sabin AB Improbability of effective vaccination against human immunodeficiency virus because of its intracellular transmission and rectal portal of entry Proc Natl Acad Sci U S A 1992 89 8852 8855 1528902 Smith DM Richman DD Little SJ HIV superinfection J Infect Dis 2005 192 438 444 15995957 Burton DR Desrosiers RC Doms RW Koff WC Kwong PD HIV vaccine design and the neutralizing antibody problem Nat Immunol 2004 5 233 236 14985706 Song XT Evel-Kabler K Rollins L Aldrich M Gao F An alternative and effective HIV vaccination approach based on inhibition of natural immune inhibitors in dendritic cells PLoS Med 2006 3 e11 10.1371/journal.pmed.0030011 16381597 Medzhitov R Preston-Hurlburt P Janeway CA A human homologue of the Drosophila Toll protein signals activation of adaptive immunity Nature 1997 388 394 397 9237759 Afonso LC Scharton TM Vieira LQ Wysocka M Trinchieri G The adjuvant effect of interleukin-12 in a vaccine against Leishmania major Science 1994 263 235 237 7904381 Weeratna RD Makinen SR McCluskie MJ Davis HL TLR agonists as vaccine adjuvants: Comparison of CpG ODN and Resiquimod (R-848) Vaccine 2005 23 5263 5270 16081189 Alexander WS Hilton DJ The role of suppressors of cytokine signaling (SOCS) proteins in regulation of the immune response Annu Rev Immunol 2004 22 503 529 15032587 Tang YW Graham BS Anti-IL-4 treatment at immunization modulates cytokine expression, reduces illness, and increases cytotoxic T lymphocyte activity in mice challenged with respiratory syncytial virus J Clin Invest 1994 94 1953 1958 7962541	16383350	PMC1323499	CC0	2021-01-05 10:40:33	no	PLoS Med. 2006 Jan 3; 3(1):e57	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030057	oa_comm
==== Front PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1638335010.1371/journal.pmed.0030057PerspectivesImmunologyInfectious DiseasesVirologyAllergy/ImmunologyHIV/AIDSHIV Infection/AIDSInfectious DiseasesImmunology and allergyMicrobiologyNew Approaches to Vaccine Adjuvants: Inhibiting the Inhibitor PerspectiveGraham Barney S Barney S. Graham is a Senior Investigator at the Vaccine Research Center, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland, United States of America. E-mail: [email protected] Competing Interests: The author declares that no competing interests exist. 1 2006 3 1 2006 3 1 e57Copyright: 2006This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration, which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. An Alternative and Effective HIV Vaccination Approach Based on Inhibition of Antigen Presentation Attenuators in Dendritic Cells Graham discusses a study that explores new approaches to enhancing vaccine-induced immune responses. ==== Body Development of an effective HIV-1 vaccine has been an elusive goal for over 20 years despite being an urgent global priority. The agonizingly slow progress is not from lack of effort, but is a consequence of the insidious biology of the virus. HIV-1 has many features that make vaccine development challenging, if not impossible [1]. Obstacles to Vaccine Development Pessimism is based in part on the empirical observation that there has never been a confirmed case of viral clearance and recovery from HIV-1 infection, and from the mounting evidence that HIV-1 superinfection (see Glossary) is common (in other words, if natural infection does not protect against infection with other HIV strains, why would we expect vaccination to offer protection?) [2]. Glossary gp120: The extracellular portion of the HIV-1 Env glycoprotein responsible for binding to CD4+ and co-receptors. gp120-pulsed bone marrow–derived dendritic cells: Dendritic cells are key antigen-presenting cells that in this case were derived from mouse bone marrow and expanded with interleukin-4 and granulocyte/monocyte colony stimulating factor in vitro before treatment with gp120. gp160: Full length HIV-1 Env glycoprotein of 160 kD molecular weight responsible for attachment to and entry into target cells. IRF-3: Transcription factor, interferon regulatory factor-3. JAKs: JAKs associate with cytokine receptors, and are important for tyrosine phosphorylation of the receptor, of each other, and of signal transduction and activator-of-transcription molecules that participate in the signaling cascade from the cytokine receptors to the nucleus. LPS: Lipopolysaccharide is analogous to endotoxin derived from Gram-negative bacteria. MyD88: Myeloid differentiation factor 88. NF-κB: Transcription factor, nuclear factor-κB. PAMP: Pathogen-associated molecular pattern. PolyI:C: Polyinosinic:polycytidylic acid is a synthetic molecule that mimics double-stranded RNA. R837: (1-(2-methyl propyl)-1H-imidazo[4,5-c]quinolin 4-amine} is a TLR-7 and TLR-8 ligand. siRNA inhibition: Small interfering RNAs are approximately 22 nucleotide-long RNA molecules that efficiently inhibit translation of their complementary mRNA. STAT: Signal transduction and activator-of-transcription molecules associate in dimers after phosphorylation by JAKs, translocate to the nucleus, and promote transcription of selected genes. Superinfection: Infection with a strain of HIV-1 that is genetically distinct from the HIV-1 present in a person with a stable immune response to the original infection. Th1 response: A T helper 1 response implies a polarized CD4+ T cell response with dominant expression of interferon-γ. In mice, this is associated with a predominant IgG2a antibody isotype response. Th2 response: A T helper 2 response implies a polarized CD4+ T cell response with dominant expression of interleukin-4, interleukin-5, interleukin-9, and interleukin-13. In mice, this is associated with a predominant IgG1 antibody isotype response. TIR: A cytoplasmic signaling domain on TLRs, Toll/interleukin-1 receptor–resistance domain. TLR-3: Toll-like receptor 3 recognizes double-stranded RNA including the synthetic molecule polyI:C. TLR-4: Toll-like receptor 4 recognizes LPS. TLR-7/8: Toll-like receptors 7 and 8 recognize single-stranded RNA and some other ligands such as imiquimod and resiquimod. Toll-like receptor: Family of proteins homologous to the Drosophila Toll receptor, found to have the capacity to recognize molecular patterns associated with pathogens. TRIF: TIR domain-containing adapter-inducing interferon, also called TICAM-1 (Toll-IL-1 receptor homology domain-containing adapter molecule). The high replication error rate of HIV-1 is the basis for two of the greatest challenges in vaccine design. First, extreme genetic diversity makes vaccine antigen selection difficult. Second, the high mutation rate provides many opportunities for the virus to escape vaccine-induced immune responses. In addition, HIV-1 glycoprotein (gp)160 can directly cause dysfunction of antigen-presenting cells, and HIV-1 can infect CD4+ T cells, which cripples the key elements required to initiate the adaptive immune response to viral pathogens. After cells are infected, there are virus-specific mechanisms that disrupt the normal regulation of immune activation, and HIV-1 can also become latent or infect immunoprivileged sites and remain hidden from the immune response. The list of barriers to vaccine development continues with the multitude of structural features of gp160 that evade antibody neutralization [3]. Therefore, an effective HIV-1 vaccine will need to induce immune responses that can swiftly respond to infection and efficiently clear or control infection at very low levels of replication. To do this may require better vaccine adjuvants or delivery vehicles than are currently available. A Novel Approach to Vaccine Development In a new study in PLoS Medicine, Song et al. explore new approaches to enhancing vaccine-induced immune responses [4]. The field of vaccine adjuvants has been rapidly evolving since the discovery of the Toll-like receptor (TLR) family of pattern recognition molecules [5]. Traditionally, adjuvants have been developed empirically from natural substances found to cause inflammation. Since cytokines were found to be the effector molecules for many adjuvant effects [6], there has been an effort to build the optimal vaccine adjuvant effect one cytokine at a time. However, this approach underestimated the complexity and timing of events necessary to augment immune responses, and has given way to using specific TLR ligands as vaccine adjuvants [7]. Activating immune responses at this level is analogous to using the original empirically derived adjuvants, but the molecular mechanisms are better defined. In the current study, Song and colleagues have achieved an even broader effect by inhibiting a natural inhibitor of TLR and cytokine signaling. The family of suppressor of cytokine signaling (SOCS) molecules targets Janus kinases (JAKs) and nuclear factor-κB (NF-κ B) pathways involved in transmitting signals from cytokine receptors and TLRs to the nucleus to control genes encoding mediators of inflammation (Figure 1) [8]. Figure 1 SOCS “Silencing” Enhances the Adjuvant Effect of TLR Ligands Substances common in viral pathogens (substances such as single-stranded RNA [ssRNA] and double-stranded RNA [dsRNA]) or in bacteria (substances such as endotoxin) are recognized by TLRs as pathogen-associated molecular patterns. When the TLRs are triggered, a series of signaling events occur that is simplified and schematized in this figure. These events lead to inflammation and activation of innate and adaptive immune responses. SOCS family members are also activated, and serve as an internal control to diminish the intensity and duration of inflammation. In the left-hand panel, R837 mimics ssRNA as a ligand for TLR-7, and initiates signaling through the MyD88 pathway, eventually resulting in the release of NF-κB and in the upregulation of SOCS1 and many genes involved in inflammation. PolyI:C is a synthetic mimic of dsRNA and triggers TLR-3-associated JAK/signal transduction and activator-of-transcription (STAT) signaling through TIR domain-containing adapter-inducing interferon (TRIF), activation of interferon regulatory factor (IRF)-3, and increased production of type 1 interferon. LPS can activate both pathways. SOCS1 specifically interferes with JAK2 and may also inhibit a step between MyD88 and NF-κB release into the nucleus, although this is controversial. This balanced internal feedback mechanism results in a controlled inflammatory process with adjuvant and antiviral effects. When SOCS1 production is blocked by siRNA (right-hand panel), the control of the inflammatory process is temporarily lost, leading to a greater adjuvant and antiviral response that appears to improve vaccine-induced immune responses. (Illustration: Giovanni Maki) The authors found that small interfering RNA (siRNA) inhibition of SOCS1 in HIV-1 gp120-pulsed bone marrow–derived dendritic cells (DCs) improved the immune response to those DCs delivered as a carrier of vaccine antigen. In addition, SOCS1 inhibition improved the adjuvant effect of polyinosinic:polycytidylic acid (polyI:C), R837, or lipopolysaccharide (LPS) on the DC vaccine. These adjuvants are recognized by TLR-3, TLR-7/8, or TLR-4, respectively. SOCS1 inhibition increased the magnitude of the antibody response and the magnitude and lytic activity of the CD8+ T cell response. It also appeared to increase the duration of antibody and T cell responses. SOCS1 inhibition increased cytokine production from in vitro–stimulated DCs and also led to increased cytokine production from CD4+ T cells. The pattern of increased cytokine production was broad and included proinflammatory cytokines, as well as cytokines traditionally associated with polarized T helper cell (Th)1 or Th2 responses. Importantly, in vivo delivery of plasmid DNA expressing the SOCS1 siRNA, together with a plasmid DNA vaccine encoding a modified HIV-1 envelope (Env) protein, improved Env-specific immune responses in mice, particularly when the polyI:C or the R837 adjuvant was administered following immunization. Implications of the Approach We maintain a tenuous balance between adequate control of pathogens or neoplasms and excessive chronic inflammation (e.g., autoimmune diseases) and uncontrolled proliferation (e.g., lymphoma) of our immune system. There are more pathways and receptors devoted to controlling the immune response than to inducing immune responses. It is critical for health maintenance to have a highly evolved process for turning off immune responses when there are so many inflammatory challenges. Therefore, it is not surprising that enhancing immune responses by removing inhibition may be more potent than the typical adjuvant concept of actively stimulating immune responses. This has been noted before when evaluating the adjuvant effects of cytokines [9]. Song and colleagues' study has shown the potential to broadly augment immune activation triggered by vaccine antigens and a variety of adjuvants. Their approach allows the antigen-presenting cell to determine the composition and kinetics of effector molecules and the co-stimulation needed for the development of potent adaptive immune responses. Before considering this approach for clinical use, it will be important to determine whether the immune response enhancement seen in mice can be achieved in higher-order animals such as nonhuman primates. The safety of this approach will need to be more fully defined in species other than mice. It will be important to show that SOCS inhibition is temporary, and that normal regulation of immune responses is maintained following the vaccination. This may require long-term evaluation in multiple species to assure there is not excessive or prolonged inflammation. Identifying pharmacological inhibitors of SOCS family members may provide a more temporally controlled approach for releasing the inhibition of TLR and cytokine signaling, and provide an additional safeguard against the theoretical risk of chronic inflammation. As the authors suggest, waking up the immune system by inhibiting the inhibitors may have therapeutic benefits in some settings. The struggle to develop a preventive vaccine for HIV-1 has taught us much of what we know about immunology, and will continue to have benefits at the conceptual level and for antiviral vaccine development in general. HIV-1 vaccine development will be facilitated by innovative approaches to vaccine formulation and delivery such as SOCS “silencing” that improve the kinetics, the magnitude, and the composition of immune responses. However, ultimate success will depend on the design of vaccine antigens that can elicit the right antibody and T cell specificity to achieve virus neutralization and clearance. I thank Kathryn L. Bonaparte for her critical review and comments. Citation: Graham BS (2006) New approaches to vaccine adjuvants: Inhibiting the inhibitor. PLoS Med 3(1): e57. Abbreviations DCdendritic cell Envenvelope gpglycoprotein JAKJanus kinase LPSlipopolysaccharide NF-κ,Bnuclear factor-κB polyI: Cpolyinosinic:polycytidylic acid siRNAsmall interfering RNA SOCSsuppressor of cytokine signaling ThT helper cell TLRToll-like receptor ==== Refs References Sabin AB Improbability of effective vaccination against human immunodeficiency virus because of its intracellular transmission and rectal portal of entry Proc Natl Acad Sci U S A 1992 89 8852 8855 1528902 Smith DM Richman DD Little SJ HIV superinfection J Infect Dis 2005 192 438 444 15995957 Burton DR Desrosiers RC Doms RW Koff WC Kwong PD HIV vaccine design and the neutralizing antibody problem Nat Immunol 2004 5 233 236 14985706 Song XT Evel-Kabler K Rollins L Aldrich M Gao F An alternative and effective HIV vaccination approach based on inhibition of natural immune inhibitors in dendritic cells PLoS Med 2006 3 e11 10.1371/journal.pmed.0030011 16381597 Medzhitov R Preston-Hurlburt P Janeway CA A human homologue of the Drosophila Toll protein signals activation of adaptive immunity Nature 1997 388 394 397 9237759 Afonso LC Scharton TM Vieira LQ Wysocka M Trinchieri G The adjuvant effect of interleukin-12 in a vaccine against Leishmania major Science 1994 263 235 237 7904381 Weeratna RD Makinen SR McCluskie MJ Davis HL TLR agonists as vaccine adjuvants: Comparison of CpG ODN and Resiquimod (R-848) Vaccine 2005 23 5263 5270 16081189 Alexander WS Hilton DJ The role of suppressors of cytokine signaling (SOCS) proteins in regulation of the immune response Annu Rev Immunol 2004 22 503 529 15032587 Tang YW Graham BS Anti-IL-4 treatment at immunization modulates cytokine expression, reduces illness, and increases cytotoxic T lymphocyte activity in mice challenged with respiratory syncytial virus J Clin Invest 1994 94 1953 1958 7962541	0	PMC1323500	CC BY	2021-01-05 10:40:32	no	PLoS Med. 2006 Jan 3; 3(1):e58	latin-1	PLoS Med	2,006	10.1371/journal.pmed.0030058	oa_comm
==== Front PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1638159710.1371/journal.pmed.0030011Research ArticleImmunologyHIV/AIDSHIV Infection/AIDSImmunology and allergyAn Alternative and Effective HIV Vaccination Approach Based on Inhibition of Antigen Presentation Attenuators in Dendritic Cells HIV Vaccine/Immune Inhibitor SilencingSong Xiao-Tong 1 2 Evel-Kabler Kevin 1 3 Rollins Lisa 1 3 Aldrich Melissa 1 3 Gao Feng 4 Huang Xue F 1 5 Chen Si-Yi 1 2 3 1Center for Cell and Gene Therapy, Baylor College of Medicine, Houston, Texas, United States of America2Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas, United States of America3Department of Immunology, Baylor College of Medicine, Houston, Texas, United States of America4Duke Human Vaccine Institute, Duke University Medical Center, Durham, North Carolina, United States of America5 Department of Pediatrics, Baylor College of Medicine, Houston, Texas, United States of AmericaLane Clifford Academic EditorNational Institutes of HealthUnited States of America To whom correspondence should be addressed. E-mail: [email protected] Competing Interests: The authors have declared that no competing interests exist. Author Contributions: XTS, KEK, and SYC designed the study and analyzed the data. XTS, KEK, LR, MA, FG, and XFH carried out experiments. SYC wrote the paper. 1 2006 3 1 2006 3 1 e116 5 2005 10 10 2005 Copyright: © 2006 Song et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. New Approaches to Vaccine Adjuvants: Inhibiting the Inhibitor Taking an Alternative Approach to HIV Vaccination Background Current efforts to develop HIV vaccines that seek to stimulate immune responses have been disappointing, underscoring the inability of natural immune responses to control HIV-1 infection. Here we tested an alternative strategy to induce anti-HIV immune responses by inhibiting a host's natural immune inhibitor. Methods and Findings We used small interfering RNA (siRNA) to inhibit suppressor of cytokine signaling (SOCS) 1, a key negative regulator of the JAK/STAT pathway, and investigated the effect of this silencing on the ability of dendritic cells (DCs) to induce anti-HIV-1 immunity. We found that SOCS1-silenced DCs broadly induced enhanced HIV-1 envelope (Env)-specific CD8+ cytotoxic T lymphocytes and CD4+ T helper cells, as well as antibody responses, in mice. Importantly, SOCS1-silenced DCs were more resistant to HIV Env-mediated suppression and were capable of inducing memory HIV Env-specific antibody and T cell responses. SOCS1-restricted signaling, as well as production of proinflammatory cytokines such as interleukin-12 by DCs, play a critical role in regulating the anti-HIV immune response. Furthermore, the potency of HIV DNA vaccination is significantly enhanced by coimmunization with SOCS1 siRNA expressor DNA. Conclusions This study demonstrates that SOCS1 functions as an antigen presentation attenuator to control both HIV-1-specific humoral and cellular responses. This study represents the first, to our knowledge, attempt to elicit HIV-specific T cell and antibody responses by inhibiting a host's antigen presentation attenuator, which may open a new and alternative avenue to develop effective therapeutic and prophylactic HIV vaccines. Si-Yi Chen and colleagues describe a possible alternative strategy for HIV vaccination in mice involving suppressor of cytokine signaling (SOCS) in dendritic cells. ==== Body Introduction Despite extensive efforts, no effective human immunodeficiency virus (HIV) vaccine has emerged or is on the horizon [1,2]. Increasing evidence indicates that the host's natural immunity has a major, albeit usually insufficient, role in limiting HIV-1 infection. CD8+ cytotoxic T cells (CTLs) are the main mediators of viral control, as demonstrated by the dramatic increase in viremia in animal models after depletion of CD8+ T cells [3,4]. Although definitive evidence for a protective role of antibodies is lacking, a number of monoclonal antibodies generated from infected individuals have broadly neutralizing activities against primary HIV-1 isolates [5]. Antibodies to the HIV-envelope (Env) protein, gp120, protect animals such as monkeys and SCID-peripheral blood lymphocyte mice from HIV or SIV infection [6,7]. Thus, there is a growing consensus that an effective HIV immunization approach should be capable of inducing vigorous protective CTL as well as antibody responses [5,8–11]. Dendritic cells (DCs), the most potent antigen-presenting cells (APCs), mediate innate and adaptive immunity against viral infection by providing proinflammatory cytokines and by processing and presenting antigens to T cells [12]. DCs use Toll-like receptors (TLRs) to recognize conserved microbial structures such as lipopolysaccharide (LPS). TLR signaling promotes DC maturation by activating mitogen-activated protein kinase and nuclear factor-κB (NF-κB), which then mediate the expression of various cytokines, resulting in the induction of innate and adaptive immunity [13,14]. Hence, exploiting the full immunostimulatory potential of DCs is likely the key to achieving an effective immune response to prevent or control HIV infection. Suppressor of cytokine signaling (SOCS) 1 is a key negative regulator of signaling of cytokines, such as interferon (IFN)-γ, interleukin (IL)-2, IL-6, IL-7, IL-12, and IL-15, through the inhibition of the Janus kinases (JAKs) in many lineages of immune cells [15,16]. Although the SOCS family includes eight members, each member plays a unique role in attenuating cellular signaling [15,16]. SOCS1 binds to the JAK activation loop as a pseudosubstrate inhibitor through its Src-homology 2 domain and targets JAK2 for degradation, leading to the inhibition of the molecule, signal transducer and activator of transcription (STAT) [15,16]. SOCS1-deficient (SOCS1−/−) mice die neonatally, with severe systemic inflammation and aberrant T cell activation, mainly as a result of unbridled cytokine signaling [17–19]. Such SOCS1−/− mice are also hypersensitive to LPS, and SOCS1−/− DCs show a more mature phenotype than do wild-type (WT) DCs and induce autoreactive antibody production [20,21], suggesting that SOCS1 plays a role in regulating DC functions by inhibiting the JAK/STAT pathway and the TLR signaling pathway directly or indirectly [15,22,23]. We have previously shown that SOCS1 plays a critical role in regulating the antigen presentation by DCs, and DCs in which SOCS1 expression is silenced by small interfering RNA (siRNA) induce enhanced CTL responses against tumor-associated antigens [24]. In agreement with this result, Hanada et al. recently reported that immunization with SOCS1−/− DCs derived from SOCS1 genetic knockout mice induced a hyper CD4+ T helper cell (Th1)-type immune response and antitumor activities [25]. Many attempts to develop HIV vaccines have sought to stimulate immune responses by manipulating HIV antigens and delivery systems and by using various adjuvants [5,8–10]. These efforts have been disappointing and have underscored the inability of natural immune responses to control HIV-1 infection in most infected or immunized individuals. We therefore tested an alternative hypothesis that anti-HIV immune responses can be enhanced by silencing the host's natural immune inhibitors. In this study, we used siRNA to silence SOCS1 and investigated the effect of this silencing on the ability of DCs to induce anti-HIV-1 antibody and T cell responses in mice. Methods Cytokine and Antibody ELISAs Cytokine levels in cell culture supernatant were quantified by ELISA analysis (BD Biosciences, San Diego, California, United States), according to the manufacturer's instructions. To determine gp120-specific antibody and subclass titers, we coated gp120 proteins (5 μg/ml in carbonate buffer [pH 9.6]) overnight at 4 °C, adding 12-fold serial dilutions of sera in PBS-5% FBS to the wells for 1 h at room temperature. After eight washes, biotinylated anti-mouse antibodies (anti-mouse IgM, IgG, IgG1, IgG2a, IgG2b, or IgG3) were added to the wells for 1 h at room temperature. Streptavidin-HRP was used as a peroxidase substrate. The reaction was stopped by addition of 50 μl of 2 M H2SO4. Optical densities were read at 450 nm on a BioAssay Reader (PerkinElmer, Wellesley, California, United States). The results are expressed as reciprocal endpoint titers, determined from a scatter plot with OD values on the y-axis and dilution-1 on the x-axis, for which the x-axis scale was logarithmic. After the data were plotted, a logarithmic curve fit was applied to each individual dilution series, and the point where the curve fit intersects the positive-negative cutoff value was determined. The cutoff value was calculated for each antibody isotype as the mean (± 3 standard deviations) of all dilutions from control mouse sera. All samples tested in each experiment were assayed at the same time. Transduction of Bone Marrow-Derived DCs with Lentiviral or Adenoviral Vectors Recombinant lentiviral vectors, LV-SOCS1-siRNA and LV-GFP-siRNA, were produced and titrated, as described previously [24]. A recombinant adenoviral vector (Ad-IL-12) expressing a biological active mouse IL-12 (a fusion protein of p35 and p40) was purchased from InvivoGen (San Diego, California, United States) and produced according to the manufacturer's instruction. Adenoviruses were titrated using Adeno-X Rapid Titer Kits (BD Bioscience). Mouse bone marrow (BM)-derived DCs were generated by culturing with granulocyte-macrophage colony-stimulating factor (GM-CSF) and IL-4, and transduced by lentiviral or adenoviral vectors as described previously [24]. T Cell Enzyme-Linked Immunospot Assays Enzyme-linked immunospot (ELISPOT) assays of isolated CD4+ or CD8+ T cells were performed as described in our previous reports [24]. Recombinant gp120 protein-pulsed BM-DCs were used for T cell stimulation. An irrelevant protein (ovalbumin [OVA]; Sigma, St. Louis, Missouri, United States) was also used as a negative control. CD4+ and CD8+ T cells were isolated from splenocytes with MACS CD4 (L3T4) or MACS CD8 (Ly-2) MicroBeads (Miltenyi Biotec, Auburn, California, United States). B Cell Isolation and gp120 Antibody-Producing B Cell ELISPOT Assay Single-cell suspensions prepared from spleens in complete RPMI 1640 medium were plated on plastic dishes for 1 h at 37 °C in 5% CO2 to remove adherent macrophages. Nonadherent cells were treated with anti-Thy1.2 and rabbit complement for 45 min at 37 °C to lyse T cells. The purity of the remaining B cells usually exceeded 90%. The B cell ELISPOT assay was performed by a modified method described before [26]. Briefly, 96-well nitrocellulose-base plates (Millipore Multiscreen PI, Billerica, Massachusetts, United States) were coated overnight with gp120 in PBS. The plates were washed six times with PBS and blocked with RPMI 1640 containing 10% FBS at 37 °C for 2 h. The isolated B cells were seeded into wells (5 × 105 cells/well) and incubated for 20 h at 37 °C in 5% CO2. The cells were then removed by six washes with PBS containing 0.5% Tween 20 (Sigma). Biotinylated anti-mouse IgG (BD Pharmingen, San Diego, California, United States), diluted in PBS containing 0.5% FBS to 1 μg/ml, was added, and the mixture incubated at 37 °C for 2 h. The avidin:biotinylated enzyme complex ([ABC]; Vector Laboratories, Burlingame, California, United States) was added for an additional hour. Anti-gp120 IgG was detected after a 4-min reaction with AEC (3-amino-9-ethylcarbazole; Sigma). The results were evaluated by ZellNet Consulting (New York, New York, United States) with an automated ELISPOT reader system (Carl Zeiss, Thornwood, New York, United States), using KS ELISPOT 4.3 software. Quantitative Real-Time PCR Analysis of BAFF and APRIL The relative expression of SOCS1 in transfected mouse BM-DCs was evaluated by quantitative real-time PCR. Total RNA was extracted from DCs, using Trizol reagent (Invitrogen, Carlsbad, California, United States), and 1.0 μg of total RNA for each sample was reverse transcribed with random hexamer primers and SuperScript First-Strand Synthesis kits (Invitrogen). Real-time 5′-nuclease fluorogenic PCR analysis was performed on an ABI 7900HT Sequence Detection System (Applied Biosystems, Foster City, California, United States) in 20-μl quadruplicate reactions with the equivalent of 5 ng of starting RNA material per reaction as template. The following primers were used for BAFF and APRIL: BAFF sense, 5′- TGCTATGGGTCATGTCATCCA-3′ and anti-sense, 5′- GGCAGTGTTTTGGGCATATTC-3′; APRIL sense, 5′- TCACAATGGGTCAGGTGGTATC-3′ and anti-sense, 5′- TGTAAATGAAAGACACCTGCACTGT-3′. TaqMan probe, forward and reverse primer for 18S were obtained from TaqMan Rodent 18S control reagents (Applied Biosystems). The PCR parameters were these recommended for the TaqMan Universal PCR Master Mix kit (Applied Biosystems), with BAFF, APRIL, and 18S reactions performed in separate tubes. BAFF and APRIL levels were normalized to 18S rRNA, while BAFF or APRIL expression (relative to the control value of mock-transfected, stimulated DCs) was calculated by the Comparative Ct method [24,27]. CTL Assays CD8+ CTL responses were assessed with a standard chromium release assay [24] that measures the ability of in vitro-restimulated splenocytes to lyse target cells. Splenocytes pooled from immunized mice were restimulated in vitro in RPMI-1640 containing gp120 proteins (20 μg/ml) for 4–6 d. Target cells pulsed with 20 μg/ml of gp120 protein overnight were labeled with 51Cr-sodium chromate solution for 90 min. Different numbers of effector cells were incubated with a constant number of target cells (1 × 104/well) in 96-well V-bottom plates (200 μl/well) for 3 h at 37 °C. The supernatants (100 μl) from triplicate cultures were collected. Percent cell lysis was calculated as (experimental release – spontaneous release)/(maximum release – spontaneous release) × 100. T and B Cell Proliferation Assay CD4+ or CD8+ T cells (1 × 106 per well) and B cells (1 × 105 per well) isolated as described above were cultured in complete medium in triplicate wells of 96-well plates with or without various stimuli. On the fourth day of culture, wells were pulsed with 1 μCi of [3H]-thymidine for 16 h. Plates were then harvested, and incorporated [3H]-thymidine was measured using a MicroBeta scintillation counter (TopCount NXT, Packard, Meriden, Connecticut, United States). DC Immunization The recombinant soluble gp120 (SF162) protein, with a purity of over 95% and mostly in a monomeric form, was produced and purified from CHO cells and kindly provided by the National Institutes of Health AIDS Research and Reference Program and Chiron Corporation. On day 5 of BM culture, DCs derived from BM of WT mice or IL-12 receptor knockout (KO) mice were transduced with LV-SOCS1-siRNA or LV-GFP-siRNA at a multiplicity of infection (MOI) of 5 [24], and pulsed with proteins for 2 h. The transduced DCs were then stimulated with LPS (100 ng/ml, Sigma) or tumor necrosis factor α (TNFα) (50 ng/ml, Peprotech, Rocky Hill, New Jersey, United States) ex vivo for 24 h, washed with PBS, and injected into mice (Jackson Laboratory, Bar Harbor, Maine, United States) via a foot pad. IL-12 receptor KO mice (B6.129S1-Il12rb1tm1Jm) in a C57BL/6 background were purchased from the Jackson Laboratory. The immunized mice were treated with LPS, PolyI:C, or R837 (30 μg/mouse) or murine IL-12 (1 μg/mouse) intraperitoneally three times on days 1, 3, and 5 after each DC immunization. DNA Vaccination The pSuper-SOCS1-siRNA expression vector was generated as described previously [24]. An HIV Env gp140 expression vector (pCMV/R-gp140CF) in which the gp120/gp41 cleavage site and fusion domain of HIV gp160 (codon usage-optimized HIV-1 strain JRFL) was deleted and the gp140CF gene was placed under control of the CMV promoter, was constructed. Endotoxin-free DNA was prepared with a DNA isolation kit from Qiagen (Valencia, California, United States), resuspended in endotoxin-free PBS (Sigma-Aldrich) at a final concentration of 1 μg/μl, and stored at −20 °C until used for injection. On the scheduled day of vaccination, 50 μg of gp140CF DNA or 200 μg of the mixture of gp140CF DNA (50 μg) and pSuper-SOCS1-siRNA expressor DNA (150 μg) [24] was injected into the quadriceps of each mouse [28,29]. The immunized mice were then treated with PolyI:C or R837 (30 μg/mouse) intraperitoneally three times on days 1, 3, and 5 after each DNA immunization [24]. Statistical Analysis We used the Student's t-test, and 95% confidence limits, to assess results for statistical significance, defined as P < 0.05. Results are typically presented as means ± standard error. Results Silencing of SOCS1 in DCs Enhances the HIV Env-Specific Antibody Response We first investigated the effect of SOCS1 silencing on the ability of DCs to induce anti-HIV antibody responses. We used HIV Env for this study, since it can induce both cellular and neutralizing antibody responses. A recombinant lentiviral vector (LV-SOCS1-siRNA) that expresses SOCS1 siRNA and has the ability to down-regulate about 90% of SOCS1 mRNA in transfected cells and a control vector (LV-GFP-siRNA) were generated, as described previously [24]. Mouse BM-derived DCs were transduced with LV-SOCS1-siRNA or LV-GFP-siRNA, loaded with recombinant HIV gp120 proteins, and matured with TNFα ex vivo. Groups of mice were then immunized with the transduced DCs twice at a weekly interval. LV-SOCS1-siRNA-DCs elicited greater gp120-specific IgG responses than did the control LV-GFP-siRNA-DCs (Figure S1A and S1B). We further tested whether in vivo stimulation with a TLR agonist, PolyI:C or R837, could further enhance anti-gp120 immune responses, since SOCS1 is an inducible feedback inhibitor [15,16] and immune responses against tumor-associated antigens induced by SOCS1-silenced DCs were preferentially enhanced by in vivo stimulation with LPS in our previous study [24]. Groups of mice were then immunized with the transduced DCs twice at a weekly interval, followed by stimulation with a low dose of PolyI:C or R837 in vivo after each DC immunization. Figure 1 shows increases in HIV Env-specific antibody titers in all IgG subclasses in mice immunized with LV-SOCS1-siRNA-DCs, compared with the corresponding IgG subclasses in LV-GFP-siRNA-DC mice. In vivo stimulation with PolyI:C or R837 preferentially enhanced HIV Env-specific antibodies, especially IgG2 and IgG3, in mice immunized with LV-SOCS1-siRNA-DCs (Figure 1). The Env-specific antibody subclass profile showed a Th1-polarized IgG response, higher IgG2a (a subclass associated with a Th1 response [30]), induced by LV-SOCS1-siRNA-DCs. Similar results were obtained in repeated experiments. In addition, in vivo stimulation with LPS also significantly enhanced HIV Env-specific antibody titers in LV-SOCS1-siRNA-DC mice (Figure S2). We did not perform neutralizing assays, since mice are not an appropriate species for reliable testing of HIV neutralizing activities [5]. We further found that SOCS1 silencing enhanced antibody responses to other strains of HIV Env proteins and antigens such as OVA (unpublished data). These results demonstrate that HIV Env-specific antibody responses are enhanced by the silencing of SOCS1 in DCs, implying a critical role for SOCS1 in DCs in controlling antigen-specific antibody responses. Figure 1 Enhanced gp120-Specific Antibody and T Cell Responses Induced by SOCS1-Silenced DCs Groups of C57BL/6 mice were immunized with gp120 (SF162) protein-pulsed, transduced BM-derived DCs (1 × 106 cells/mouse) twice at a weekly interval, followed by PolyI:C or R837 stimulation (30 μg/mouse) in vivo three times on days 1, 3, and 5 after each DC immunization, and sera and splenocytes were collected from each group of mice 14 d later. HIV gp120-specific IgG subclass titers (A) from the pooled sera of each group (4–6 mice/group) were quantified by capture ELISA. CD8+ T cells (B) and CD4+ T cells (C) isolated from pooled splenocytes were used for IFN-γ ELISPOT assays stimulated with gp120 proteins. Intracellular IFN-γ staining of CD8+ T cells from the pooled splenocytes were also performed (D). Representative data from one of three experiments are presented. NS, no stimulation. P < 0.01, LV-SOCS1-siRNA-DCs versus LV-GFP-siRNA-DCs. Silencing of SOCS1 in DCs Enhances HIV gp120-Specific T Cell Responses We next asked whether SOCS1 silencing could enhance HIV Env-specific CTL responses by using IFN-γ ELISPOT, intracellular cytokine staining, and CTL assays to test the functional status of CD8+ T cells in the immunized mice. CTL activities against gp120-pulsed target cells in the LV-SOCS1-siRNA-DC mice were more potent than those in the LV-GFP-siRNA-DC mice (see Figure S1). The CTL activity detected in these assays was gp120-specific, since splenocytes from LV-SOCS1-siRNA-DC mice lacked any apparent CTL activity against non-gp120-pulsed target cells (unpublished data). Natural killer cell (NK) activities were also enhanced in mice immunized with SOCS1-silenced DCs (Figure S3). In vivo stimulation with PolyI:C or R837 further enhanced the CD8+ T cell responses in LV-SOCS1-siRNA-DC-immunized mice (Figure 1). Intracellular staining of splenocytes with IFN-γ also showed higher percentages of IFN-γ+ CD8+ T cells in LV-SOCS-siRNA-DC mice (Figure 1). Various TLR agonists had a comparable ability to enhance the immunostimulatory potency of LV-SOCS1-siRNA-DCs (Figure S4). In addition, the percentage of perforin-positive CD8+ T cells was significantly increased in mice immunized with gp120-pulsed LV-SOCS1-siRNA-DCs (Figure S5), suggesting that SOCS1-silenced DCs may qualitatively enhance CTL responses as well. LV-SOCS1-siRNA-DC immunization induced enhanced HIV gp120-specific CD4+ T cells (Figure 1). We compared the potency of SOCS1-silenced DC immunization and protein adjuvant immunization. Figure 2A–2C show that gp120-pulsed SOCS1-silenced DCs induced potent CD8+ and CD4+ T cell responses as well as antibody responses, especially IgG2a, IgG2b, and IgG3. In contrast, immunization with the same amount of recombinant gp120 proteins formulated in incomplete Freund's adjuvant (IFA) only induced weaker antibody responses and barely detectable CD8+ and CD4+ T cell responses. Taken together, these results demonstrate a balanced and enhanced antibody and T cell response against HIV Env in mice immunized with SOCS1-silenced DCs, especially when stimulated in vivo with a low dose of TLR agonists, suggesting that SOCS1 in DCs critically regulates both anti-HIV humoral and cellular immunity. Figure 2 Comparison of gp120-Specific Antibody and T Cell Responses Induced by Protein Immunization and SOCS1-Silenced DCs Groups of C57BL/6 mice were immunized with gp120 protein (20 μg/ml)-pulsed, transduced BM-derived DCs (1 × 106 cells/mouse) or the same amount of gp120 protein formulated in IFA (20 μg/mouse) twice at a weekly interval. All of the mice were injected with PolyI:C or R837 (30 μg/mouse) in vivo three times on days 1, 3, and 5 after each immunization, and sera and splenocytes were collected from each group of mice 14 d later. HIV gp120-specific IgG subclass titers (A) from the pooled sera of each group (4–6 mice/group) were quantified by capture ELISA. CD8+ T cells (B) and CD4+ T cells (C) isolated from pooled splenocytes were used for IFN-γ ELISPOT assays stimulated with gp120 proteins. Representative data from one of three experiments are presented. P < 0.01, gp120 protein + IFA versus gp120-pulsed LV-SOCS1-siRNA-DCs. Enhanced gp120-Specific Th Response Induced by SOCS1-Silenced DCs Given the role of cytokines in programming Th1 versus Th2 responses [31,32], we reasoned that SOCS1 silencing might affect CTL and antibody responses by regulating the production of cytokines by DCs. Figure 3A and 3B shows significantly increased levels of IL-12 (p70), IFN-γ, and TNFα, which promote Th1-polarized responses, produced by LV-SOCS1-siRNA-BM-DCs generated in the culture containing GM-CSF and IL-4, compared with GFP-siRNA-DCs generated with GM-CSF and IL-4 culture after stimulation with LPS or LPS with anti-CD40 antibodies. The expression of EOMES mRNA, a transcription factor involved in the regulation of IFN-γ, was enhanced in LV-SOCS1-siRNA-DCs compared with LV-GFP-siRNA-DCs after stimulation with LPS (Figure S6), in agreement with a recent report by Hanada et al [25]. In addition, significant increases of IL-4, IL-6, and IL-10, which promote Th2-polarized responses, were also seen in the SOCS1-silenced BM-DCs (P < 0.01). Interestingly, we found that both LV-SOCS1-siRNA-DCs and LV-GFP-siRNA-DCs generated in a culture containing GM-CSF alone only produced low levels of IL-12 (p70), even when stimulated with LPS and anti-CD40 antibodies. These results suggest that IL-4 is a potent enhancer of IL-12 production, which is supported by an earlier finding [33]. Collectively, these data suggest that the higher levels of both Th1- and Th2-promoting cytokines produced by SOCS1-silenced DCs may account for the enhanced ability of SOCS1-silenced DCs to induce both HIV Env-specific CTL and antibody responses. Figure 3 Enhanced Production of Both Th1- and Th2-Polarizing Cytokines by SOCS1-Silenced DCs and Activated CD4+ Th (A) Enhanced production of both Th1- and Th2-polarizing cytokines by SOCS1-silenced DCs. BM-DCs transfected with SOCS1 siRNA or control [24] were stimulated with LPS (100 ng/ml). Concentrations of various cytokines in the culture media were analyzed by ELISA 24 h after stimulation. NS, no stimulation. P < 0.01, LV-SOCS1-siRNA-DCs versus LV-GFP-siRNA-DCs. (B) IL-12 production by transduced DCs. BM cells were cultured with mGM-CSF (20 μg/ml only or mGM-CSF and mIL-4 (20 μg/ml) [24] for 6 d and then transduced with LV-SOCS1-siRNA or LV-GFP-siRNA. The transduced DCs (5 × 105/ml) were then stimulated with LPS (100 ng/ml) and plate-coated anti-CD40 mAb (5 μg/ml, BD Bioscience) in the presence or absence of IL-4. Concentrations of IL-12 (p70) in the culture media were analyzed by ELISA 24 h after stimulation and are presented from one of three independent experiments. P < 0.01, LV-SOCS1-siRNA (GM-CSF + IL-4) versus LV-GFP-siRNA (GM-CSF + IL-4). (C–E) CD4+ Th responses induced by LV-SOCS1-siRNA-DCs. CD4+ T cells were isolated from pooled splenocytes of different groups of mice and subjected to the following assays. Numbers of IFN-γ-producing CD4+ T cell precursors were determined with the ELISPOT assay (C). 3H-thymidine incorporation rates of the isolated CD4+ T cells were determined on the fourth day of restimulation with gp120-pulsed DCs (D). Cytokine levels in the culture medium of isolated CD4+ cells stimulated with gp120-pulsed DCs for 48 h were determined by ELISA (E). The mean results (+ standard error) from one of three experiments are presented. P < 0.01, LV-SOCS1-siRNA-DC versus LV-GFP-siRNA-DC mice. SOCS1 silencing in DCs clearly promoted antibody and CTL responses, but it was unclear whether HIV Env-specific CD4+ Th responses, which are intimately involved in the induction of antibody and CTL responses, are also enhanced by SOCS1 silencing. We therefore isolated CD4+ T cells from immunized mice using CD4+ microbeads and analyzed them with various assays. As depicted in Figure 3C, the frequencies of gp120-specific CD4+ T cells were significantly higher in LV-SOCS1-siRNA-DC mice than in LV-GFP-siRNA-DC mice. 3H-thymidine incorporation assays showed that the CD4+ T cells from LV-SOCS1-siRNA-DC mice proliferated more actively than those from LV-GFP-siRNA-DC mice in response to stimulation with gp120-pulsed DCs (Figure 3D). Analysis of the cytokine profiles produced by CD4+ T cells isolated from LV-SOCS1-siRNA-DC mice after stimulation with gp120-pulsed DCs revealed increased levels of both Th1-polarizing (IFN-γ, IL-12, and TNFα) and Th2-polarizing (IL-4 and IL-10) cytokines (Figure 3E). These results indicate that SOCS1-silenced DCs induce an enhanced Th1-polarized, but a mixed Th1 and Th2, response against HIV Env, which is consistent with the Th1-polarized gp120-specific IgG subclass profile (higher IgG2a) shown in Figures 1B and 2B. Enhanced gp120-Specific B Cell Activation by SOCS1-Silenced DCs DCs have been shown to directly trigger B cell proliferation, maturation, and class-switch recombination by producing APRIL (a proliferation-inducing ligand) and BAFF (B-cell activating factor of the TNF family, also known as BLyS), members of the TNF superfamily [34–36]. We examined the effect of SOCS1 silencing on the production of APRIL and BAFF by DCs using real-time RT-PCR. LV-SOCS1-siRNA-DCs expressed higher levels of APRIL and BAFF mRNA upon LPS stimulation than did LV-GFP-siRNA-DCs (Figure 4A), in agreement with the increased expression of BAFF and APRIL in SOCS1−/− DCs [20]. Figure 4 Enhanced HIV-Specific B Cell Responses (A) Enhanced production of BAFF and APRIL by SOCS1-silenced DCs. The transduced BM-DCs were stimulated with LPS (100 ng/ml) for 24 h. Relative expression levels of BAFF and APRIL mRNA were then determined by real-time quantitative PCR as described in Methods, and normalized to mock-transfected DCs after LPS stimulation using the Comparative Ct method [27]. Representative data from two independent experiments are presented. P < 0.01, LV-SOCS1-siRNA-DC versus LV-GFP-siRNA-DCs. (B–D) Enhanced activation of gp120-specific B cells by SOCS1-silenced DCs. Frequencies of anti-gp120 antibody-producing cells in different groups of mice were determined and reported as the number of cells secreting gp120-specific IgG per 5 × 105 B cells (B). The proliferation rates (C) and cytokine production (D) of B cells (5 × 104/well) isolated from the spleens of different groups of mice after stimulation with anti-CD40 (5 μg/ml), IL-4 (20 ng/ml), or costimulation with anti-CD40 and IL-4 for 48 h were determined, and results from one of three independent experiments are presented. P < 0.01, LV-SOCS1-siRNA-DC mice versus LV-GFP-siRNA-DC mice under various in vitro stimulation conditions. To test the ability of SOCS1-silenced DCs to enhance activation of gp120-specific B cells, we used an anti-gp120 IgG-specific B cell ELISPOT assay to directly examine the frequencies of anti-gp120 IgG-producing B cells in the immunized mice. Frequencies of anti-gp120 IgG-producing B cells were significantly higher in LV-SOCS1-siRNA-DC mice than in LV-GFP-siRNA-DC mice (P < 0.01) (Figure 4B). Higher percentages of B cells exhibited an activated phenotype characterized by high levels of CD69, CD40, and CD86 in LV-SOCS1-siRNA-DC mice, compared with B cells from LV-GFP-siRNA-DC mice (unpublished data). We further purified B cells from the spleens of immunized mice and stimulated them with various stimuli. Figure 4C shows that B cells from LV-SOCS1-siRNA-DC mice proliferated more vigorously when costimulated with anti-CD40 and IL-4 than did B cells from LV-GFP-siRNA-DC mice. Interestingly, B cells from LV-SOCS1-siRNA-DC mice, but not those from LV-GFP-siRNA-DC mice, responded strongly to IL-4 or anti-CD40 only, suggesting that increased numbers of B cells were already activated in vivo by immunization with LV-SOCS1-siRNA-DCs. We also found that B cells from LV-SOCS1-siRNA-DCs mice produced higher levels of various cytokines, including IL-6, IL-2, and TNF-α, in response to various stimuli (Figure 4D). The enhanced B cell activation and antibody production induced by SOCS1-silenced DCs were likely CD4+ T cell-dependent, since the antibody production was compromised in CD4 knockout mice immunized with SOCS1-silenced DCs (unpublished data). Collectively, these results suggest that SOCS1-silenced DCs produce enhanced levels of B lymphocyte stimulators (BAFF and APRIL) and Th2-polarizing cytokines, leading to more effective activation of HIV Env-specific B cells and Th cells. Long-Term HIV Env-Specific CTL and Antibody Responses Induced by SOCS1-Silenced DCs Having shown that SOCS1 silencing in DCs enhances the primary HIV Env-specific CTL and antibody responses, we tested whether SOCS1-silenced DCs would induce memory HIV-specific CTL and antibody responses. Figure 5A shows that mice immunized with LV-GFP-siRNA-DCs had very low levels of gp120-specific antibodies at 6 mo after immunization, while LV-SOCS1-siRNA-DC mice still retained significant titers of gp120-specific IgG1 and IgG2 antibodies in their sera. At 1 wk after booster immunization, the LV-SOCS1-siRNA-DC mice showed strong recall antibody responses, with a mean titer of anti-gp120 IgG1 at 2 × 105 and anti-gp120 IgG2 at 1 × 105, while the LV-GFP-siRNA-DC mice showed poor recall antibody responses, with a mean titer of IgG1 at 3 × 103 and IgG2 at 4 × 102. These data show that SOCS1-silenced DCs exhibit about 64- and 255-fold increases in the titers of IgG1 and IgG2a antibodies, respectively, compared to LV-GFP-siRNA-DCs. Figure 5 Long-Term gp120-Specific Antibody and CTL Responses Induced by SOCS1-Silenced DCs IgG subclass titers (A) from pooled sera of different groups of mice and frequencies of IFN-γ-positive T cells of CD8+ T cells (B) and CD4+ T cells (D) isolated from pooled splenocytes (two mice per group) were determined at 6 mo after DC immunization and on day 7 after booster immunization with recombinant gp120 emulsified in IFA (20 μg protein/mouse). Intracellular IFN-γ and surface CD44 costaining of gated CD8+ T cells from splenocytes at 6 mo after immunization are shown (C). The data are representative of two experiments. The maintenance of memory HIV-specific CTLs and Th was assessed by examining gp120-specific CD8+ and CD4+ T cell responses with IFN-γ ELISPOT assays. Figure 5B shows that strong gp120-specific CTL responses were detected in LV-SOCS1-siRNA-DC mice, but not in LV-GFP-siRNA-DC mice, at 6 mo after immunization (249 IFN-γ spots per 5 × 105 CD8+ T cells in LV-SOCS1-siRNA-DC mice versus three IFN-γ spots in LV-GFP-siRNA-DC mice). Vigorous gp120-specific CTL responses were rapidly induced by booster immunization in LV-SOCS1-siRNA-DC mice, but not in LV-GFP-siRNA-DC mice (446 IFN-γ spots per 5 × 105 CD8+ T cells in LV-SOCS1-siRNA-DC mice versus 16 IFN-γ spots in LV-GFP-siRNA-DC mice on day 7 post-boosting) (Figure 5B). Costaining of intracellular IFN-γ and the surface CD44 memory marker of CD8+ T cells also showed a higher percentage of CD44hi and IFN-γ+ CD8+ T cells in LV-SOCS1-siRNA-DC mice, compared with LV-GFP siRNA-DC mice at 6 mo post-immunization (Figure 5C). Similarly, gp120-specific CD4+ Th responses were maintained and rapidly induced in LV-SOCS1-siRNA-DCs mice at 6 mo after immunization (391 IFN-γ spots per 5 × 105 CD4+ T-cells in LV-SOCS1-siRNA-DC mice versus 37 IFN-γ spots in LV-GFP-siRNA-DC mice on day 7 post-boosting) (Figure 5D). Thus, immunization with SOCS1-silenced DCs effectively induces long-term HIV Env-specific CTL, Th, and antibody responses. No apparent toxicity was observed in the mice immunized with LV-SOCS1-siRNA-DCs pulsed with gp120 up to 7 mo after immunization. Histological analysis of all major organs and tissues of the immunized mice revealed no pathologic inflammation (unpublished data). Levels of IgG and anti-dsDNA were comparable in LV-SOCS1-siRNA-DC and mock DC mice. These data suggest that gp120-pulsed LV-SOCS1-siRNA-DC immunization does not cause pathological inflammation in mice. Resistance of SOCS1-Silenced DCs to HIV Env-Mediated Immune Suppression The HIV Env protein gp120 can suppress the ability of DCs to produce proinflammatory cytokines and to stimulate T cells [37–40]. We therefore asked whether the enhanced activation of DCs by SOCS1 silencing might overcome the inhibitory effects of gp120 proteins on the cytokine production and immunostimulatory capacity of DCs. IL-12 was selected as a representative cytokine for these experiments, because DC-derived IL-12 was found to play a dual role, driving Th1 development as well as directly signaling B cells for developing humoral response [41–43]. As shown in Figure 6A, LV-SOCS1-siRNA-DCs in the presence of gp120 proteins retained the ability to respond to LPS. In contrast, the response of LV-GFP-siRNA-DCs to LPS stimulation was severely compromised by the presence of gp120 proteins. The susceptibility of SOCS1-silenced DCs to gp120-mediated suppression was further investigated in vivo. Mice were immunized with OVA-pulsed transduced DCs with or without pretreatment of gp120 proteins ex vivo. Pre-exposure to gp120 proteins did not have apparent effects on the ability of LV-SOCS1-siRNA-DCs to induce OVA-specific antibody responses (Figure 6B and 6C), nor did it compromise OVA-specific CD8+ CTL and CD4+ Th responses induced by LV-SOCS1-siRNA-DCs (P > 0.05) (Figure 6D and 6E). However, such pretreatment significantly reduced the ability of LV-GFP-siRNA-DCs to induce OVA-specific antibody and CTL responses (P < 0.05) (Figure 6B–6E). These results indicate that SOCS1 silencing renders DCs resistant to HIV gp120-mediated suppression, probably because of the enhanced cytokine production and hyperactivated state of SOCS1-silenced DCs [20,24]. Figure 6 Resistance of SOCS1-Silenced DCs to HIV gp120-Mediated Suppression (A) Effects of gp120 on cytokine production by DCs. BM-DCs were transfected with SOCS1-siRNA or a SOCS1-siRNA mutant oligonucleotide as described previously [24], and then cultured with or without SF162 gp120 (20 μg/ml) or LPS (100 ng/ml), and cytokine levels were determined at the different times of cultures, as indicated. (B–E) Effects of gp120 on DC antigen presentation in vivo. Transfected BM-DCs were pulsed with OVA, incubated with or without gp120 for 2 d, and then stimulated with LPS (100 ng/ml) ex vivo overnight. Mice were then immunized with the transduced DCs twice, following three in vivo LPS stimulations. OVA-specific antibody IgG (B) and IgG1 (C) titers and frequencies of IFN-γ-producing OVA-specific CD8+ T cells (D) and CD4+ T cells (E) were examined 2 wk after the second DC immunization. Data are representative of two repeats. Regulation of gp120-Specific Cellular Responses by SOCS1-Restricted IL-12 Signaling in DCs Since IL-12 is a potent stimulator of Th1 immune responses [44] and is regulated by SOCS1 [45], we further tested whether in vivo stimulation with IL-12, which may be applicable in the clinic, is also effective in enhancing the potency of LV-SOCS1-siRNA-DCs. HIV gp120-pulsed, transduced DCs that were matured ex vivo with TNFα were transferred into mice. The recipient mice were then stimulated in vivo three times with a low dose of recombinant mouse IL-12 (1 μg/mouse/injection). Figure 7A shows that gp120-specific CD8+ CTL activities in LV-SOCS1-siRNA-DC-immunized mice were significantly enhanced by in vivo IL-12 stimulation, as demonstrated by IFN-γ ELISPOT assay. In contrast, in vivo administration with IL-12 only had a modest effect on CTL activities in LV-GFP-siRNA-DC-immunized mice. Figure 7B also shows that in vivo IL-12 stimulation preferentially enhanced gp120-specific CD4+ Th responses induced by LV-SOCS1-siRNA-DCs. These results suggest that in vivo administration with a representative proinflammatory cytokine (IL-12) preferentially enhances the immunostimulatory ability of SOCS1-silenced DCs. Figure 7 The Role of IL-12 Signaling in Enhanced Anti-HIV Immunity (A and B) In vivo injection with IL-12 preferentially enhanced gp120-specific CTL and Th responses induced by SOCS1-silenced DCs. C57BL/6 mice were immunized with 1 × 106 of HIV gp120-pulsed (50 μg/ml), transduced DCs derived from BM of WT mice or IL-12 receptor KO mice with ex vivo TNFα maturation (50 ng/ml). On days 1, 3, and 5 after DC immunization, murine IL-12 (1 μg/mouse, Peprotech) was administered intraperitoneally. CD8+ T cells (A) or CD4+ T cells (B) isolated 2 wk later from the pooled splenocytes of immunized mice (2–3 each group) were subjected to IFN-γ ELISPOT assays. An irrelevant protein, OVA, was used as a negative control. Representative data from two independent experiments are presented. P < 0.01, LV-SOCS1-siRNA-DC versus LV-SOCS1-siRNA-DC + IL-12, or IL12R KO LV-SOCS1-siRNA-DC + IL-12 versus LV-SOCS1-siRNA-DC + IL-12. (C and D) gp120-specific CTL and Th responses induced by SOCS1-silenced DCs or Ad-IL-12-DCs. BM-derived DCs from WT mice were transfected with LV-SOCS1-siRNA (MOI of 5) or Ad-IL-12 with various MOIs of 10–1,000 or cotransfected with LV-SOCS1-siRNA (MOI of 5) and Ad-IL-12 (MOI of 10) for 4 h. DCs derived from BM of IL-12 receptor KO mice were cotransfected with LV-SOCS1-siRNA (MOI of 5) and Ad-IL-12 (MOI of 10) for 4 h. Groups of C57BL/6 mice were immunized with 1 × 106 of gp120-pulsed (50 μg/ml), transfected DCs with ex vivo TNFα maturation. CD8+ T-cells (C) or CD4+ T cells (D) isolated 2 wk later from the pooled splenocytes of immunized mice (2–3 each group) were subjected to IFN-γ ELISPOT assays. An irrelevant protein, OVA, was used as a negative control. Representative data from two independent experiments are presented. P < 0.01, Ad-IL-12/SOCS1-siRNA-DC versus IL-12-DCs, or Ad-IL-12/SOCS1-siRNA-DC versus IL12R KO Ad-IL-12/SOCS1-siRNA-DC. The preferential enhancement of the immunostimulatory ability of SOCS1-silenced DCs by a low dose of IL-12 in vivo suggests that SOCS1-restricted cytokine signaling in antigen-presenting DCs is critical in regulating antigen presentation and anti-HIV immunity. To further investigate this possibility, we compared the immunostimulatory ability of DCs transfected with a recombinant adenovirus expressing mouse IL-12 cytokine (Ad-IL-12) or with LV-SOCS1-siRNA. Mice were immunized with gp120-pulsed DCs transfected with various MOIs of Ad-IL-12 (10 to 1,000), DCs transfected with LV-SOCS1-siRNA (MOI of 5), or DCs cotransfected with LV-SOCS1-siRNA (MOI of 5) and Ad-IL-12 (MOI of 10). DCs transfected with Ad-IL-12 at a MOI of 300 constitutively produced a high level of IL-12, comparable to that produced by LV-SOCS1-siRNA-DCs after stimulation with LPS (unpublished data). Figure 7C and 7D show that gp120-specific CTL and Th responses were enhanced in mice immunized with DCs transfected with Ad-IL-12 (MOIs of 10 to 1,000), when compared with control LV-GFP-siRNA-DC. Interestingly, DCs transfected with either low or high MOIs of Ad-IL-12 induced comparable levels of gp120-specific CD8+ and CD4+ T cell responses (Figure 7C and 7D), suggesting that gp120-specific immune responses cannot be simply boosted by administration of increasing doses of proinflammatory cytokines. However, SOCS1-silenced DCs cotransfected with a low MOI (10) of Ad-IL-12 induced significantly more potent gp120-specific CD8+ CTL and CD4+ Th responses than SOCS1-silenced DCs or WT DCs transfected with either low or high MOIs (10 to 1,000) of Ad-IL-12 (Figure 7C and 7D), supporting a critical role of SOCS1-restricted IL-12 signaling in DCs for the induction of anti-HIV cellular responses. To further determine the role of autocrine signaling of IL-12 in antigen-presenting DCs, we compared the CTL and Th responses induced by DCs derived from IL-12 receptor KO mice or WT mice. These DCs were cotransfected with LV-SOCS1-sRNA and Ad-IL-12 and then injected into WT mice. Ad-IL-12/SOCS1-siRNA-DCs derived from IL-12 receptor KO mice exhibited a significantly reduced ability to induce gp120-specific CTL responses, compared with Ad-IL-12/SOCS1-siRNA DCs from WT mice (Figure 7C and 7D). Similarly, in vivo stimulation with IL-12 showed a significantly reduced ability to induce gp120-specific CTL and Th responses induced by IL-12R KO LV-SOCS1-siRNA-DCs, compared with WT LV-SOCS1-siRNA-DCs (Figure 7A and 7B). Taken together, these results indicate that SOCS1-restricted, autocrine signaling of proinflammatory cytokines such as IL-12 in DCs, in addition to the SOCS1-restricted production of proinflammatory cytokines by DCs, plays a critical role in controlling anti-HIV immune responses. Potency of HIV DNA Vaccine Enhanced by Coimmunization with SOCS1-siRNA DNA The ability of SOCS1-silenced DCs to enhance both HIV Env-specific CTL and antibody responses suggests that our SOCS1 silencing approach might be useful in improving the potency of HIV DNA vaccination. We generated a HIV gp140CF expression vector, in which the gp120/gp41 cleavage site and fusion domain of gp160 were deleted and the gp140CF gene was placed under control of the CMV promoter. To test the effect of SOCS1 siRNA on DNA vaccination, we injected mice with gp140CF DNA only or with a mixture of gp140CF DNA and pSuper-SOCS1-siRNA expressor DNA, which was constructed previously [24], weekly for 3 wk, followed by PolyI:C or R837 stimulation (30 μg/mouse) in vivo after each DNA immunization. Enhanced HIV Env-specific antibody titers were evident in mice coimmunized with pSuper-SOCS1-siRNA DNA (Figure 8). gp120-specific IgG2a antibodies, indicative of a Th1-polarized immune response [46], were preferentially enhanced, indicating that SOCS1-siRNA DNA preferentially enhances Th1-polarized anti-HIV immune responses induced by DNA vaccines. HIV Env-specific CTL responses were significantly enhanced by co-injection of pSuper-SOCS1-siRNA DNA, as demonstrated by ELISPOT assays (Figure 8). Intracellular IFN-γ staining also showed enhanced gp120-specific CD8+ T cell responses in mice coimmunized with pSuper-SOCS1 siRNA DNA (Figure 8). Moreover, HIV Env-specific CD4+ Th responses were enhanced by co-injection of SOCS1-siRNA DNA (Figure 8). These results indicate that pSuper-SOCS1 siRNA DNA coimmunization enhances the potency of HIV DNA vaccination, probably due to the enhanced immunostimulatory capacity of the cotransfected APCs in the immunized mice. Thus, our SOCS1 silencing strategy is applicable to ex vivo DC-based and in vivo vaccination settings. Figure 8 Enhancement of HIV DNA Vaccine by SOCS1-siRNA DNA Groups of C57BL/6 mice were immunized with gp140CF DNA [74] only or a mixture of gp140CF DNA and pSuper-SOCS1-siRNA expressor DNA [24] weekly for 3 wk, followed by PolyI:C or R837 stimulation (30 μg/mouse) in vivo three times on days 1, 3, and 5 after each DNA immunization. HIV gp120-specific IgG subclass titers (A), IFN-γ spot numbers of CD8+ T cells (B) and CD4+ T cells (C) from pooled splenocytes of different groups of mice (4–6 mice per group) 1 wk after the last immunization are shown from one of three independent experiments. Intracellular IFN-γ staining of CD8+ T cells from the pooled splenocytes was also performed (D). P < 0.01, gp140CF and GFP siRNA versus gp140CF and SOCS1 siRNA coimmunization. Discussion In this study we found that silencing of the negative signaling regulator SOCS1 in DCs results in drastic enhancement of both HIV Env-specific CTL and antibody responses in mice. We demonstrated that SOCS1-silenced DCs have an enhanced ability to generate memory HIV Env-specific T cell and B cell responses. We also found that SOCS1-restricted, autocrine signaling of proinflammatory cytokines, such as IL-12 in DCs, as well as production of proinflammatory cytokines by DCs, play a critical role in inducing anti-HIV immune response. In addition, we demonstrated that coimmunization with SOCS1 siRNA DNA significantly enhances the potency of HIV DNA vaccination. Thus, a balanced memory humoral and cellular response against HIV can be induced by SOCS1-silenced DCs and SOCS1-siRNA DNA. This study indicates that SOCS1 functions as a critical antigen presentation attenuator. Furthermore, this SOCS1 silencing strategy is broadly applicable to enhancing both therapeutic and prophylactic vaccines against HIV and other pathogens. The role of DCs in the induction of humoral responses has been traditionally viewed as a consequence of CD4+ Th priming for cognate interaction between T cells and B cells. However, the direct role of DCs in stimulation of the humoral response has been documented in vitro and in vivo [42,47]. Notably, DCs were found to strongly enhance both proliferation and antibody production of CD40-activated B cells [42]. Immunization with DCs loaded with antigens can induce a protective humoral response [48]. Here, we found that SOCS1-silenced DCs enhance the production of Th2-polarizing cyokines as well as B lymphocyte stimulatory cytokines (BAFF and APRIL), which is likely responsible for the enhanced Th and B cell activation seen in SOCS1-silenced DC-immunized mice. Our finding is supported by a previous report that SOCS1−/− DCs induce aberrant expansion of B cells and autoreactive antibody production [20]. Hence, this study demonstrates the critical role of SOCS1 in DCs in controlling HIV-specific antibody responses and implies that the silencing of SOCS1 can be generically used to boost antibody responses against antigens other than HIV Env. An important finding of this study is that SOCS1-silenced DCs induce balanced, memory HIV Env-specific antibody and CTL responses, which may be desirable for preventing or controlling HIV infection [5,8–11,49,50]. Although the mechanism(s) by which SOCS1 silencing induces a balanced, memory humoral and cellular response is unclear, it may involve the enhanced production of a mixed pattern of Th1- and Th2-polarizing cytokines by SOCS1-silenced DCs. These results are consistent with mixed antibody and CTL responses naturally generated against many pathogens such as viruses [30], indicating that Th1 and Th2 polarization is not mutually exclusive [32,51]. Hanada et al. recently found that the expression of Eomes, a transcription factor, was selectively overexpressed in SOCS1−/− BM-DCs, which may be responsible for the enhanced production of IFN-γ [25]. Baetz et al. [22] and Gingras et al. [23] recently reported that SOCS1 indirectly regulates TLR signaling in macrophages by inhibiting the signaling of type I IFN that is induced by TLR signaling. However, the Gingras et al. studies [15,22,23] showed that levels of IL-12 produced by SOCS1−/− BM-derived macrophages were comparable to those by WT BM-derived macrophages in response to LPS. We found that SOCS1-silenced BM-DCs produced higher levels of proinflammatory cytokines including IL-12 (p70 heterodimer) than did LV-GFP-siRNA-DCs (see Figure 2B). Our results are also consistent with two recent reports that SOCS1 KO macrophages produced excessive amounts of IL-12 and other cytokines in response to stimuli, and that enhanced levels of IL-12 and other cytokines were found in sera of conditional SOCS1 KO mice [52,53]. We noticed that transduction with control GFP-siRNA also enhanced the immunostimulatory potency of DCs, because siRNA molecules can stimulate DCs via the activation of TLR signaling [54,55] and the direct activation of some cellular genes such as the IFN-stimulated genes [56]. Thus, the enhanced immunostimulatory potency of DCs by SOCS1-siRNA is likely a collective result of SOCS1 silencing and nonspecific stimulatory effect of siRNA molecules. The nonspecific stimulatory ability of siRNA molecules may be an added benefit of using siRNA to silence a signaling inhibitor to enhance anti-HIV immunity. The results of this study underscore the importance of autocrine SOCS1-restricted signaling of proinflammatory cytokines such as IL-12 in DCs for the induction of anti-HIV immune responses. Proinflammatory cytokines have been used to enhance the potency of HIV vaccines [57,58]. Here we found that in vivo stimulation with IL-12 or cotransfection of recombinant Ad-IL-12 viruses only modestly enhance anti-HIV immune responses induced by WT DC immunization. In contrast, the anti-HIV immune responses induced by SOCS1-silenced DC immunization were significantly enhanced by in vivo stimulation with IL-12 or cotransfection of Ad-IL-12. These results indicate that the signaling of proinflammatory cytokines and their stimulatory effects on APCs are tightly restricted by SOCS1, which may reflect the modest or no enhancement of the potency of HIV vaccines by coexpression of various proinflammatory cytokines [59]. The importance of autocrine signaling of proinflammatory cytokines in antigen presentation was reported in several recent studies [60,61]. In addition, we found that SOCS1-silenced DCs have a superior ability to generate HIV-specific memory T cell and B cell responses. The increase of memory T cell and antibody responses was more profound than that of primary responses induced by SOCS1-silenced DCs. Although the molecular mechanisms for memory generation are still poorly defined, the fate of activated T lymphocytes is likely determined by the strength under which an antigen is presented by APCs [62–66]. The results of our study suggest that unrestricted autocrine signaling, as well as enhanced production of both Th1- and Th2-promoting cytokines by SOCS1-silenced DCs, may account for the enhanced ability of SOCS1-silenced DCs to induce memory HIV-specific CTL and antibody responses. It is tempting to postulate that enhanced expression and signaling of cytokines such as IL-7 and IL-15 and other surface molecules, which are involved in the generation of memory response [67–69], by SOCS1-silenced DCs may be responsible for the enhanced memory responses. Our results also suggest that the quality of gp120-specific T cells, in addition to their frequencies, may be enhanced by SOCS1-silenced DCs (Figure S6). It would be interesting to systematically examine the expression of granzyme, perforin, Fas ligand, and other molecules on these antigen-specific T cells induced by SOCS1-silenced DCs in a future study. Functional defects and depletion of DCs are common in HIV-infected individuals, likely contributing to the progressive immunodeficiency. Peripheral mononuclear cells from HIV-1-infected patients have been shown to produce significantly less IL-12 than those from uninfected controls [70]. HIV gp120 protein can suppress the ability of DCs to produce proinflammatory cytokines and to stimulate T cells [37,71], although the mechanism by which gp120 suppresses the production of proinflammatory cytokines is largely uncharacterized. We found that SOCS1-silenced DCs are more resistant to HIV gp120-mediated suppression, since SOCS1-silenced DCs in the presence of gp120 proteins still produced higher levels of IL-12 in response to LPS and induced stronger CTL and antibody responses (see Figure 6). This finding is especially relevant to the development of therapeutic HIV vaccines, which would be used in immunosuppressed HIV-infected individuals [72]. The enhanced resistance to gp120-mediated suppression may be due to the unbridled proinflammatory STAT (1/2/4) signaling in SOCS1-silenced DCs to antagonize anti-inflammatory signaling mediated by HIV gp120. In this study we treated DCs with high concentrations of soluble recombinant HIV gp120 proteins largely in a monomeric form, which may not reflect the physiologic conditions of HIV infection. Further studies are needed to investigate precise molecular mechanisms responsible for the enhanced resistance of SOCS1-silenced DCs to HIV suppression in a condition closely resembling natural HIV infection. The vaccination strategy described here, to our knowledge, represents the first effort to enhance anti-HIV immune responses by inhibiting the host's immune inhibitors in DCs. Since natural immunity is ineffective in controlling HIV-1 infection, disabling the host's immune inhibitors may be critical to generate effective anti-HIV immune responses. However, mere enhancement of HIV-specific immune responses may not lead to the induction of protective HIV antibodies and CTL responses. In this regard, our approach offers the opportunity for combinational immunization with currently available vaccines, as demonstrated by our coimmunization of DNA vaccine and SOCS1 siRNA DNA. When used with improved HIV immunogens and delivery systems [5,73], this vaccination approach may provide a new avenue to enhance weak protective immune responses or generate broader and stronger responses not only against dominant epitopes, but also against weakly immunogenic or cryptic, yet protective epitopes. Current efforts in HIV immunotherapy and vaccine development are largely aimed at stimulating anti-HIV immune responses by modifying HIV antigens and using various delivery systems and adjuvants. This study demonstrates the principle of disabling a signaling inhibitor in host DCs as an alternative and effective HIV immunization approach, which could be translated into the clinic. Indeed, our recent study identified a human SOCS1 siRNA and demonstrated that human SOCS1-silenced DCs also had enhanced immunostimulatory capacity to activate antigen-specific CTL responses (unpublished data). The SOCS1 siRNA molecule or expression cassette can be incorporated into various forms of prophylactic HIV vaccines, including DNA-, adenovirus-, vaccinia-, poxvirus-, virus-like particle-, and other vector-based vaccines. Furthermore, immunization of HIV-infected patients with DCs loaded with inactivated HIV viruses led to a reduction in viral loads and an increase in CD4+ T cell numbers in the blood [72], suggesting that SOCS1 silencing could augment the effect of therapeutic HIV vaccines for the long-term control of HIV infection. Thus, further investigations are warranted to determine if protective anti-HIV responses would be induced by this SOCS1 silencing strategy in monkeys and ultimately in humans. Supporting Information Figure S1 Enhanced gp120-Specific Antibody and CTL Responses Induced by SOCS1-Silenced DCs Groups of C57BL/6 mice were immunized with gp120 (SF162) protein-pulsed, transduced BM-derived DCs (1 × 106 cells/mouse) twice at a weekly interval without in vivo LPS stimulation, and sera and splenocytes were collected from each group of mice 14 d later. HIV gp120-specific IgG (A) and subclass (B) titers from the pooled sera of each group (4–6 mice/group) were quantified by capture ELISA. Antibody titers are reported as the mean ± standard deviation of endpoint titers [24]. Pooled splenocytes from the immunized mice were subjected to CTL assays against gp120 protein-pulsed syngeneic TC-1 cells (C). Representative data from one of three experiments are presented. P < 0.01, LV-SOCS1-siRNA-DCs versus LV-GFP-siRNA-DCs. (74 KB PDF). Click here for additional data file. Figure S2 gp120-Specific Antibody and CTL Responses Enhanced by SOCS1-Silenced DCs and In Vivo LPS Stimulation Groups of C57BL/6 mice were immunized with gp120 (SF162) protein-pulsed, transduced BM-derived DCs (1 × 106 cells/mouse) twice at a weekly interval, followed by LPS stimulation (30 μg/mouse) in vivo three times on days 1, 3, and 5 after each DC immunization, and sera and splenocytes were collected from each group of mice 14 d later. HIV gp120-specific IgM and IgG (A) and IgG subclass (B) titers from the pooled sera of each group (4–6 mice/group) were quantified by capture ELISA. Pooled splenocytes from the immunized mice were subjected to CTL assays against gp120 protein-pulsed syngeneic TC-1 cells (C). CD8+ T cells isolated from the splenocytes were used for IFN-γ ELISPOT assays stimulated with gp120 proteins or control BSA (D). Representative data from one of three experiments are presented. P < 0.01, LV-SOCS1-siRNA-DCs versus LV-GFP-siRNA-DCs. (72 KB PDF). Click here for additional data file. Figure S3 Enhanced NK Activities in Mice Immunized with SOCS1-Silenced DCs Splenocytes pooled from each group of mice immunized with gp120-pulsed BM-DCs (1 × 106 cells/mouse) were examined for NK activity using a 5-h 51Cr release assay against Yac-1 cells. Data are representative of three independent experiments. P < 0.05, LV-SOCS1-siRNA-DCs versus LV-GFP-siRNA-DCs. (50 KB PDF). Click here for additional data file. Figure S4 Comparison of T Cell Responses Boosted by Various TLR Agonists Groups of C57BL/6 mice were immunized with gp120 protein-pulsed, LV-SOCS1-siRNA-transduced DCs (1 × 106 cells/mouse) twice at a weekly interval, followed by in vivo stimulation with LPS, PolyI:C, or R837 (30 μg/mouse) three times on days 1, 3, and 5 after each DC immunization. CD8+ T cells isolated from the splenocytes were used for IFN-γ ELISPOT assays stimulated with gp120 proteins 14 d later. NS, no in vivo stimulation with any TLR agonist. *P < 0.01, NS versus LPS, PolyI:C, or R837. (48 KB PDF). Click here for additional data file. Figure S5 Enhanced Perforin Expression in T Cells of Mice Immunized with SOCS1-Silenced DCs Groups of C57BL/6 mice were immunized with gp120-pulsed, transduced BM-derived DCs (1 × 106 cells/mouse) or the same amount of gp120 protein formulated in IFA (20 μg/mouse) twice at a weekly interval. All of the mice were injected with PolyI:C (30 μg/mouse) in vivo three times on days 1, 3, and 5 after each immunization, and splenocytes were collected from each group of mice 14 d later. The splenocytes were in vitro restimulated with gp120 protein-pulsed BM-DCs for 5 h and then costained with anti-CD8-FITC and anti-Perforin-PE (BD Pharmingen) for FACS analysis. (66 KB PDF). Click here for additional data file. Figure S6 Enhanced Expression of Eomes in LV-SOCS1-siRNA-Transduced DCs The expression of Eomes and T-bet in the transduced DCs after 24 h of LPS stimulation was examined by RT-PCR, as described by Hanada et al. [25]. GAPDH was used as an internal control. A pair of primers used for T-bet amplification were 5′- CCCACAAGCCATTACAGG-3′ and 5′- AGTGATCTCTGCGTTCTGGT-3′; a pair of primers for Eomes amplification was 5′- TGAATGAACCTTCCAAGACTCAGA-3′ and 5′- GGCTTGAGGCAAAGTGTTGACA-3′; and a pair of primers for GAPDH amplification was 5′- ACCACAGTCCATGCCATCAC-3′ and 5′- TCCACCACCCTGTTGCTGTA-3′. (75 KB PDF). Click here for additional data file. Patient Summary Background When you are vaccinated against a virus, you are given a harmless form of the virus. A particular part of the virus—a part called an “antigen”—triggers your immune system to produce a response that includes antibodies (proteins secreted into the blood that bind just to that antigen) and specific cells. If one day you are infected with the actual virus, your body now has the means to fight it off. Despite a great deal of work going into the development of a vaccine against HIV, so far no vaccine has been produced. There are a number of possible reasons for this lack of success. For example, it has not been easy to identify the ideal part on the virus (the ideal antigen) that would trigger the immune system to produce a response. Also, we have not yet found a way to trigger a strong enough immune response against any experimental HIV vaccine. One group of cells involved in producing an immune response is known as antigen-presenting cells. These cells are responsible for handling antigens from viruses (and other microbes, such as bacteria) so that the body's immune system can respond efficiently and eliminate the microbe. The activity of these cells is controlled by molecular signals (cytokines) inside the cell. Why Was This Study Done? The authors wanted to take a different approach from the one described above to producing a vaccine. Our immune systems are kept in check by molecules called “immune inhibitors.” These prevent our immune systems from attacking our own bodies. These inhibitors might also be preventing our own immune systems from mounting an effective immune response to an HIV vaccine. The authors thought that by “switching off” or silencing these inhibitors, this might help the immune system to trigger a response against an HIV vaccine. They tested their hypothesis in a laboratory study of mice. They switched off a particular immune inhibitor called “SOCS1” inside a type of antigen-presenting cell called a dendritic cell, and looked at whether this would boost the mouse immune response in general, and whether it would lead to a more effective immune response against an HIV vaccine. What Did the Researchers Do and Find? They found that silencing SOCS1 in dendritic cells increased the immune response, and that this response was long-lasting. They also showed that when this strategy was combined with vaccination using genetic material (DNA) from HIV, the vaccine was more effective. What Do These Findings Mean? It seems that it may be possible to boost the body's response to HIV vaccination by switching off an immune inhibitor. However, these results are preliminary, as this approach has at the moment not been tested in humans, and before it is clear whether it will work in humans much research will need to be done, including assessing whether the approach is safe. Where Can I Get More Information Online? Medline Plus has many links to pages of information on HIV: http://www.nlm.nih.gov/medlineplus/aids.html The Body has a page of links on the search for an HIV vaccine: http://www.thebody.com/treat/vaccines.html We thank Wenhong Ren, Lei Shen, and Natasha Lapteva for technical assistance and valuable suggestions. We also thank the National Institutes of Health (NIH) AIDS Research and Reference Reagent Program, National Institute of Allergy and Infectious Diseases (NIAID) Division of AIDS, and Chiron Corporation for providing recombinant proteins and reagents; and Gary Nabel at the NIAID Vaccine Research Center, for providing pCMV/R vector. This work was supported by grants from the NIH (R01AI48480, R01AI48711, and R21 AI055386). KEK and MA were supported by a NIH training grant (T32-AI07495). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Citation: Song XT, Evel-Kabler K, Rollins L, Aldrich M, Gao F, et al. (2006) An alternative and effective HIV vaccination approach based on inhibition of antigen presentation attenuators in dendritic cells. PLoS Med 3(1): e11. Abbreviations APCantigen-presenting cell BMbone marrow CTLCD8+ cytotoxic T lymphocyte DCdendritic cell ELISPOTenzyme-linked immunospot Envenvelope GM-CSFgranulocyte-macrophage colony-stimulating factor HIVhuman immunodeficiency virus IFAincomplete Freund's adjuvant IFNinterferon ILinterleukin JAKJanus kinase KOknockout LPSlipopolysaccharide LVlentiviral vector MOImultiplicity of infection NKnatural killer cell OVAovalbumin siRNAsmall interfering RNA SOCSsuppressor of cytokine signaling STATsignal transducer and activator of transcription ThCD4+ T helper cell TLRToll-like receptor WTwild-type ==== Refs References Pantaleo G Koup RA Correlates of immune protection in HIV-1 infection: What we know, what we don't know, what we should know Nat Med 2004 10 806 810 15286782 Desrosiers RC Prospects for an AIDS vaccine Nat Med 2004 10 221 223 14991035 Schmitz JE Kuroda MJ Santra S Sasseville VG Simon MA Control of viremia in simian immunodeficiency virus infection by CD8+ lymphocytes Science 1999 283 857 860 9933172 Jin X Bauer DE Tuttleton SE Lewin S Gettie A Dramatic rise in plasma viremia after CD8+ T cell depletion in simian immunodeficiency virus-infected macaques J Exp Med 1999 189 991 998 10075982 Burton DR Desrosiers RC Doms RW Koff WC Kwong PD HIV vaccine design and the neutralizing antibody problem Nat Immunol 2004 5 233 236 14985706 Mascola JR Stiegler G VanCott TC Katinger H Carpenter CB Protection of macaques against vaginal transmission of a pathogenic HIV-1/SIV chimeric virus by passive infusion of neutralizing antibodies Nat Med 2000 6 207 210 10655111 Gauduin MC Parren PW Weir R Barbas CF Burton DR Passive immunization with a human monoclonal antibody protects hu-PBL-SCID mice against challenge by primary isolates of HIV-1 Nat Med 1997 3 1389 1393 9396610 McMichael AJ Hanke T HIV vaccines 1983–2003 Nat Med 2003 9 874 880 12835708 Nabel GJ Challenges and opportunities for development of an AIDS vaccine Nature 2001 410 1002 1007 11309631 Letvin NL Barouch DH Montefiori DC Prospects for vaccine protection against HIV-1 infection and AIDS Annu Rev Immunol 2002 20 73 99 11861598 Zolla-Pazner S Identifying epitopes of HIV-1 that induce protective antibodies Nat Rev Immunol 2004 4 199 210 15039757 Banchereau J Steinman RM Dendritic cells and the control of immunity Nature 1998 392 245 252 9521319 Beutler B Rietschel ET Innate immune sensing and its roots: The story of endotoxin Nat Rev Immunol 2003 3 169 176 12563300 Akira S Takeda K Toll-like receptor signalling Nat Rev Immunol 2004 4 499 511 15229469 Kubo M Hanada T Yoshimura A Suppressors of cytokine signaling and immunity Nat Immunol 2003 4 1169 1176 14639467 Alexander WS Hilton DJ The role of suppressors of cytokine signaling (SOCS) proteins in regulation of the immune response Annu Rev Immunol 2004 22 503 529 15032587 Marine JC Topham DJ McKay C Wang D Parganas E SOCS1 deficiency causes a lymphocyte-dependent perinatal lethality Cell 1999 98 609 616 10490100 Alexander WS Starr R Fenner JE Scott CL Handman E SOCS1 is a critical inhibitor of interferon gamma signaling and prevents the potentially fatal neonatal actions of this cytokine Cell 1999 98 597 608 10490099 Naka T Tsutsui H Fujimoto M Kawazoe Y Kohzaki H SOCS-1/SSI-1-deficient NKT cells participate in severe hepatitis through dysregulated cross-talk inhibition of IFN-gamma and IL-4 signaling in vivo Immunity 2001 14 535 545 11371356 Hanada T Yoshida H Kato S Tanaka K Masutani K Suppressor of cytokine signaling-1 is essential for suppressing dendritic cell activation and systemic autoimmunity Immunity 2003 19 437 450 14499118 Kinjyo I Hanada T Inagaki-Ohara K Mori H Aki D SOCS1/JAB is a negative regulator of LPS-induced macrophage activation Immunity 2002 17 583 591 12433365 Baetz A Frey M Heeg K Dalpke AH Suppressor of cytokine signaling (SOCS) proteins indirectly regulate Toll-like receptor signaling in innate immune cells J Biol Chem 2004 279 54708 54715 15491991 Gingras S Parganas E de Pauw A Ihle JN Murray PJ Re-examination of the role of suppressor of cytokine signaling 1 (SOCS1) in the Regulation of Toll-like receptor signaling J Biol Chem 2004 279 54702 54707 15491990 Shen L Evel-Kabler K Strube R Chen SY Silencing of SOCS1 enhances antigen presentation by dendritic cells and antigen-specific anti-tumor immunity Nat Biotechnol 2004 22 1546 1553 15558048 Hanada T Tanaka K Matsumura Y Yamauchi M Nishinakamura H Induction of hyper Th1 cell-type immune responses by dendritic cells lacking the suppressor of cytokine signaling-1 gene J Immunol 2005 174 4325 4332 15778397 Le Bon A Schiavoni G D'Agostino G Gresser I Belardelli F Type i interferons potently enhance humoral immunity and can promote isotype switching by stimulating dendritic cells in vivo Immunity 2001 14 461 470 11336691 Livak KJ Schmittgen TD Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method Methods 2001 25 402 408 11846609 Hauser H Shen L Gu QL Krueger S Chen SY Secretory heat-shock protein as a dendritic cell-targeting molecule: A new strategy to enhance the potency of genetic vaccines Gene Ther 2004 11 924 932 15085173 You Z Huang X Hester J Toh HC Chen SY Targeting dendritic cells to enhance DNA vaccine potency Cancer Research 2001 61 3704 3711 11325842 Allen JE Maizels RM Th1-Th2: Reliable paradigm or dangerous dogma? Immunol Today 1997 18 387 392 9267081 MacDonald AS Pearce EJ Cutting edge: Polarized Th cell response induction by transferred antigen-pulsed dendritic cells is dependent on IL-4 or IL-12 production by recipient cells J Immunol 2002 168 3127 3130 11907061 Gor DO Rose NR Greenspan NS TH1-TH2: A procrustean paradigm Nat Immunol 2003 4 503 505 12774069 D'Andrea A Ma X Aste-Amezaga M Paganin C Trinchieri G Stimulatory and inhibitory effects of interleukin (IL)-4 and IL-13 on the production of cytokines by human peripheral blood mononuclear cells: Priming for IL-12 and tumor necrosis factor alpha production J Exp Med 1995 181 537 546 7836910 Balazs M Martin F Zhou T Kearney J Blood dendritic cells interact with splenic marginal zone B cells to initiate T-independent immune responses Immunity 2002 17 341 352 12354386 Litinskiy MB Nardelli B Hilbert DM He B Schaffer A DCs induce CD40-independent immunoglobulin class switching through BLyS and APRIL Nat Immunol 2002 3 822 829 12154359 MacLennan I Vinuesa C Dendritic cells, BAFF, and APRIL: Innate players in adaptive antibody responses Immunity 2002 17 235 238 12354377 Fantuzzi L Purificato C Donato K Belardelli F Gessani S Human immunodeficiency virus type 1 gp120 induces abnormal maturation and functional alterations of dendritic cells: A novel mechanism for AIDS pathogenesis J Virol 2004 78 9763 9772 15331709 Soumelis V Scott I Gheyas F Bouhour D Cozon G Depletion of circulating natural type 1 interferon-producing cells in HIV-infected AIDS patients Blood 2001 98 906 912 11493432 Barron MA Blyveis N Palmer BE MaWhinney S Wilson CC Influence of plasma viremia on defects in number and immunophenotype of blood dendritic cell subsets in human immunodeficiency virus 1-infected individuals J Infect Dis 2003 187 26 37 12508143 Pacanowski J Kahi S Baillet M Lebon P Deveau C Reduced blood CD123+ (lymphoid) and CD11c+ (myeloid) dendritic cell numbers in primary HIV-1 infection Blood 2001 98 3016 3021 11698285 Dubois B Massacrier C Vanbervliet B Fayette J Briere F Critical role of IL-12 in dendritic cell-induced differentiation of naive B lymphocytes J Immunol 1998 161 2223 2231 9725215 Dubois B Vanbervliet B Fayette J Massacrier C Van Kooten C Dendritic cells enhance growth and differentiation of CD40-activated B lymphocytes J Exp Med 1997 185 941 951 9120400 Skok J Poudrier J Gray D Dendritic cell-derived IL-12 promotes B cell induction of Th2 differentiation: A feedback regulation of Th1 development J Immunol 1999 163 4284 4291 10510367 Trinchieri G Interleukin-12 and the regulation of innate resistance and adaptive immunity Nat Rev Immunol 2003 3 133 146 12563297 Eyles JL Metcalf D Grusby MJ Hilton DJ Starr R Negative regulation of interleukin-12 signaling by suppressor of cytokine signaling-1 J Biol Chem 2002 277 43735 43740 12221108 Feltquate DM Heaney S Webster RG Robinson HL Different T helper cell types and antibody isotypes generated by saline and gene gun DNA immunization J Immunol 1997 158 2278 2284 9036975 Inaba K Steinman RM Van Voorhis WC Muramatsu S Dendritic cells are critical accessory cells for thymus-dependent antibody responses in mouse and in man Proc Natl Acad Sci U S A 1983 80 6041 6045 6351074 Flamand V Sornasse T Thielemans K Demanet C Bakkus M Murine dendritic cells pulsed in vitro with tumor antigen induce tumor resistance in vivo Eur J Immunol 1994 24 605 610 8125131 Imami N Pires A Hardy G Wilson J Gazzard B A balanced type 1/type 2 response is associated with long-term nonprogressive human immunodeficiency virus type 1 infection J Virol 2002 76 9011 9023 12186885 Letvin NL Walker BD Immunopathogenesis and immunotherapy in AIDS virus infections Nat Med 2003 9 861 866 12835706 Colonna M Can we apply the TH1-TH2 paradigm to all lymphocytes? Nat Immunol 2001 2 899 900 11577341 Catlett IM Hedrick SM Suppressor of cytokine signaling 1 is required for the differentiation of CD4+ T cells Nat Immunol 2005 6 715 721 15924143 Chong MM Metcalf D Jamieson E Alexander WS Kay TW Suppressor of cytokine signaling-1 in T cells and macrophages is critical for preventing lethal inflammation Blood 2005 106 1668 1675 15899915 Judge AD Sood V Shaw JR Fang D McClintock K Sequence-dependent stimulation of the mammalian innate immune response by synthetic siRNA Nat Biotechnol 2005 23 457 462 15778705 Hornung V Guenthner-Biller M Bourquin C Ablasser A Schlee M Sequence-specific potent induction of IFN-alpha by short interfering RNA in plasmacytoid dendritic cells through TLR7 Nat Med 2005 11 263 270 15723075 Sledz CA Holko M de Veer MJ Silverman RH Williams BR Activation of the interferon system by short-interfering RNAs Nat Cell Biol 2003 5 834 839 12942087 Barouch DH Santra S Schmitz JE Kuroda MJ Fu TM Control of viremia and prevention of clinical AIDS in rhesus monkeys by cytokine-augmented DNA vaccination Science 2000 290 486 492 11039923 Oh S Berzofsky JA Burke DS Waldmann TA Perera LP Coadministration of HIV vaccine vectors with vaccinia viruses expressing IL-15 but not IL-2 induces long-lasting cellular immunity Proc Natl Acad Sci U S A 2003 100 3392 3397 12626740 Dale CJ De Rose R Wilson KM Croom HA Thomson S Evaluation in macaques of HIV-1 DNA vaccines containing primate CpG motifs and fowlpoxvirus vaccines co-expressing IFNgamma or IL-12 Vaccine 2004 23 188 197 15531036 Grohmann U Belladonna ML Bianchi R Orabona C Ayroldi E IL-12 acts directly on DC to promote nuclear localization of NF-kappaB and primes DC for IL-12 production Immunity 1998 9 315 323 9768751 Qu C Moran TM Randolph GJ Autocrine type I IFN and contact with endothelium promote the presentation of influenza A virus by monocyte-derived APC J Immunol 2003 170 1010 1018 12517968 Gett AV Sallusto F Lanzavecchia A Geginat J T cell fitness determined by signal strength Nat Immunol 2003 4 355 360 12640450 Iezzi G Karjalainen K Lanzavecchia A The duration of antigenic stimulation determines the fate of naive and effector T cells Immunity 1998 8 89 95 9462514 Iezzi G Scotet E Scheidegger D Lanzavecchia A The interplay between the duration of TCR and cytokine signaling determines T cell polarization Eur J Immunol 1999 29 4092 4101 10602020 Langenkamp A Casorati G Garavaglia C Dellabona P Lanzavecchia A T cell priming by dendritic cells: Thresholds for proliferation, differentiation and death and intraclonal functional diversification Eur J Immunol 2002 32 2046 2054 12115626 Lanzavecchia A Sallusto F Dynamics of T lymphocyte responses: Intermediates, effectors, and memory cells Science 2000 290 92 97 11021806 Schluns KS Lefrancois L Cytokine control of memory T-cell development and survival Nat Rev Immunol 2003 3 269 279 12669018 Becker TC Wherry EJ Boone D Murali-Krishna K Antia R Interleukin 15 is required for proliferative renewal of virus-specific memory CD8 T cells J Exp Med 2002 195 1541 1548 12070282 Madakamutil LT Christen U Lena CJ Wang-Zhu Y Attinger A CD8alphaalpha-mediated survival and differentiation of CD8 memory T cell precursors Science 2004 304 590 593 15105501 Ma X Montaner LJ Proinflammatory response and IL-12 expression in HIV-1 infection J Leukoc Biol 2000 68 383 390 10985255 Carbonneil C Donkova-Petrini V Aouba A Weiss L Defective dendritic cell function in HIV-infected patients receiving effective highly active antiretroviral therapy: Neutralization of IL-10 production and depletion of CD4+ CD25+ T cells restore high levels of HIV-specific CD4+ T cell responses induced by dendritic cells generated in the presence of IFN-alpha J Immunol 2004 172 7832 7840 15187167 Lu W Arraes LC Ferreira WT Andrieu JM Therapeutic dendritic-cell vaccine for chronic HIV-1 infection Nat Med 2004 10 1359 1365 15568033 Yang X Lee J Mahony EM Kwong PD Wyatt R Highly stable trimers formed by human immunodeficiency virus type 1 envelope glycoproteins fused with the trimeric motif of T4 bacteriophage fibritin J Virol 2002 76 4634 4642 11932429 Gao F Weaver EA Lu Z Li Y Liao H-X Antigenicity and immunogenicity of an HIV-1 group M consensus envelope J Virol 2005 79 1154 1163 15613343	16381597	PMC1323501	CC BY	2021-01-05 11:13:40	no	PLoS Med. 2006 Jan 3; 3(1):e11	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030011	oa_comm
==== Front PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1637950010.1371/journal.pmed.0030044Research ArticleCardiology/Cardiac SurgeryEpidemiology/Public HealthHealth PolicyHealth PolicyCardiovascular MedicineCohort studiesPublic HealthResearch designExclusion and Inclusion of Nonwhite Ethnic Minority Groups in 72 North American and European Cardiovascular Cohort Studies Cohort Studies and EthnicityRanganathan Meghna 1 Bhopal Raj 2 1The Robert Wood Johnson Foundation, Princeton, New Jersey, United States of America2Public Health Sciences Section, Division of Community Health Sciences, University of Edinburgh, Edinburgh, United KingdomGill Paramjit Academic EditorUniversity of BirminghamUnited Kingdom To whom correspondence should be addressed. E-mail: [email protected] Competing Interests: RB served on the Biobank UK Science Committee (2003–July 2004) and is currently a member of the Scottish Regional Collaborating Centre of Biobank UK. The conception of this paper precedes these roles. The views expressed here are personal and do not represent those of Biobank UK. Author Contributions: MR and RB designed the study and analyzed the data. MR and RB contributed to writing the paper. 3 2006 3 1 2006 3 3 e4415 5 2005 4 11 2005 Copyright: © 2006 Ranganathan and Bhopal.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Cardiovascular Disease Research: Time to Focus on Minority Ethnic Groups Background Cohort studies are recommended for understanding ethnic disparities in cardiovascular disease. Our objective was to review the process for identifying, including, and excluding ethnic minority populations in published cardiovascular cohort studies in Europe and North America. Methods and Findings We found the literature using Medline (1966–2005), Embase (1980–2001), Cinahl, Web of Science, and citations from references; consultations with colleagues; Internet searches; and RB's personal files. A total of 72 studies were included, 39 starting after 1975. Decision-making on inclusion and exclusion of racial/ethnic groups, the conceptual basis of race/ethnicity, and methods of classification of racial/ethnic groups were rarely explicit. Few publications provided details on the racial/ethnic composition of the study setting or sample, and 39 gave no description. Several studies were located in small towns or in occupational settings, where ethnic minority populations are underrepresented. Studies on general populations usually had too few participants for analysis by race/ethnicity. Eight studies were explicitly on Caucasians/whites, and two excluded ethnic minority groups from the whole or part of the study on the basis of language or birthplace criteria. Ten studies were designed to compare white and nonwhite populations, while five studies focused on one nonwhite racial/ethnic group; all 15 of these were performed in the US. Conclusions There is a shortage of information from cardiovascular cohort studies on racial/ethnic minority populations, although this has recently changed in the US. There is, particularly in Europe, an inequity resulting from a lack of research data in nonwhite populations. Urgent action is now required in Europe to address this disparity. A systematic review reveals a shortage of information on racial and ethnic minority populations. This compromises the relevance of the evidence underlying health policies and guidelines for nonwhite patients. ==== Body Introduction Cardiovascular disease is the most common cause of death in most industrialised societies and is either the leading or a dominant cause of death for all racial and ethnic groups in the US and the UK. The risk is especially high amongst those originating from the Indian subcontinent—South Asians [1]. Research on ethnic group differences and similarities may potentially help advance understanding of the relationships between risk factors and cardiovascular disease. Cardiovascular cohort studies have been one of the key approaches for achieving such understanding [2,3]. Most such studies started after World War II, when coronary heart disease mortality increased in many western countries [2]. This period coincided with an expansion of migration from developing to industrialised countries, leading to a marked increase in ethnic diversity in Europe and North America in the late 20th century (http://www.migrationinformation.org/GlobalData/countrydata/data.cfm). The inclusion of minority groups in such cohort studies is important not only to compare differences in health status between groups but also to assess risk factor-outcome relationships within such groups. Levy [3] has called for cohort studies to seek answers to ethnic disparities in cardiovascular risks identified in cross-sectional work, while Bhopal and Senior have outlined the problems and potential of ethnicity as an epidemiological variable [4]. The main objective of this review was to identify how the major cardiovascular cohort studies in North America and Europe included or excluded ethnic minority populations. The methods and aims of this review could be extended, but these geographical areas were chosen because cardiovascular cohort studies have been pioneered by groups in these locations [2]. There is no clearly defined line between what is, and what is not, a cardiovascular cohort study, and individual judgment is required to make that determination. For the purposes of this review, cardiovascular cohort studies were defined as prospective studies in defined populations, with a primary aim of studying risk factor-outcome relationships for major diseases such as stroke and coronary heart disease. Studies included are summarised in Table S1 [5–76]. Cohort studies with a multipurpose aim, those focused on other diseases, and those arising from studies originally designed as cross-sectional surveys or trials were generally excluded, as were studies of populations in which the investigators had little or no control over the sample (e.g., volunteers), although they may have yielded some cardiovascular data. A list of the studies that were given careful consideration but excluded, with reasons given, is in Table S2. Our reasoning for focusing on cardiovascular cohort studies, in addition to personal and academic interest, was this: Ethnic variations in cardiovascular disease give a clear rationale for inclusion of ethnic and racial minority groups, which may not be present for other conditions. This review may help health and research policy makers and the research community to judge whether there is equity, by which we mean needs of different populations have been met equally well, and, if not, whether we need new studies. Methods Search Strategy The starting point was a preliminary list prepared by RB in 1999. Both authors searched for studies independently between the period April 2000 through September 2005, using a variety of sources and repeated searches. Articles were identified using the electronic databases Medline (1966–2001, repeated in 2003 and 2005), Embase (1980–2001), Web of Science, and Cinahl using the following keywords: “cardiovascular disease” or “atherosclerosis” or “coronary heart disease,” and “cohort studies” or “epidemiological studies” or “prospective studies”. The search was repeated with the words “ethnicity” or “ethnic groups” or “racial groups” added. In Medline the search used free text and MESH terms. This led to more than 150 references in each database. The keywords “ethnicity and cardiovascular disease” used in the Web of Science database yielded more than 300 references. We examined the bibliographies of retrieved articles such as the meta-analysis of prospective observational studies by the Oxford Collaborative Group [2], and searched the Internet using the search engine Google and the Web sites of the British Medical Journal, National Research Register, Medical Research Council, and National Heart, Lung, and Blood Institute. Also, the names of specific cohort studies were keyed into the Internet search engines, e.g., for the Atherosclerosis Risk in Communities (ARIC) Study. Finally, colleagues were consulted, RB's literature files were examined, and referees pointed to additional studies. Grey literature (unpublished reports and abstracts) and editorial correspondence were not included. The search was limited at the outset to papers in English, as the authors cannot read other languages and most major cohort studies are published in English-language journals. Database filters for English papers only were not applied. Papers with titles in non-English languages were not considered further. Although a count of such papers was not made, our impression is that they were few. Nonetheless, it is unlikely that any important European studies published in languages other than English have been missed. At the Migrant Health in Europe International Conference held in Rotterdam in 2004, RB led a workshop discussing the potential development of a European multiethnic cardiovascular cohort study. Separately, RB presented this paper. The audience of knowledgeable participants were unaware of similar studies in Europe. At the conference many papers on cardiovascular diseases were presented, but none reported such studies. Professor Marc Bruijneels has collected information across Europe, for a proposed European project to compile data by ethnic group but he found no cardiovascular cohort data by ethnic group (personal communication). Studies were eligible for consideration that were designed to examine prospectively the relationship between risk factors and cardiovascular disease outcomes (coronary heart disease and stroke) in population samples. By population samples we mean natural living populations and exclude studies of people with existing diseases. In view of the nature of this study reflecting how investigators made decisions on who to include and exclude in the sample, studies without a sampling frame based on volunteers were excluded, as investigators would have little decision-making latitude in such circumstances. Some studies were concerned with multiple outcomes, and where publications showed an emphasis on cardiovascular diseases these were included, e.g., the Nurses Health Study [63]. Cohort studies where the sample was defined retrospectively based on records (investigators have limited choices over sampling and on ethnic coding in such studies), trials, cross-sectional studies, case control studies, and studies based solely on routine statistics were excluded. Inclusion decisions required a degree of judgment and flexibility because, as stated above, there is no firm definition for a cardiovascular cohort study. Furthermore, investigators themselves conducted and analysed their studies flexibly, and assessing the study design was not always easy, e.g., with follow-up of studies that were originally cross-sectional. We also followed advice from referees, e.g., The Women's Health Initiative observational study was included, although a major component of this study is a trial. Multiple Risk Factor Intervention Trial (MRFIT) was designed as a trial, although it contained an observational component (with volunteer samples), and was excluded. Studies designed to study cancer, e.g., the European Prospective Investigation of Cancer (EPIC) [77] and the Multi-Ethnic Los Angeles Cohort Study, were excluded (Table S2), although such studies may shed light on cardiovascular diseases. In total, 72 cardiovascular cohort studies in the US/North America and Europe were included. This review focused on papers describing the design and rationale of the study. For example, the Framingham Study has hundreds of secondary papers, but a few discussing methods were identified. This approach is justified on both scientific and pragmatic grounds. Scientifically, inclusion/exclusion of particular populations is a design issue that is handled at the planning stage. Pragmatically, it would have been inefficient to examine multiple papers for information that ought to be provided in the baseline paper. Only one paper per study is cited here, although sometimes several were examined. Studies based on combining existing cohorts, e.g. the Sleep, Heart and Health Study (Table S2), are not included here, but the original studies are when appropriate. Research Questions and Data Extraction The research questions that guided data collection from the studies are listed in Table 1. Information extracted from publications was directly entered into Table S1. Both authors independently examined all the papers with virtually complete agreement. The few disagreements were resolved by conferring. Table 1 Research Questions Terminology and Concepts of Race and Ethnicity Wherever possible and appropriate, the terminology used for ethnic group classification has been quoted directly from the paper, even when this is not in agreement with currently accepted terminology and is potentially offensive. Similarly, we have accepted the concepts of race and ethnicity provided by authors, but for reasons discussed by Senior and Bhopal we have tended to use the word ethnicity rather than race, and we apply the concepts as discussed in their paper [4]. Our use of the term nonwhite reflects our focus on populations that do not have European ancestral origins (described, using current conventions, as white), and would not describe themselves, or be perceived as, white. This focus reflects long-standing, widespread concern about inequities in health and health care that are particular to such populations. Results Overview The main aim of each study, with slight variations, was to determine the incidence of coronary heart disease and/or stroke and study risk factor-outcome relationships. Table 2 summarises and Table S1 lists the 72 studies included [5–76]. The studies started between 1946 and 2000, with 39 starting after 1975, by which time ethnic minority populations were becoming well established in Western Europe (http://www.migrationinformation.org/GlobalData/countrydata/data.cfm, and knowledge of ethnic variations in cardiovascular disease was appearing in Europe. Studies numbered 41 in Europe, 31 in North America, and one (the Seven Countries study) in both. Ten studies were designed to compare white and nonwhite populations, while five studies focused on one nonwhite racial/ethnic group; all 15 of these were conducted in the US. Table 2 Summary of Key Findings in Table S1: Inclusion and Exclusion of Racial/Ethnic Minorities in Cardiovascular Cohort Studies Studies seldom provided details on the racial or ethnic composition of the study setting or sample, and when they did the details were minimal; 39 gave no description at all. Several studies were located in small towns or in occupational settings, whereas minority populations tend to live in cities and work in a restricted range of workplaces. The investigators in some studies saw the population homogeneity of such locations as valuable. Studies that were based on general populations usually had too few participants for analysis by race or ethnicity. The process by which decisions on inclusion and exclusion of racial/ethnic groups were rarely made explicit. Eight studies explicitly stated they were on Caucasians or whites, and two excluded ethnic minority groups using language and birthplace criteria. One study [61] excluded the nonwhite population (6,236 people) from the incidence component of the research because of small numbers and low response rates to mailed questionnaires. There were other examples of studies including ethnic minority groups in the baseline phase of the study but reporting cohort analyses in the white population. There were major differences in the extent of inclusion of minority groups between Europe and the US. Europe None of the studies done in Europe mentioned studying nonwhite racial or ethnic variations in their aims. The ethnic composition of the source population for the sample was not described or discussed, but sometimes the text showed awareness of the issue of ethnic heterogeneity; e.g., the Paris Prospective Study [12] was explicitly of native-born French men, while the Second Manifestations of ARTerial disease (SMART) study [43] was of Dutch speakers. Two studies intended to examine European origin ethnic groups. The Yugoslavia Cardiovascular Disease Study [7] contrasted Muslim and Roman Catholic populations, but there was little detail on their characteristics. The Cardiovascular Disease in Norwegian Counties Study [21] collected information by ethnic group but provided no analysis on this variable. Few authors specified the ethnic composition of their sample, and when they did it was usually with the label Caucasian or white with no or sparse detail on the assignment of ethnic/racial group. North America Most North American studies gave some attention to the issue of ethnicity and race, usually in relation to the sample rather than the racial/ethnic composition of the setting of the study. The assignment of racial/ethnic group, and its validity, were not made explicit. Five studies focused on one nonwhite group: the Meharry Cohort Study [50] focused on African-American students; the Jackson Heart Study [75] focused on a black population; the Gila River Indians Community Study [65] studied adult American Indians (Pimas), as did the Strong Heart Study (12 American Indian tribes in Arizona, Oklahoma and North and South Dakota) [66]; and the NI-HON-SAN project [55] studied adults of Japanese origins, comparing American and Japanese locations. Ten studies compared one or more ethnic groups simultaneously: the Evans County Study [52], the ARIC study [69], and the Charleston Heart Study [53] compared black and white adults; the CARDIA Study [67] compared black and white students; the Bogalusa Heart Study [60] compared black and white school children; the San Antonio Heart Study compared cardiovascular risk factors in Mexican Americans and Anglos [68]; the National Health and Nutrition Examination Survey (NHANES) [57], the Multi-Ethnic Prospective Cohort [62], the Women's Health Initiative [73], and the Multi-Ethnic Study of Atherosclerosis [76], were multiethnic studies. Discussion The study of ethnic variations in disease is long established, with a research base founded in the 19th century and strengthening in the 20th century, particularly in North America. The cardiovascular research community has been aware for some decades of important variations in the frequency, causes and consequences of cardiovascular diseases in non-European origin ethnic minority populations. Well-publicised government reports and academic papers discussed these issues in the mid-1970s and 1980s, e.g., heart attacks in East London were shown to be high in Asians (mainly from the Indian subcontinent) and low in Caribbeans [78]. Mortality rates may vary threefold or more between minority ethnic groups; e.g., in a comparison of Chinese and South Asian populations living in the UK [1], the variations were much larger than those among white minority populations, e.g., the Irish-born people living in England and Wales compared to the whole population there [79]. There is no published major cohort study focusing on ethnic group variations in Europe, but a growing information base is developing in North America. This observation is important for both policy and practice. For example, risk prediction models have been developed from data on white European origin populations, and their unreliability in relation to racial/ethnic minority groups is recognised [80, 81]. Within the cardiovascular field there is also concern about possible racial disparities in health care and outcomes [82,83]. Ethnic minority groups in the US and in Europe are “at-risk” of differential treatment, particularly for surgical therapies, and several explanations, including institutional discrimination, are being pursued [83]. The decisions of most individual investigators undertaking cohort studies to concentrate on white European origin populations may have been scientifically sound and well meant, but collectively, especially in Europe but also for some ethnic groups in the US, it may have resulted in a lack of attention to the needs of nonwhite populations. Limitations of This Study Since this review was limited to papers published in English, there was the potential to miss relevant, large-scale cardiovascular cohort studies in Europe and North America published in other languages. The reference lists of papers examined did not, however, cite them, and consultations around Europe did not identify them (see Methods). Bias from such omissions, if any, is unlikely to alter the conclusions of this paper. One or a few papers from each study, usually those giving adequate detail on the rationale and design of the study (sample, participants, methods used, etc.), were studied. Secondary papers may mention ethnic groups, as in the Whitehall Study, where cross-sectional analyses were done [84]. This said, the papers we studied clarified the primary intentions and design of each study. It is axiomatic that unless the race/ethnicity component is considered at the design stage, and the ethnic group of participants is identified, useful data on this issue are unlikely to accrue later. Our search strategy, which excluded manual searching of journals, may have missed some studies, as was also acknowledged by The Prospective Studies Collaboration [2]. It would be an inefficient, and possibly futile, exercise to catalogue every study, especially as several cohorts have led to hundreds of papers. However, to our knowledge, this is the most complete list of cardiovascular cohort studies available. We have excluded small-scale studies that were combined to create a cohort, e.g., the Italian RIFLE project consisting of 45 “cohorts” and a total sample of 32,726. There is no mention of race or ethnicity in the paper [85]. It is improbable that such small individual studies included the ethnic dimension. The studies included meet the highest standards, as indicated by publication in journals indexed by electronic databases. There are many other cohort studies that are multipurpose or focus on noncardiovascular diseases. In theory these could potentially yield data on ethnic variations in cardiovascular diseases. Our assessment of such studies suggests that they give no more detail on the racial/ethnic issues than those we have examined (Table S2). For example, the EPIC-Norfolk study did not discuss ethnicity [77]. Unpublished (grey) literature has not been included in this review, with the exception of the Jackson Study [75], which links to the published ARIC study and is fully described on a website. This exception was made because of its obvious importance. We are aware of some cross-sectional studies (generally small) that are designed with linkage to mortality follow-up, e.g., the Southall Study [86] and the Newcastle Heart Project [87], that will publish risk factor-outcome data in due course. These were, however, designed with the power for cross-sectional and not cohort analyses. A cohort study of Indian Asians in West London is ongoing (Jaspal Kooner, personal communication), but this study will not address the needs of other minority ethnic groups in the UK. There are also many studies designed as trials that have long-term follow-up and provide opportunities for cohort type analysis, e.g., MRFIT [88]. Analysis of inclusion and exclusion of ethnic minority populations in trials and other study designs was beyond the scope of this study, although it may be that the findings will be similar. Further research on cardiovascular trials, cross-sectional studies, and case-control studies might be illuminating. These limitations do not, however, alter our main conclusions. Interpretation of Main Results The review answered the research questions (Table 1). There are many cardiovascular cohort studies, indicating their perceived and actual importance. The ethnic composition of the population where the studies were based, and the process of inclusion/exclusion of ethnic minority groups, was not a point of emphasis in publications. Many studies gave little or no data on the ethnic composition of the sample, or the description was limited and based on ethnic group labels. With the exemplary exception of the San Antonio study, which developed an algorithm based on a range of data, studies did not provide details of the processes for ethnic coding. Cohort-based analysis by ethnic group is available in the US for a number of ethnic groups, but not in Europe. Although the sample size may be too small to produce analysis by ethnic group, inclusion of minority populations is still important. Such cohorts are population-based and should be generalisable to populations similar to that from which the sample has been drawn; without information on ethnic composition, generalisability becomes more difficult. Such studies can also provide a foundation for larger studies focusing on minority populations, and potentially could lead to analysis by ethnic group after pooling of data. Analysis of data by ethnic group requires adequately powered studies, and these will be large-scale, expensive, and challenging. Such studies will be funded only when there is agreement on the need for them. There are more than 30 European cohort studies, many started within the last 20 years when ethnic variations were already described, yet collectively or singly they are unable to provide analysis by ethnic group. This paper contributes to the needs assessment. Many cohort studies have focused on white populations despite being set in multiracial/ethnic nations and regions (http://www.migrationinformation.org/GlobalData/countrydata/data.cfm). This is especially so in European studies, e.g., those set in major cities such as London, Amsterdam, and Paris. This observation applies to studies started after the mid-1970s when understanding about the needs of minority groups was substantial, many cohort studies on white populations were in place, and knowledge about causes and control of cardiovascular disease in white populations was already advanced. The studies that were designed to study racial/ethnic minority groups were all in the US, and were started more recently, in response to an increasing recognition of the needs of ethnic minorities. Some studies were openly exclusive, e.g., including only the native born or native language speakers, or simply being confined to whites/Caucasians. Some studies used members of occupational groups as study participants. This can lead to exclusion of minority populations, perhaps unwittingly, in that unemployment is usually higher in ethnic minority populations, some of which have comparatively high levels of self-employment and employment in small workplaces and are less likely to be a substantial proportion of the work force of large employers [89]. While recruiting randomly is, arguably, fair, it rarely permits analyses by ethnic group, because the resultant sample size is too small, except for European-origin populations. This can lead to incidence data analyses in white populations and more limited analysis, e.g., cross-sectional in ethnic minority groups [84]. By choosing small towns or rural areas, as their base, as in the Framingham, Tecumseh, Seven Countries, Caerphilly, and British Regional Heart Studies, investigators gain population stability and homogeneity but miss multiethnic populations living predominantly in inner city areas, where the cardiovascular disease prevention challenge is greatest. In these circumstances there is a case for more purposive sampling, including weighting the sample to augment the number of ethnic minority participants. Studies that started after 1975 and which, in retrospect, might have been designed in this way include Whitehall 2, Rotterdam Elderly Study, British Women's Heart and Health Study, Nurses Health Study, Iowa Women's Health Study, and the Health Professionals Heart Study. With the exception of black/white comparisons, the opportunity for multiethnic comparison has not been fully exploited, although several recent studies in the US promise a truly multiethnic approach. Explanations for the findings here include scientific pragmatism, shortage of resources, potential difficulties in accessing populations and in gaining informed consent, insufficient expertise and experience, lack of interest, a resistance to dividing populations by ethnic or racial status (particularly in some countries of mainland Europe), and the possibility of indirect or direct discrimination. In many ways the issues highlighted echo those applying to women until recently. These explanations require further analysis and research. Data are vital to assess the needs of ethnic minority groups; to implement, evaluate, and adjust the necessary health policies; and to provide excellent clinical care based on valid risk prediction models. The Race Relations (Amendment) Act 2000 in the UK [90] and laws in Europe will mandate a change of strategy for all public sector organisations, including those that commission or fund research, such as the Medical Research Council. In the US, the NIH Strategic Research Plan (2002–2006) promises to spearhead further change, building on previous NIH policies [91]. This paper indicates that researchers in the US have responded to NIH policies promoting the inclusion of ethnic minority populations in research. Studies exploring ethnic variations may lead to insights that are generalisable to the whole population, in terms of both disease causation and the effectiveness of interventions and healthcare systems. Inclusion of ethnic minorities groups in research, therefore, is likely to benefit the whole population. A Lancet commentary has called for cohort studies in such groups [3]. The traditional approach, whereby researchers', peer reviewers', and funding bodies' interests drive the research agenda, needs to be balanced by a strategic needs-based approach if the inequity described in this paper is to be addressed. The planned Biobank UK study of a cohort of 500,000 people in the UK offers an opportunity to redress the gap in the UK—but only if it achieves its stated goals of recruiting ethnic minority groups with due emphasis on population heterogeneity, attention to cross-cultural comparability of data, and high retention of ethnic minority populations in the cohort (http://www.ukbiobank.ac.uk/). This paper raises broader questions that merit debate on how the research community responds to the increasing ethnic diversity of populations internationally. Supporting Information Table S1 Cardiovascular Cohort Studies Included by Area of Study (Europe and North America) and in Chronological Order of Fieldwork Found at 10.1371/journal.pmed.0030044.st001 (139 KB DOC). Click here for additional data file. Table S2 List of Studies Excluded from the Analysis Found at 10.1371/journal.pmed.0030044.st002 (41 KB DOC). Click here for additional data file. Patient Summary Background As a result of migration, mostly for economic reasons, populations in many countries around the world are becoming increasingly diverse. In many cases, ethnic minorities differ in their socio-economic circumstances, culture, lifestyle, and genetic make-up from the majority population (and there are, of course, also differences within the majority group and within minority groups). These differences in risk factors (such as diet and smoking) can influence a person's susceptibility to disease. For some diseases, such as heart disease, it is well known that particular ethnic groups are at higher risk than others. It is not always clear, though, whether this high risk is due to socio-economic circumstances, culture, genes, lifestyle, or a combination of these factors. Why Was This Study Done? Cardiovascular disease, which can cause heart attacks and strokes, is the most common cause of death in the US and most European countries. Rates of cardiovascular diseases vary substantially between different countries, and also between different ethnic groups in ethnically diverse countries such as the UK and the US. Researchers and health policy makers need to understand more about the variations in cardiovascular disease. To ensure the best possible health care for the entire population, they need to know how exactly the risks differ between different ethnic groups and what causes those differences. One key research tool to address these questions are so-called cohort studies (a cohort refers to a specific group of people that are studied over time). Cohort studies are “forward looking”—they typically enroll a large number of healthy participants who are then followed over a number or years to study long-term health outcomes. Over the past decades, a number of cohort studies have focused on cardiovascular disease. In this study, researchers wanted to find out whether these cohort studies included or excluded ethnic minority groups. What Did the Researchers Do and Find? They searched the medical literature for cohort studies on cardiovascular disease. They found 72 that met their criteria and analyzed them in detail. The researchers discovered that most of the cohort studies did not provide detailed information on the ethnic composition of the broader populations from which the participants were recruited. Most of them also did not state whether minority groups were included in or excluded from the study. Additionally, the researchers found that many of the studies did not give details on the ethnic composition of the participants themselves, or on how the participants' ethnicity was determined. Studies with participants that were representative of diverse populations usually were not large enough to answer the kinds of questions necessary to determine differences between different ethnic groups. However, ten of the studies were designed specifically to compare white and nonwhite participants, and five studies focused on nonwhite minority groups specifically. All 15 of those studies were done in the US. What Does This Mean? Despite the knowledge that ethnicity matters in cardiovascular disease, most cohort studies have not been designed to further explore this connection. The situation in the US seems to be changing, with a number of recent studies designed to add data on cardiovascular disease risks and causes among minority populations. No such studies have yet been reported in Europe. Research strategies in Europe should be adjusted to meet this need. Where Can I Find More Information Online? The following Web sites provide information on minority participation in health research. Office of Minority Health Research at the US National Heart, Lung, and Blood Institute: http://www.nhlbi.nih.gov/about/omha/ Center of Excellence in Minority Health and Health Disparities at Harvard Medical School: http://www.mfdp.med.harvard.edu/coe/ The Migration Information Source provides data on the composition of the populations for many countries: http://www.migrationinformation.org/ Department of Health, England: http://www.dh.gov.uk/PolicyAndGuidance/EqualityAndHumanRights/fs/en/ South Asian Health Foundation: http://www.sahf.org.uk/events.html We thank Marshall Dozier for help with Medline searches and records, Simon Capewell for his list of cohort studies, Gerry Fowkes and Peter Whincup for constructive criticism of an earlier previous draft, referees for valuable constructive criticism and information, and Lori Malatesta, Hazel King, and Tori Hastie for secretarial help. There was no specific funding, and the research was supported by the authors and their employers. The employers had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Citation: Ranganathan M, Bhopal R (2006) Exclusion and inclusion of nonwhite ethnic minority groups in 72 North American and European cardiovascular cohort studies. PLoS Med 3(3): e44. Abbreviations ARICAtherosclerosis Risk in Communities MRFITMultiple Risk Factor Intervention Trial ==== Refs References Gill PS Kai J Bhopal RS Wild S Raftery J Health care needs assessment: Black and minority ethnic groups The epidemiologically based needs assessment reviews. Third series 2002 In press. Available at: http://hcna.radcliffe-oxford.com/bemgframe.htm . Accessed 29 November 2005 Prospective Studies Collaboration Cholesterol, diastolic blood pressure, and stroke: 13,000 strokes in 450,000 people in 45 prospective cohorts Lancet 1995 346 1647 1653 8551820 Levy D Kannel WB Searching for answers to ethnic disparities in cardiovascular risk Lancet 2000 356 266 267 11071176 Senior PA Bhopal RS Ethnicity as a variable in epidemiological research BMJ 1994 309 327 330 8086873 Morris JN Kagan A Pattison DC Incidence and prediction of ischaemic heart disease in London busmen Lancet 1996 2 553 559 Keys A Seven countries. A multivariate analysis of death and coronary heart disease 1980 Cambridge (Massachusetts) Harvard University Press 361 Kozarevic D Pirc B Dawber TR Kahn HA Prevalence and incidence of coronary disease in a population study. The Yugoslavia Cardiovascular Disease Study J Chronic Dis 1971 24 495 505 5136242 Kornitzer M Dramaix M Beriot I Lannoy M Gheyssens H Twenty-five-year mortality follow-up in the Belgian Bank Study J Cardiol 1993 82 153 171 Reunanen A Aromaa A Pyvrdld K Punsar S Maatela J The Social Insurance Institution's coronary heart disease study. Baseline data and 5-year mortality experience Acta Med Scand 1983 673 Suppl 1 120 Jónsdóttir LS Sigfusson N Sigvaldason H Thorgeirsson G Incidence and prevalence of recognised and unrecognised myocardial infarction in women. The Reykjavik Study Eur Heart J 1998 19 1011 1018 9717035 Reid DD Bert GZ Hamilton PJS Jarett KJ Keen H Cardio-respiratory disease and diabetes among middle-aged male civil servants Lancet 1974 1 469 473 4131881 Ducimetiere P Richard J Cambien F The pattern of subcutaneous fat distribution in middle-aged men and the risk of coronary heart disease: The Paris Prospective Study Int J Obes 1986 10 229 240 3759330 Bengtsson C Blohme G Hallberg L Hallstrom T Issakson B The study of women in Gothenburg 1968–69—A population study. General design, purpose, and sampling results Acta Med Scand 1973 193 311 318 4717311 Tomas-Abadal L Varas-Lorenzo C Bernades-Bernat E Balaguer-Vintro I Coronary risk factors and a 20-year incidence of coronary heart disease and mortality in a Mediterranean industrial population. The Manresa Study, Spain Eur Heart J 1994 15 1028 1036 7988593 Gyntelberg F Screening for hypertension in an epidemiological study Acta Med Scand 1973 193 393 397 4717320 Meade TW North WRS Chakrabarti R Stirting Y Hemostatic function and cardiovascular death: Early results of a prospective study Lancet 1980 2 533 537 6157066 Leren P Askervold EM Foss OP Froli A Grymyr D The Oslo Study. Cardiovascular disease in middle-aged and young Oslo men Acta Med Scand 1975 588 1 38 Hawthorne VM Watt GCM Hart CC Hole DJ Cardiorespiratory disease in men and women in urban Scotland: Baseline characteristics of the Renfrew/Paisley (MIDSPAN) study population Scott Med J 1995 40 102 107 8787108 Vartiainen E Sarti C Tuomilehto J Kuulasmaa K Do changes in cardiovascular risk factors explain changes in mortality from stroke in Finland? BMJ 1995 310 901 904 7719179 Bonaa KH Arnesen E Association between heart rate and atherogenic blood lipidfractions in a population. The Tromso Study Circulation 1992 86 394 405 1638708 Bjartveit K Foss OP Gjervig T Lund-Larsen PG The cardiovascular disease study in Norwegian counties. Background and organisation Acta Med Scand (Suppl) 1979 634 5 70 Law MR Wald NJ Wu T Hackshaw A Bailey A Systematic underestimation of association between serum cholesterol concentration and ischemic heart disease in observational studies: Data from the BUPA study BMJ 1994 308 363 366 8124143 Rosolova H Simon J Sefrna F Impact of cardiovascular risk factors on morbidity and mortality in Czech middle-aged men: Pilsen Longitudinal Study Cardiology 1994 85 61 68 7954559 Hansen AT Schnohr P Jensen G Nyboe J Appleyard M The Copenhagen City Heart Study. Osterbroundersogelsen. A book of tables with data from the first examination (1976–78) and a five-year follow up (1981–83) Scand J Soc Med 1989 170 Suppl 41 11 180 De Maat MPM Bladbjerg EM Drivsholm T Borch-Johnsen K Moller L Inflammation, thrombosis and atherosclerosis: Results of the Glostrup Study J Thromb Haemost 2003 1 950 957 12871360 Bakx JC Veldstra MI Van de Hoogen HJM Zielhuis GA Thien T Blood pressure and cardiovascular morbidity and mortality in a Dutch Population: The Nijmegen Cohort Study Prev Med 2001 32 142 147 11162339 Kemper HCG The Amsterdam Growth Study: A longitudinal analysis ofhealth, fitness and lifestyle. Volume 6, HK Sport Science Monograph Series 1995 Champaign (Illinois) Human Kinetics 288 Shaper AG Pocock SJ Walker M Cohen M Wale CJ British Regional Heart Study: Cardiovascular risk factors in middle-aged men in 24 towns BMJ 1981 288 179 186 Schulte H Cullen P Assmann G Obesity, mortality and cardiovascular disease in the Munster Heart Study (PROCAM) Atherosclerosis 1999 144 199 209 10381293 De Backer G Regional differences in dietary habits, coronary risk factors and mortality rates in Belgium. I. Design and methodology Acta Cardiol 1984 39 285 292 6333125 Merlo J Ranstram J Liedholm H Hedblad B Lindberg G Incidence of myocardial infarction in elderly men being treated with anti-hypertensive drugs-population-based cohort study BMJ 1996 313 457 461 8776312 Caerphilly and Speedwell Collaborative Group Caerphilly and Speedwell Collaborative Heart Disease Studies J Epidemiol Community Health 1984 38 259 262 6332166 Puddu PE Lanti M Menotti A Mancini M Zanchetti A Serum uric acid for short-term prediction of CVD incidence in the Gubbio Population Study Acta Cardiology 2001 56 243 251 Salonen JT Tuomainen TP Salonen R Lakka TA Nyysonen R Donation of blood is associated with reduced risk of myocardial infarction Am J Epidemiol 1998 148 445 451 9737556 Marmot MG Davey Smith G Stansfield S Patel C North F Health inequalities among British civil servants: The Whitehall-II Study Lancet 1991 337 1387 1393 1674771 Ducimetiere P Guize L Milon MH Richard J Rufat P Arteriographically documented coronary artery disease and alcohol consumption in French men: The CORALI Study Eur Heart J 1995 14 727 733 Smith WC Crombie IK Tavendake R Irving JM Kenicer MB The Scottish Heart Health Study: Objectives and development of methods Health Trends 1987 45 211 217 Beks PJ Mackay AJC de Neeling JND de Vries H Bouter LM Peripheral arterial disease in relation to glycaemic level in an elderly Caucasian population: The Hoorn Study Diabetologia 1995 38 86 96 7744233 Houterman S Verschuren WM Hofman A Witteman JC Serum cholesterol as a risk factor for myocardial infarction in elderly men and women: The Rotterdam Study J Int Med 1999 246 25 33 Kiechl S Willeit J Egger G Oberhollenzer M Aichner F Alcohol consumption and carotid atherosclerosis: Evidence of dose-dependant atherogenic and antiatherogenic effects. Results from the Bruneck Study Stroke 1994 25 1593 1598 8042209 Fowkes FG Housley E Cawood EH Macintyre CC Ruckley CV Edinburgh Artery Study: Prevalence of asymptomatic and symptomatic peripheral arterial disease in the general population Int J Epidemiol 1991 20 384 391 1917239 Scarabin PY Aillaid MF Amouyel P Evans A Luc G Association of fibrinogen, factor 7 and PAI-1 with baseline findings among 10,500 male participants in a prospective study of myocardial infarction—The PRIME Study. Prospective Epidemiological Study of Myocardial Infarction Thromb Haemost 1998 80 749 756 9843166 Simons PC Algra A van de Laak MF Grobbee DE van der Graaf Second Manifestations of ARTerial disease (SMART) Study: Rationale and design Eur J Epidemiol 1999 15 773 781 10608355 Cremer P Nagel D Mann H Labrot B Muller-Berninger R Ten year follow-up results from the Goettingen Risk, Incidence and Prevalence Study (GRIPS). 1. Risk factors for myocardial infarction in a cohort of 5790 men Atherosclerosis 1997 129 221 230 9105565 Lawlor DA Taylor M Bedford C Ebrahim S Is housework good for health? Levels of physical activity and factors associated with activity in elderly women. Results from the British Women's Heart and Health Study J Epidemiol Community Health 2002 56 473 478 12011209 Mathewson FA Brereton CC Keltie WA Paul GI The University of Manitoba Follow-up study. A prospective investigation of cardiovascular disease Can Med Assoc J 1965 92 947 953 14285286 Klag MJ Ford DE Mead LA Jiang H Whelton PK Serum cholesterol in young men and subsequent cardiovascular disease N Engl J Med 1993 328 313 318 8419817 Dawber TR Kannel WB Lyell LP An approach to longitudinal studies in a community: The Framingham Study Ann N Y Acad Sci 1963 107 539 555 14025561 Paul O Lepper MH Phelan W A longitudinal study of coronary heart disease Circulation 1963 28 20 31 13941964 Thomas J Semenya KA Neser WB Thomas DJ Green DR Precursors of hypertension in black medical students: The Meharry Cohort Study J Natl Med Assoc 1984 76 111 112 6708119 Epstein FH Ostrander LD Johnson BC Payne MW Hayner NS Epidemiological studies of cardiovascular disease in a total community—Tecumseh, Michigan Ann Intern Med 1965 62 1170 1185 14295500 McDonough JR Hames CG Stulb SC Garrison GE Cardiovascular disease field study in Evans County, Ga Public Health Rep 1963 78 1051 1059 14090074 Keil JE Loadholt CB Weinrich C Sandifer SH Boyle E Incidence of coronary heart disease in blacks in Charleston, South Carolina Am Heart J 1984 108 779 786 6475747 Sesso HD Paffenbarger RS Ha T Lee I-M Physical activity and cardiovascular disease risk in middle-aged and older women Am J Epidemiol 1999 150 408 415 10453817 Donahue RP Abbott RD Reed D Yano K Physical activity and coronary heart disease in middle-aged and elderly men: The Honolulu Heart program Am J Public Health 1988 78 683 685 3369600 Stamler J Dyer AR Shekelle RB Neaton J Stamler R Relationship of baseline major risk factors to coronary and all-cause mortality, and to longevity: Findings from long-term follow-up of Chicago cohorts Cardiology 1993 82 191 222 8324780 He J Ogden LG Bazzano LA Vupputuri S Loria C Risk factors for congestive heart failure in US men and women: NHANES I epidemiologic follow-up study Arch Intern Med 2001 161 996 1003 11295963 Lauer MR Connor WE Leaverton PE Reiter MA Coronary heart disease risk factors in school: The Muscatine Study J Pediatr 1975 86 697 706 1133650 Barrett-Connor E Criqui MH Klauber MR Holdbrook M Diabetes and hypertension in a community of older adults Am J Epidemiol 1981 113 276 284 7468582 Croft JB Webber LS Parker FC Berenson GS Recruitment and participation of children a long-term study of cardiovascular disease: The Bogalusa Heart Study, 1973–1982 Am J Epidemiol 1984 120 436 448 6332531 Beeson WL Mills PK Phillips RL Andress M Fraser GE Chronic disease among Seventh-day Adventists, a low-risk group. Rationale, methodology, and description of the population Cancer 1989 64 570 581 2743251 Maskarinec G Meng L Kolonel LN Alcohol intake, body weight, and mortality in a multiethnic prospective cohort Epidemiology 1998 9 654 661 9799177 Manson JE Willett WC Meir J Stampfer MJ Colditz GA Body weight and mortality among women N Engl J Med 1995 333 677 685 7637744 Lee IM Paffenbarger R Physical activity and stroke incidence: The Harvard Alumni Health Study Stroke 1998 29 2049 2054 9756580 Howard VB Davis PM Pettitt DJ Knowler WC Bennett PH Plasma and lipoprotein cholesterol and triglyceride concentrations in the Pima Indians: Distributions differing from those of Caucasians Circulation 1983 68 714 724 6577980 Lee TE Welty TR Fabsitz R Cowan LD Le NA The Strong Heart Study. A study of cardiovascular disease in American Indians: Design and methods Am J Epidemiol 1990 132 1141 1155 2260546 Friedman GD Cutter GR Donahue RP Hughes GH Hulley SB CARDIA: Study design and recruitment and some characteristics of the examined subjects J Clin Epidemiol 1988 41 1105 1116 3204420 Stern MP Rosenthal M Haffner SM Hazuda HP Franco LJ Sex difference in the effects of sociocultural status on diabetes and cardiovascular risk factors in Mexican Americans. The San Antonio Heart Study Am J Epidemiol 1984 120 834 851 6507426 ARIC Investigators The Atherosclerosis Risk in Communities (ARIC) Study: Design and objectives Am J Epidemiol 1989 129 687 702 2646917 French SA Folsom AR Jeffery RW Williamson DR Prospective study of intentionality of weight loss and mortality in older women: The Iowa Women's Health Study Am J Epidemiol 1999 149 504 514 10084239 Rimm EB Ascherio A Giovannucci E Spiegelman D Stampfer Vegetable, fruit, and cereal fiber intake and risk of coronary heart disease among men JAMA 1996 275 447 451 8627965 Fried LP Borhani NO Enright P Furberg CD Gardin JM The Cardiovascular Health Study: Design and rationale Ann Epidemiol 1991 1 263 276 1669507 The Women's Health Initiative Study Group Design of the women's health initiative clinical trial and observational study Controlled Clin Trials 1997 19 61 109 Dwyer JH Navab M Dwyer KM Hassan K Sun P Oxygenated carotenoid lutein and progression of early atherosclerosis: The Los Angeles atherosclerosis study Circulation 2001 103 2922 2927 11413081 National Heart, Lung, and Blood Institute Jackson Heart Study Available: http://www.nhlbi.nih.gov/about/jackson/index.htm . Accessed 29 November 2005 Bild DE Bluemke DA Burke GL Detrano R Diez Roux AV Multi-ethnic study of atherosclerosis: Objectives and design Am J Epidemiol 2002 156 871 881 12397006 Day N Oakes S Luben R Khaw KT Bingham S EPIC-Norfolk: Study design and characteristics of the cohort. European Prospective Investigation of Cancer Br J Cancer 1999 80 95 103 10466767 Pedoe HT Clayton D Morris JN Brigden W McDonald L Coronary heart-attack in East London Lancet 1975 2 833 838 53328 Wild S McKeigue P Cross sectional analysis of mortality by country of birth BMJ 1997 314 705 710 9116545 British Cardiac Society, British Hyperlipidaemia Association, British Hypertension Society, British Diabetic Association Joint British recommendations on prevention of coronary heart disease in clinical practice: Summary BMJ 2000 320 705 707 10710586 Cappuccio FP Oakeshott P Strazzullo P Kerry SM Application of Framingham risk estimates to ethnic minorities in United Kingdom and implications for primary prevention of heart disease in general practice: Cross sectional population based study BMJ 2002 325 1 6 12098707 Feder G Crook AM Magee P Banerjee S Timmis AD Ethnic differences in invasive management of coronary disease: Prospective cohort study of patients undergoing angiography BMJ 2002 324 511 516 11872548 Bhopal RS Spectre of racism in health and healthcare: Lessons from history and the United States BMJ 1998 316 1970 1973 9641943 Whitty CJ Brunner EJ Shipley MJ Hemingway H Marmot MG Differences in biological risk factors for cardiovascular disease between three ethnic groups in the Whitehall II study Atherosclerosis 1998 142 279 286 RIFLE Research Group Menotti A Farchi G Seccareccia F The prediction of coronary heart disease mortality as a function of major risk factors in over 30,000 men in the Italian RIFLE pooling project. A comparison with the MRFIT primary screenees J Cardiovasc Risk 1994 1 263 270 7621307 McKeigue PM Shah B Marmot MG Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians Lancet 1991 337 382 386 1671422 Bhopal RS Unwin N White M Yallop J Walker L Heterogeneity of coronary heart disease risk factors in Indian, Pakistani, Bangladeshi and European origin populations: cross sectional study BMJ 1999 319 215 220 10417082 Sherwin R Kaelber CT Kezdi P Kjelsberg MO Thomas HE The Multiple Risk Factor Intervention Trial (MRFIT) II. The Development of the Protocol Prev Med 1981 10 402 425 7027236 Chaturvedi N McKeigue PM Methods of epidemiological surveys of ethnic minority groups J Epidemiol Community Health 1994 48 107 111 8189161 Home Office Race Relations (Amendment) Act 2000. New laws for a successful multi-racial Britain 2001 London Home Office Communication Directorate 7 63 National Institutes of Health Strategic research plan to reduce and ultimately eliminate health disparities. Washington (DC): United States Department of Health and Human Services. 47 p 2000 Available: http://www.nih.gov/about/hd/strategicplan.pdf . Accessed 29 November 2005	16379500	PMC1324792	CC BY	2021-01-05 11:13:50	no	PLoS Med. 2006 Mar 3; 3(3):e44	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030044	oa_comm
==== Front PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0030081SynopsisCardiology/Cardiac SurgeryEpidemiology/Public HealthHealth PolicyCardiovascular MedicineCohort studiesHealth PolicyPublic HealthResearch designCardiovascular Disease Research: Time to Focus on Minority Ethnic Groups Synopsis3 2006 3 1 2006 3 3 e81Copyright: © 2006 Public Library of Science.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Exclusion and Inclusion of Nonwhite Ethnic Minority Groups in 72 North American and European Cardiovascular Cohort Studies ==== Body Long-standing inequities exist in the amount of research on diseases that affect the developed world and those—such as leishmaniasis and schistosomiasis—that affect nonwhite populations who live in the developing world. These inequities have also been perceived to exist when white and nonwhite populations coexist in developed countries. Ethnic groups differ in their susceptibility to particular diseases. These differences can be genetic—the result of gene mutations that are more prevalent in some ethnic groups. They may also be due to social factors—in industrialized countries, for example, ethnic minorities are often poorer, less educated, and more frequently unemployed than their white counterparts. Studies in white populations, therefore, cannot necessarily be extrapolated to other ethnic groups. Cardiovascular disease is a major cause of death for all ethnic groups in developed countries, and the risk is especially high in those originating from South Asia. In the UK, for example, early deaths from coronary heart disease in Indians, Bangladeshis, Pakistanis, and Sri Lankans are about 50% higher than the national average. For Caribbeans and West Africans, on the other hand, the rates are much lower than average. Meghna Ranganathan and Raj Bhopal systematically reviewed the scientific literature over the past decades to assess the extent to which different ethnic minorities were included in cardiovascular cohort studies in Europe and North America. They identified 72 studies, 39 of which started after 1975 (at that time, it was well-known that different ethnic groups have different risk factors, and quite a bit was known about causes and control of the disease in white populations). Forty-one studies were conducted in Europe and 31 in North America, and one study involved participants from both continents. Overall, the researchers found, there is little information on cardiovascular research in ethnic populations—just ten studies compared white and nonwhite populations, and only five focused on one nonwhite ethnic group. All 15 of those were conducted in the United States. Despite the high risk of cardiovascular disease in ethnic minorities in Europe, not one European study so far has investigated the disease specifically in these populations. In general, it seems that issues of race or ethnicity have rarely been taken into account. Decisions about which ethnic groups to include were not often a part of the study design nor were they made explicit in the study report, and few articles gave details of the ethnic composition of the study population. Even when nonwhite participants were included in studies, there were often too few of them to allow for analysis by ethnicity. In some cases, researchers were open about their aim to study only white populations. In others, by selecting participants based on employment status or by conducting studies in rural settings, researchers were unlikely to include many participants from nonwhite ethnic groups that tend to cluster in cities. In cardiovascular disease, in particular, clear ethnic variations in risk mean that studying nonwhite populations is crucial if their health needs are not to be neglected, say the authors. The first step toward better understanding why some ethnic groups are more susceptible to disease is acknowledging the need for studying the question. Because such studies are expensive and challenging, say Ranganathan and Bhopal, they will only be done when there is a demand for them.	0	PMC1324793	CC BY	2021-01-05 10:40:37	no	PLoS Med. 2006 Mar 3; 3(3):e81	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030081	oa_comm
==== Front PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1637949710.1371/journal.pbio.0040022Research ArticleBiophysicsNeurosciencePhysiologyNeurology/NeurosurgeryRattus (Rat)Two-Photon Imaging of Cortical Surface Microvessels Reveals a Robust Redistribution in Blood Flow after Vascular Occlusion Flow in Cortical Surface ArteriolesSchaffer Chris B 1 2 Friedman Beth 3 4 Nishimura Nozomi 1 2 Schroeder Lee F 5 Tsai Philbert S 1 2 Ebner Ford F 6 Lyden Patrick D 3 4 5 Kleinfeld David [email protected] 1 2 5 1Department of Physics, University of California San Diego, La Jolla, California, United States of America2Center for Theoretical Biological Physics, University of California San Diego, La Jolla, California, United States of America3Department of Neurosciences, University of California San Diego, La Jolla, California, United States of America4Department of Neurology, Veterans Affairs Medical Center, San Diego, California, United States of America5Graduate Program in Neurosciences, University of California San Diego, La Jolla, California, United States of America6Department of Psychology, Vanderbilt University, Nashville, Tennessee, United States of AmericaCorbetta Maurizio Academic EditorWashington University School of MedicineUnited States of America2 2006 3 1 2006 3 1 2006 4 2 e228 8 2005 11 11 2005 Copyright: © 2006 Schaffer et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Redundancy in Cortical Surface Vessels Supports Persistent Blood Flow A highly interconnected network of arterioles overlies mammalian cortex to route blood to the cortical mantle. Here we test if this angioarchitecture can ensure that the supply of blood is redistributed after vascular occlusion. We use rodent parietal cortex as a model system and image the flow of red blood cells in individual microvessels. Changes in flow are quantified in response to photothrombotic occlusions to individual pial arterioles as well as to physical occlusions of the middle cerebral artery (MCA), the primary source of blood to this network. We observe that perfusion is rapidly reestablished at the first branch downstream from a photothrombotic occlusion through a reversal in flow in one vessel. More distal downstream arterioles also show reversals in flow. Further, occlusion of the MCA leads to reversals in flow through approximately half of the downstream but distant arterioles. Thus the cortical arteriolar network supports collateral flow that may mitigate the effects of vessel obstruction, as may occur secondary to neurovascular pathology. The authors quantify changes in blood flow in the pial arteriolar network of rodent cortex following targeted occlusions to individual vessels. ==== Body Introduction Normal brain function requires adequate levels of blood flow to ensure the delivery of oxygen and nutrients to cells and to facilitate the removal of metabolites and heat. The vasculature that supplies and regulates this flow is comprised of a succession of feeder vessels and networks with extensive redundant connections [1,2]. At the level of the supply to the brain, a pair of “communicating” arteries connects the carotid arteries to form the Circle of Willis. This loop is known to ensure a substantial level of fault tolerance to an occlusion of one of the member vessels [3]. At the level of cerebral cortex, the branches of each cerebral artery form the artery's own network of communicating arterioles on the surface of its cortical territory [4]. This network, in turn, gives rise to arterioles that plunge into the cortex and branch into the capillary bed. Each of the surface communicating networks is highly interconnected and, in addition, connects with the communicating networks of neighboring cerebral arteries through pial arteries called leptomeningeal anastomoses [5]. In analogy with the large-scale redundancy afforded by the Circle of Willis, the communicating network has been hypothesized to provide a robust source of collateral flow in the event of an occlusion of a microvessel [5–8]. Previous efforts to test this hypothesis, as well as the more general issue of the relationship between network topology and compensatory flow after an occlusion [3,9,10], have been hampered by a lack of methodology in which small vascular occlusions can be precisely targeted in time and space and in which flow can be quantified throughout multiple neighboring branches in a network. Here we address the issue of flow redistribution in a network of cortical communicating arterioles in rat parietal cortex that is supplied primarily by the middle cerebral artery (MCA). Our approach makes use of optical-based technologies. Two-photon laser scanning microscopy (TPLSM) [11,12] in conjunction with labeling of the blood plasma with fluorescein-dextran is used to form maps of the angioarchitecture as well as quantify the transport of individual red blood cells (RBCs) [13–15]. Occlusions are performed in two complementary ways. In the first method, occlusions to individual targeted microvessels at the pial surface are achieved through the introduction of a photosensitizer to the blood stream and the subsequent irradiation of the microvessel with focused laser light. This technique requires, on the one hand, suprathreshold illumination of a surface vessel to form a clot and, on the other hand, subthreshold illumination of deep and lateral vessels to avoid undesired clots. Fulfillment of these conditions represents an extension of the conventional application of the photothrombotic clot method [16–20] to allow for examination of dynamic flow changes in surface communicating arterioles after focal clot formation in a single surface arteriole. In the second method, an overall decrease in perfusion through the arterial source to parietal cortex is achieved by threading a filament along the carotid artery to block the base of the MCA [21]. This decrease necessarily leads to a change in the overall pressure balance throughout parietal cortex [22]. Results Mapping and Perturbing Blood Flow in Communicating Arteriole Networks Large-scale maps of the fluorescent brain vasculature in the field of a craniotomy (Materials and Methods) (Figure S1) reveal distal branches of the MCA as well as communicating and diving arterioles (Figure 1A). A high degree of redundancy in the surface arteriole network is apparent, with anastomoses formed between vessels from both the same and different MCA branches (Figure 1B). Candidate vessels for photothrombotic clotting are identified from the maps (Figure 1A) and traced back to a readily identifiable artery or vein; we considered only target vessels that are arterioles. High-resolution planar images (Figure 1C) are used to determine the baseline diameter of the target as well as of the vessels that lie upstream and downstream from the target. Line scans, in which the laser focus is repetitively scanned along the axis of each vessel (Figure 1C), are used to form a space–time image in which the non-fluorescent RBCs are represented as dark streaks (Figure 1D) [14,15,23,24]. The sign and magnitude of the slope of the streaks reflect the direction and speed, respectively, of RBC motion (Figure 1D, Figure S2, and Protocol S1). The analysis of this data yields a time series of the RBC speed in an arteriole (Figure 1E). We average over time, typically 40-s epochs, to suppress the underlying cardiac contributions (Figure 1E). We observe that the average speed increases with increasing vessel diameter (Figure S3), consistent with previous studies [25]. Figure 1 TPLSM of Fluorescently Labeled Cortical Vasculature In Vivo (A) Low-magnification TPLSM image of fluorescently labeled brain vasculature in rat parietal cortex. The axes indicate the rostral (R) and medial (M) directions. In the inset is an image of latex-filled brain vasculature taken from Scremin [1], with a box that indicates the approximate size and location of a typical craniotomy and an arrow that identifies the MCA. (B) Tracing of the surface arterial vascular network from the image in (A). Branches of the MCA are indicated, as are representative examples of the communicating arterioles (CA) that form the surface network and diving arterioles (DA) that supply cortex. (C) Maximal projection of a TPLSM image stack through a cortical arteriole. The dark line indicates the location where the line-scan data were taken, and the arrow represents the direction of flow obtained from these scans. (D) Line-scan data from the vessel in (C) to quantify the flow of RBCs. Each scan is displayed below the previous one, forming a space–time image with time increasing from top to bottom of the image. The dark streaks running from upper right to lower left are formed by the motion of the non-fluorescent RBCs. The RBC speed is given by the inverse of the slope of these streaks; the direction of flow is discerned from the sign of the slope. (E) RBC speed along the center of the arteriole shown in (C) and (D) as a function of time. The periodic modulation of the RBC speed occurs at the approximately 6-Hz heart rate. The dotted line represents the temporal average of the speed. (F) RBC speed in an arteriole, averaged over 40 s, as a function of the transverse position in the vessel along horizontal (y) and vertical (z) directions. The parabolic curve represents the laminar flow profile that most closely matches the data, i.e., s = A·(1 − r/R)2 where s is the speed of the RBCs, r is a radius from the origin and corresponds to either the y or z direction, R is the measured vessel radius of 26 μm, and A is a free parameter (A = 10 mm/s). The line-scan method is accurate for large vessels only if the speed varies slowly across the diameter of the vessel. We thus measured the spatial profile of the average speed across both the lateral (y) and axial (z) directions and observe a profile that peaks in the middle of the vessel and smoothly decreases toward the vessel wall (Figure 1F). The observed profile is flatter than that for laminar flow of a Newtonian fluid in a stiff pipe (Figure 1F), as expected for particle flow in a soft vessel [26]. The smooth nature of the spatial variation in speed implies that the extraction of speed from the line-scan measurements is relatively insensitive to misalignment between the scan and vessel axes so long as the measurements are performed close to the center of the vessel. Once the baseline measurements of vessel diameter and RBC speed were completed, the photosensitizer rose bengal was added to the blood stream (Methods). Green laser light, tightly focused on the target vessel (Figure 2A), was used to excite the dye (Figure 2B). We monitored the clot formation in real time, as illustrated for the example of flow into a trifurcating vessel in Figure 2 (Figure 2C; Video S1), and adjusted the laser power to near-threshold levels for clot formation. Once the clot was completed, the diameters of neighboring vessels and the speed of flow in these vessels was again measured (Figure 2D). Further, in control experiments the RBC speed decreased only slightly (5% per h) over 3 h (Figure S4), which far exceeds the duration of a set of measurements after a clot. Figure 2 Photothrombotic Clotting of Individual Targeted Surface Cortical Blood Vessels in Anesthetized Rat (A) Maximal projection of TPLSM image stack showing several surface arterioles (red A) and venules (blue V). The green circle indicates the region of the targeted arteriole that will be irradiated with green laser light. The white box indicates the region and orientation of the images in (C). (B) Schematic illustration of the targeted photothrombotic occlusion of a vessel and experiment timeline. After baseline imaging and blood flow measurements, rose bengal is intravenously injected into the animal. Green laser light is focused onto the wall of the target vessel, which excites the rose bengal and ultimately triggers the natural clotting cascade. Surface vessels adjacent to the target vessel are not occluded because they are not exposed to the 532-nm irradiation. (C) Planar TPLSM images of photothrombotic clotting of a surface arteriole. The frame on the left is taken at baseline. The green circle indicates the region of the targeted arteriole that will be irradiated, whereas the white arrows indicate the blood flow direction, as determined from line-scan measurements in the targeted vessel and in the vessels downstream from the target. The numbers over the downstream vessels correspond to the numbered line-scan data shown in (D). The streaked appearance of the vessels is due to the motion of RBCs during the acquisition of the image. The center frame is taken after an intravenous injection of rose bengal and 2-min irradiation with 0.5 mW of 532-nm laser light. The vessel is partially occluded (indicated by green double arrow). The right frame is taken after one more minute of irradiation. The target vessel is completely clotted (indicated by red X) whereas surrounding vessels are unaffected. Stalled blood flow is indicated by the dark mass of clotted cells in the target region and the brightly fluorescent region of blood plasma upstream from the target region. Note that blood flow is maintained in the branches downstream from the target vessel by a reversal in the direction of blood flow in the center branch, as determined from the line-scan data in (D). (D) Baseline and post-clot line-scan data for the numbered vessels downstream from the target vessel shown in (C). The average RBC speed determined from the line-scan data is indicated for each case, with a positive speed taken to be along the baseline direction of flow. Localization of Photothrombotic Effects A combination of in vivo flow measurements and post-mortem histology was used to determine the effects of focal irradiation of the target surface vessel on non-targeted surface vessels and the underlying cortical parenchyma. Surface vessels located lateral to the target vessel were never observed to clot, consistent with their exposure to insufficient laser power to trigger photothrombosis (Figure 2C). To reduce parenchymal damage deep to the irradiation site, the incident laser power on subsurface vessels was minimized by three physical mechanisms: (1) the absorption of incident light by the target vessel; (2) the divergence of the beam beyond the target vessel as a result of the strong focus; and (3) the scattering of light by brain tissue. Flow measurements were assessed in parenchymal capillaries situated less than 150 μm beneath the target vessel and within a lateral distance of 100 μm from the target vessel (55 capillaries, across six clots in four rats). In addition, flow was measured in control capillaries located at the same depth but 2 mm or more from the target vessel (18 capillaries). Flow was measured at baseline and after occlusion of the target vessel. The average flow speed across all capillaries was reduced to 57% ± 8% (mean ± standard error of the mean) of the baseline value and was highly variable (standard deviation = 64%). This reduction appeared to result from an increase in the normal occurrence of capillary stalls [14,27], from a baseline level of 8% to a post-clot level of 30%. Flow in control vessels was 84% ± 5% of the baseline value, consistent with slow rundown over the course of the experiment (Figure S4), and was less variable (standard deviation = 20%). In toto, these control data show that neighboring surface arterioles and venules remain unclotted after a target vessel is occluded. Further, clots in sub-surface capillaries are limited to vessels that lie directly beneath the target vessel, typically far from the downstream region perfused by the surface vessel. We assessed acute ischemic effects of photothrombosis across cell types with the hypoxia marker pimonidazole hydrochloride (Hypoxyprobe; Chemicon International, Temecula, California, United States). This probe forms an immobile adduct in viable but hypoxic tissue, that is visualized post hoc by antibody staining [28]. This marker was injected intravenously after clot formation and allowed to circulate for 1 h prior to sacrifice. As illustrated by the example of Figure 3A, the cortical distribution of Hypoxyprobe is highly sequestered in vessels and cells immediately below the photothrombotic clot (Figure 3A2 and 3A4) within a volume of ~500 pl (four clots across four animals). Additional analysis revealed an increased tendency for the retention of fluorescein-dextran in the extracellular space that borders vessels that lie directly beneath a photothrombotic clot (Figure 3A3). In control experiments, we observed that irradiation of the brain surface in a region between surface vessels also led to fluorescein-dextran retention in subsurface capillaries (Figure S5). These data imply that the pathology seen in capillaries beneath the target vessel is the result of photochemical damage, consistent with previous studies [29] and with our findings of increased capillary stalls beneath the target vessel. Critically, the territory of labeled Hypoxyprobe and fluorescein retention overlap (Figure 3A3). Thus there was no indication of wide-spread tissue hypoxia as a result of the focal clot formation in the surface arteriole. Figure 3 Assay of Oxidative Stress and Vascular Disruption Pimonidazole hydrochloride was introduced into the bloodstream at 1 h after targeted photothrombosis. At the end of this period, the animal was euthanized and transcardially perfused for brain tissue fixation. Labeling is illustrated for a closely spaced series of sections (within 150 μm). (A1–A4) The section, immunolabeled with an antibody against pimonidazole adducts (Hypoxyprobe), shows localization of adducts at uptake sites that are largely restricted to zones beneath the clot. Lateral and medial directions are labeled by M and L, respectively. The intermediate- and high-magnification views show immunolabeling across neural compartments, including parenchyma and blood vessels (A2 and A4). A mixed brightfield-fluorescent image shows that the fluorescein-dextran retained in the vessels and extravasated into the parenchyma overlaps the pimonidazole labeling (A3). (B1–B3) Immunolabeling of the platelet marker CD41 indicates sparse damage. A mixed image of fluorescein-dextran vascular retention and the CD41 immunoreactivity indicates that the damage is confined to just below the clot (B2). (C1–C3) Immunolabeling of tissue with the reactive astrocyte marker vimentin increased only marginally the labeling of vessels just below the clot (indicated by an asterisk). A mixed image of fluorescein-dextran retention and the vimentin immunoreactivity indicates that the damage is co-localized (C2). To probe ischemic cascades in hypoxic parenchymal vessels, we localized activated platelets by immunostaining with CD41 antibodies. Anti-CD41 immunoreactivity is also seen just below the clot (Figure 3B1 and 3B3) and overlaps with retention of fluorescein-dextran (Figure 3B2). To assay potential damage to other neural elements, we probed for reactive astrocytes by immunostaining for the filamentous astrocyte protein vimentin, as this is strongly up-regulated in reactive astrocytes that proliferate in response to injury [30]. Vimentin immunostaining is relatively evenly distributed in the cortex, most likely in glial endfeet on vessels, but shows only negligibly increased staining (Figure 3C1 and 3C3) among those ischemic vessels that retain fluorescein (Figure 3C2). Together these data indicate that photothrombosis results in a restricted zones of marked vascular ischemia (five clots across five animals). To selectively assess neuronal integrity in the vicinity of photothrombosis, we stained tissue sections with antibodies that recognize microtubule associated protein 2 (MAP2); MAP2 staining intensity is a sensitive and early indicator of neuropathology associated with ischemia [31]. As illustrated by the example in Figure 4, relevant tissue could be localized near a clotted vessel that remained intact (Figure 4A) and contained intra-luminal aggregates of nucleated blood cells (Figure 4B). The tissue near the clotted vessel exhibits near uniform staining for MAP2 (Figure 4A). Only subtle neuropathology is observed below the clot as indicated by a slight loss of staining for MAP2 in layer 1 (Figure 4A) and angulated neuronal somata with eccentric nucleus location and a corkscrew appearance in scattered dendrites (Figure 4C). In general, there was no indication of severe neuropathology in neurons that lay directly beneath the targeted vessel (seven clots across three rats) (Figure S6). Thus, we have achieved the desired goal of a localized clot to a single surface vessel with minimum collateral damage. Figure 4 Neuronal Integrity near Photothrombotic Clot The animal was sacrificed 1 h after the disruption, and the tissue was harvested from below the target vessel shown in Figure 5C. (A) The section, immunolabeled with an antibody against MAP2, shows widespread staining of neurons. (B) Fluorescent imaging of propidium iodide counterstain. The stained endothelial cells demarcate vessels. Note the numerous nuclei with segmented lobes characteristic of aggregates of leucocytes within the vessel (clot). (C) High-magnification view shows subtle neuropathology in some cells, i.e., corkscrew dendrites (single arrows) and shrunken neurons with eccentric nuclei (double arrow). Qualitative Aspects of Flow Rearrangement after a Single Arteriole Occlusion The essential consequence of an occlusion to a surface arteriole is illustrated by the example in Figure 2. We observe that blood flow downstream from the occlusion is maintained through a reversal in the direction of flow in a vessel at the first downstream branch (middle branch, Figure 2C and 2D). The reversal occurs within 1 s, i.e., within the acquisition time of an image, after the target vessel was occluded. As a population across all experiments (47 clots in 34 rats with arterioles that ranged from 15 to 140 μm in diameter), flow reversal occurred in one vessel at the first downstream branch after 100% of the clots. In a subset of targeted vessels (n = 7), the clot broke free midway through the experiment and had to be reformed. In each of these cases, the pattern of downstream flow observed after reclotting was the same as that observed after the first clot. Three additional examples illustrate typical flow changes in vessels upstream as well as downstream from an occlusion of a communicating arteriole (Figure 5; RBC speeds are shown for baseline and post-clotting conditions). In the circular architecture of Figure 5A, the reversal of flow direction at the first branch downstream from the clot is accompanied by a decrease in the speed of the RBCs. An anastomosis that connects to the same MCA branch as the targeted vessel is clearly the origin of this reversed flow. In the tree-like structure of Figure 5B, the reversal of flow direction at the first branch downstream is again accompanied by a decrease in RBC speed. A complicated pattern of reversed and non-reversed flow is observed in vessels farther downstream from the clot. In addition, flow is slowed in the upstream vessel that originally was a source to the targeted vessel (** in Figure 5B), whereas flow is essentially unchanged in a parallel vessel that shares the same source as the targeted vessel (* in Figure 5B). For the final example (Figure 5C), the targeted vessel was originally fed by the confluence of flow from two branches of the MCA. A photothrombotic clot results in a reversal in downstream flow, as well as a reversal in flow in one of the vessels that was a source of blood to the target vessel. The confluence of separate sources was observed to move to another point along the anastomosis between the MCA branches. Figure 5 Examples of Flow Changes that Result from Localized Occlusion of a Cortical Surface Arteriole (A–C) On the left and right are TPLSM images taken at baseline and after photothrombotic clotting of an individual vessel, respectively. Left center and right center are diagrams of the surface vasculature with RBC speeds (in mm/s) and directions indicated. The red X indicates the location of the clot, and vessels whose flow direction has reversed are indicated with red arrows and labels. In the examples of panels (A) and (B) we show maximal projections of image stacks whereas the example in panel (C) shows single TPLSM planar images; the streaks evident in the vessels in these latter frames are due to RBC motion, and the dashed box in the diagrams represents the area shown in the images. Quantitative Aspects of Flow Rearrangement after a Single Arteriole Occlusion We now consider the systematics of changes in RBC speed as well as potential changes in vessel diameter that resulted from occlusion of the target. Vessels in the neighborhood of the target vessel were classified according to their topological relationship to the target (Figure 6A): Upstream vessels (U) provided a source of blood to the target vessel; parallel vessels (P) shared the same source as the target; downstream vessels drained from the target vessel and are grouped as those immediately downstream (D1) and those two to four branches downstream (D2 to D4) from the target. We found that the RBC speed in upstream vessels slows substantially, yet the RBC speed in parallel vessels was increased slightly (Figure 6B, left). In two cases the target vessel was fed by a confluence of sources, and one of the upstream vessels reversed direction (as in Figure 5C). For vessels that lie at the first downstream branch from the target, the average RBC speed is reduced and half of the vessels reverse their flow direction (Figure 6B, middle). For vessels that lie farther downstream, the average speed is unchanged and the flow direction is reversed in approximately half the vessels (Figure 6B, right). Summary statistics are given in Table 1. Figure 6 Compendium of Flow Changes following Localized Photothrombotic Clotting of Communicating Surface Arterioles (A) Illustration of the four different classes of vessels considered, each delineated by their connectivity to the target vessel (indicated by an X). (B) Plots of post-clot RBC speed as a function of baseline RBC speed for each vessel class. The post-clot and baseline speeds were significantly correlated for the upstream and parallel vessels, but uncorrelated for the downstream vessels (Table 1). (C) Plots of post-clot vessel diameter as a function of baseline diameter. The pre- and post-clot diameters were correlated for all cases (Table 1). (D) Plots of post-clot volume blood flux as a function of the baseline value. The diagonal lines represent post-clot flux levels of 10% and 100% of baseline. As for the diameters, the pre- and post-clot fluxes were correlated for all cases (Table 1). Table 1 Summary Statistics for Changes in Vascular Parameters In contrast to the substantial changes in the speed of RBC flow in vessels in the neighborhood of the target vessel, only small changes are induced in the diameter of individual vessels as a result of clotting the target vessel (Figure 6C). On average, the diameter of all classes of vessels remained unchanged by an occlusion (see Table 1). Our measurements of RBC speed, denoted s, and vessel diameter, denoted d, may be combined to calculate the volume flux of blood, F, where under the approximation of laminar flow. We find that the flux is reduced by ~55%, relative to the baseline value, for upstream vessels and slightly increased for parallel vessels (Figure 6D, left, and Table 1). For vessels downstream from the clot, the post-clot flux ranges between 5% and 200% of the baseline value, with the flux reduced by approximately 40% at the first downstream branch (Figure 6D, center, and Table 1) and unchanged for vessels lying farther downstream (Figure 6D, right and Table 1). We estimate that random errors in the measurement of speed and diameter contribute a 10% uncertainty to the estimate of the flux for each vessel and that drift in the flux contributes an additional 10% uncertainty (see Figure S4 and Protocol S1). The observed spread in the baseline and post-clot values of the flux (Figure 6D and Table 1) is thus dominated by biological variability. Lastly, the changes in flux were correlated with changes in RBC speed for all vessels (p < 0.001), but the changes in flux were correlated with changes in vessel diameter only for the case of parallel vessels (p < 0.01). Flow Rearrangement after Occlusion of the MCA As a means to probe the possible ubiquity of reversals in flow in the arteriole communicating network, we studied the effects of a filament occlusion of the MCA on the flow in individual surface arterioles (Figure 7A). This artery is the primary supply to the surface network. The arterioles measured in response to MCA occlusion were, on average, of larger diameter than those measured in response to photothrombotic occlusion. Similar to the case of the localized occlusion (see Figure 2), we measured the speed of RBC flow and the diameter of vessels before and after occlusion of the MCA. The example data of Figure 7A, in which flow velocities are indicated in six of the surface arterioles, illustrate the essential results. The speed of RBCs was substantially reduced in five of these arterioles, consistent with a drop in the source perfusion. Critically, the direction of flow in four of the six arterioles was reversed after the occlusion. Figure 7 Quantitative Measurements of Flow Changes in Cortical Arterioles after Filament Occlusion of the MCA (A) Example of flow changes observed following MCA occlusion. Left: projection of a TPLSM image stack taken at baseline. Center: tracing of the surface arteriole network from the image with the baseline blood flow speed and direction indicated in some vessels. Right: blood flow speed and direction during MCA occlusion. Red arrows and speed labels indicate vessels whose direction has reversed. The axes indicate the rostral (R) and medial (M) directions. (B–D) RBC speed, vessel diameter, and volume blood flux, respectively, during MCA occlusion as a function of baseline values. The baseline and occlusion values of the diameter and flux are significantly correlated (p < 0.005), with r2 = 0.92 and 0.26, respectively; the baseline and occlusion values of the speed are not significantly correlated. As a population across all measurements (42 vessels across 11 rats), we observed an overall reduction in the magnitude of RBC speed and the presence of vessels with reversed as well as non-reversed flow in each experiment (Figure 7B). The diameter of the vessels was essentially unaffected by the occlusion (Figure 7C and Table 1), and there was no correlation between changes in the volume flux and vessel diameter for individual vessels. The volume flux was reduced by 45% after the occlusion (Figure 7D and Table 1). On average, approximately half the vessels reversed flow direction during the MCA occlusion. Discussion We examined the resilience of blood flow in a network of communicating arterioles that lies in the territory fed primarily by branches of the MCA (see Figure 1). Blood flow downstream from a targeted, localized occlusion does not stop, but rather is reestablished at the first downstream branch by a reversal in the direction of flow in one of the downstream branches (see Figures 2 and 5). Such flow reversals are common phenomena downstream from localized microvessel occlusions (see Figure 5). They are also a feature within this territory in response to occlusion of the MCA (Figure 7). The magnitude of the blood flow change following a localized occlusion in a surface arteriole depends on the topological relationship of a vessel to the clotted vessel (see Figure 6). The average flow at the upstream and the first downstream branch is reduced by less than half, whereas the average flow in vessels parallel to the clotted vessel and far downstream from the occlusion remain near baseline levels (Figure 6). In general terms, this study has revealed the persistent nature of perfusion at the level of cortical surface communicating arteriole networks (Figure 8). Figure 8 Summary of Quantitative Measurements of Changes in Volume Blood Flux in Response to Single Microvessel Occlusion and MCA Occlusion (A) Illustration showing topological relationship of vessels in (B) to (E) relative to the clotted arteriole (indicated with an X). (B–E) Histograms of the ratio of the post-clot flux to the baseline flux for vessels with different topological relationships to a photothrombotically clotted cortical surface arteriole: (B) upstream, (C) parallel, (D) first branch downstream, and (E) second to fourth branch downstream. (F) Histogram of the ratio of the flux measured during intra-luminal filament occlusion of the MCA to the baseline flux for cortical arterioles. In the interests of clarity, outliers with post-clot flux greater than twice the baseline flux were excluded from these histograms: one parallel vessel (ratio = 7), two D1 vessels (ratio = 2.5, 2.8), one D2 – 4 vessel (ratio = 11), and one filament occlusion vessel (ratio = 2.1). The arrows point to the mean values across all data points for each vessel class. Our observation that, on average, the diameters of vessels remained constant after localized microvessel occlusion (see Figure 6C and Table 1) is somewhat surprising. Complementary studies on the acute effects of mechanical occlusion on vessel diameter report that the cessation of flow in MCA tributaries, realized through multiple ligations to the surface branches [32], leads to an immediate, apparoximately 20% increase in vessel size. An even larger increase has been reported immediately following the occlusion of the common carotid arteries [33]. In contrast, a decrement in diameter of the MCA is seen after its occlusion by photothrombotic formation of a large, approximately 1.5 mm–long clot [34]. On the chronic timescale of 1 mo, multiple ligations to the surface branches of the MCA led to dilation to 200% of the initial vessel diameter [35]. In addition, other work has shown that cerebral arterioles, especially leptomeningeal anastomoses that connect the anterior cerebral with the MCA [5], show an approximately 50% increase in diameter 1 mo after MCA occlusion [36,37]. The dilation observed in previous acute and chronic studies has been attributed to active vascular regulation and/or to remodeling mechanisms. In contrast to these past results, we observe no systematic change in the diameter of downstream pial arterioles over a few hours in response to a small, approximately 100 μm–long clot formed in single surface arterioles by photothrombotic occlusion (Figure 6C and Table 1). One possible mechanism for this observed stability in vessel diameter may be that the new flow pattern is rapidly established, in about 1 s after the formation of an occluding clot (Video S1). Of interest, a classic means to demonstrate vascular auto-regulation in humans is based on modulation of the partial pressure of inspired CO2. However, recent work now shows that although cerebral blood flow undergoes predictable changes, the diameter of large cerebral arteries may be unchanged [38,39] or vary widely [40]. With regard to the magnitude of reperfusion in the vicinity of a localized clot, the observed flux was 60% of normal flow at the first downstream branch and 100% of normal flow at more distal downstream branches (Figures 6 and 8). The results of past studies suggest that flow rates must drop to 10% to 30% of baseline values [41–43] before irreversible neuropathology occurs. Such damage is neither expected nor observed to occur under the conditions of this study (see Figures 3 and 4), consistent with a past study of distal MCA occlusions [44]. In particular, we observed limited pathology directly below the clotted surface vessel, which was caused by photochemical damage to the capillaries, and no pathology in cortical regions that lie downstream from the site of clot formation (Figures 3 and 4). Thus the surface network of communicating arterioles appears to be well protected against single-point occlusions of surface arterioles by virtue of its architecture alone. The redundancy of brain vasculature changes as one proceeds from the level of the internal carotid arteries to that of the brain capillaries. At the level of the supply to the brain, the Circle of Willis provides sufficient redundancy so that all cephalic arteries are perfused in the event of an occlusion to one of the internal carotids. Blockages immediately downstream from the Circle of Willis, at the level of a cerebral artery, lead to a variety of neuropathologies. In particular, an occlusion to the base of the MCA typically results in infarction and widespread cell death in the basal ganglia whereas the same blockage produces less-severe deficits in the cerebral cortex or penumbral region [42,44]. Compensation at the level of cortex has been attributed to flow through leptomeningeal anastomoses that connect the territories of the intact anterior cerebral artery and the MCA [5]. The present work demonstrates that functional compensations can also occur at the scale of surface communicating arterioles, formed through anastomoses between branches of the MCA (Figure 1A and 1B), since perfusion is maintained in neighboring branches after photothrombosis of a single arteriole (see Figures 5, 6, and 8). This failsafe mechanism most likely results from the many short-range loops in this arterial network, which may also play a role in rerouting a fixed supply of blood among different cortical columns during shifts in cortical electrical activity [45]. The highly redundant cortical surface vasculature discussed here for rats is also present in humans [4,46]. In contrast to the present results for the surface vasculature of cortex, the apparent lack of a similar system of anastomoses in the basal ganglia may contribute to its greater vulnerability after a blockage of the MCA [47]. The present work focused on the surface network of communicating arterioles (Figure 1B, CA), a two-dimensional network formed by anastomoses between branches of a cerebral artery. This network delivers blood to the cortical parenchyma through a series of diving arterioles (Figure 1B, DA). The topology and resilience of the three-dimensional, subcortical microvascular network remain an open issue. From a technical perspective, the present methodology to form localized occlusions is ideal for surface vessels but cannot selectively target vessels at depth because, as for all techniques that rely on the absorption of a single photon, all vessels superficial to the target will also be affected. Thus a technology that depends on the nonlinear absorption of light is required. Such a technique, which makes use of high-fluence ultrashort pulses of light that can precisely ablate tissue [48], has been demonstrated to work with vessels in superficial cortical layers [49]. This method cannot, in turn, be used to target clots in the surface communicating arteriole network, as the lack of containment by parenchymal tissue causes irradiated vessels to burst. Preliminary data suggest that clots to subsurface microvessels leads to perfusion failure in the immediate downstream vessels. This is in marked contrast to the redistribution of flow that preserves perfusion in all downstream branches of the surface communicating arteriole network (see Figures 6, 8D, and 8E). In humans, damage to microvessels is a known pathological condition [50–54]. In particular, occlusion of small-scale arterioles is a likely cause of clinically silent lacunar infarcts [55] that are correlated with an increased risk of dementia and cognitive decline [56–58]. It is thus interesting that the Rotterdam Scan study [57], which identified clinically silent lacunar infarcts through magnetic resonant imaging, found that few cortical infarcts were located near the surface and thus where the vasculature appears to be most redundant [46]. Our results for surface blood flow dynamics (see Figures 6, 7, and 8) suggest an emerging relation between vascular topology and susceptibility to stroke in different regions of the brain. Materials and Methods Surgery Our subjects were 51 Sprague-Dawley rats. In 34 animals of both sexes, 150 to 350 g in mass, one to five individual microvessels were occluded photothrombotically. In 11 male animals, 290 to 310 g in mass, the MCA was occluded by the filament method. The remaining six rats were used for control experiments. All animals were anesthetized by interperitoneal injection of urethane (150 mg per 100-g rat), supplemented as necessary; urethane is reported to maintain autoregulation of cerebral blood flow [59]. Atropine sulfate was delivered by subcutaneous injection (5 μg per 100-g rat) at the start of surgery and supplemented hourly at a reduced dose (1 μg per 100-g rat) to reduce secretions. We further supplemented by subcutaneous injection of 5% (w/v) glucose in phosphate buffered saline (PBS) (0.5 ml per 100-g rat) every hour. Body temperature was maintained at 37.5 °C and heart rate was continuously monitored. A 4-mm by 6-mm craniotomy was prepared over parietal cortex, with the center at medial-lateral equal to 4.5 mm and anterior-posterior equal to −4.5 mm relative to Bregma. The dura was removed, a chamber consisting of a metal frame and a removable coverglass (no. 1) lid was glued to the skull [60], and the space between the exposed brain surface and the coverglass was filled with 1.5% (w/v) low–melting point agarose in an artificial cerebro-spinal fluid [61]. A 0.3-ml bolus of a 5% (w/v) solution of 2 MDa fluorescein-conjugated dextran in PBS was injected into the tail vein to label the blood plasma. The care and experimental manipulation of our animals have been reviewed and approved by the Institutional Animal Care and Use Committee at the University of California, San Diego. TPLSM Images were obtained with a two-photon laser scanning microscope of local design [62] that was modified to include a path for a green laser beam (see Figure S1 and Protocol S1). A 0.12 numerical aperture (NA), 5×-magnification air objective was used to obtain images of the surface vasculature across the entire cranial window to aid in navigating around the cortical vasculature. We changed to a 0.8-NA, 40×-magnification water-immersion objective for high-resolution imaging, line-scan measurements, and photothrombotic clotting with the green laser. The line-scan rate was 1.3 kHz for measuring RBC speed, and we typically acquired line-scans for 40 s and report the average speed over this period. Photothrombotic clotting A continuous wave green-light laser (λ = 532 nm) (TIM-622; Transverse Industries, Taipei Hsien, Taiwan) was directed onto the beam axis of the microscope with a dichroic mirror (see Figure S1). The green-light beam underfilled the back aperture of the objective and was aligned so it focused at the center of the same plane as the near-infrared pulsed laser beam and formed an approximately 5-μm diameter spot. As the green-light overlapped with the fluorescent spectrum of fluorescein, the green-light was delivered in pulses of 1-s duration that were interspersed with TPLSM imaging. Vessels targeted for clotting were centered in the imaging field (see Figure 2A). The rat was then given a 0.3-ml intravenous injection of 1% (w/v) rose bengal (Na salt) in PBS, and the wall of the target vessel was irradiated with 0.1 to 5 mW of green laser light for a total of 30 to 600 s. Irradiation of a photosensitizer leads to the production of singlet oxygen [63], which damages the wall of the vessel and subsequently triggers a clotting cascade that leads to an occlusion [16,17]. We slowly increased the green laser power while monitoring clot formation in near real time and used the minimum power required to trigger clot formation in the target vessel. This procedure led to the formation of a single, localized clot (see Figure 2C and Video S1). In control experiments, green laser irradiation at a typical power led to no visible effects on the target vessel in the absence of rose bengal (Figure S7) and no observable retention of fluorescein-dextran in the tissue below the target vessel (see Figure S5). Filament occlusion of MCA A filament is advanced inside the internal carotid artery until it blocks the origin of the MCA, as described [21]. Briefly, an incision is made in the neck, exposing the left common carotid artery. The external carotid and pterygopalatine arteries are ligated with 4–0 silk suture. An incision is made in the wall of the common carotid artery and a 4–0 nylon suture, whose tip has been blunted by heating on a microforge, is advanced 17.5 mm from the bifurcation point of the external and internal carotid arteries, thereby blocking the ostium of the MCA. Flow velocities are measured at baseline and during the MCA occlusion. A large-scale TPLSM image (see Figure 7A) is used to ensure the same vessels are measured at baseline and during the occlusion. The occlusion measurements are performed within two hours of the insertion of the filament. Postmortem histology The cortex of seven of the animals that received localized photothrombotic clots, four with pimonidazole hydrochloride (Hypoxyprobe-1 [90201]; Chemicon) injected 1 h before sacrifice, were perfused transcardially with 100 ml of PBS, followed by 100 ml of 4% (w/v) paraformaldehyde in PBS. Fiducial marks were made in the corners of the craniotomy by passing −20 μA though a single tungsten electrode that translated at 2 mm/s. The brain was cryoprotected with sucrose and then 50 μm–thick sections were cut in a coronal plane on a freezing-sliding microtome. Sections near the location of the clots were selected based on the location of the targeted vessels relative to the fiducial marks, as determined from widefield images of the craniotomy taken before and after the perfusion and the retention of fluorescein-dextran in capillaries beneath the target vessel (see Figure 3A3). To confirm the completeness of a clot in the surface vessels, we looked for intra-luminal aggregates of blood-borne white cells by staining sections with propidium iodide (1 μg/ml) mixed into aqueous mountant (see Figure 4B). Immunohistochemistry Sections lying beneath photothrombotic occlusions were mounted onto Fisher Superfrost Plus slides (Pittsburgh, Pennsylvania, United States) and underwent antigen retrieval in 10 mM citrate buffer (pH 6.0) heated to boiling in a microwave oven (300 s at full power). The slides were incubated overnight in one of four monoclonal antibodies in diluent with PBS and 0.2% (v/v) Triton X–100 followed by incubation with a biotinylated anti-mouse secondary antibody: (1) Hypoxyprobe antibody (90204; Chemicon); (2) MAP2 antibody (M1406; Sigma, St. Louis, Missouri, United States); (3) CD41 antibody (MAB1207; Chemicon); and (4) vimentin antibody (MAB3400; Chemicon). In all cases, bound antibody was visualized with the Vector ABC Kit (Vector Laboratories, Burlingame, California, United States) using diaminobenzadine as the chromagen. The sections were cover-slipped with Prolong mountant (Molecular Probes, Eugene, Oregon, United States). Supporting Information Figure S1 Schematic of the Two-Photon Laser Scanning Microscope (TPLSM) Our realization is based on an 800-nm fs (femtosecond) laser with an integrated continuous-wave (CW) 532-nm laser for photothrombotic clotting. The CW laser, attenuated using neutral density (ND) filters, is directed onto the beam axis of the TPLSM with a dichroic mirror (600 nm–long pass, dichroic 1). An approximately 2-mm hole was etched in the coating of the TPLSM dichroic (dichroic 2) to allow transmission of the green laser beam. The 532-nm laser was aligned so it focused in the same plane as and at the center of the TPLSM image. The rat is bolted, via a metal head frame affixed to the skull, onto a two-dimensional translation stage that allows precise positioning of the rat relative to the TPLSM field of view and the CW laser focus. HWP, half-wave plate; PMT, photomultiplier tube. (141 KB PDF). Click here for additional data file. Figure S2 Illustration of Automated Algorithm for Finding Slope of Streaks Formed by Moving RBCs in TPLSM Line-Scan Data The data are for the same vessel shown in Figure 1C, 1D, and 1E, although not from the same time point as Figure 1D. (A) Line-scan data from an epoch in time are transformed to a square matrix with normalized axes. In the left image, an abrupt change in the slope of the streaks due to a heartbeat is indicated. (B) The central region of the square matrix is rotated, and we search for the angle that yields horizontal streaks, as in the middle panel. (C) Separability of line-scan data as a function of rotation angle; separability is maximal for vertical or horizontal streaks (Protocol S1). The rotation angle corresponding to horizontal streaks is chosen, yielding the RBC speed and direction, in this case: 11.9 mm/s and a flow direction of right to left. (262 KB PDF). Click here for additional data file. Figure S3 Baseline Measurements of the Time-Averaged RBC Speed at the Center of a Vessel as a Function of the Diameter of the Vessel The data include all arterioles in this study. Arterioles measured as part of the photothrombotic (rose bengal) and MCA (filament) occlusion studies are indicated separately. The line represents a best-fit linear regression to the data, and shows a statistically significant correlation, valid for diameters between 10 and 130 μm, between speed, S, and diameter, D, given by S = (4.9 mm/s) + (124/s) × D (p < 0.0001). (765 KB PDF). Click here for additional data file. Figure S4 Normalized Volume Blood Flux in Five Arterioles as a Function of Time For each vessel, each measurement of the flux was normalized to the average over all measurements for that vessel. The five vessels varied substantially in their average volume flux: 0.17 (circles), 0.013 (diamonds), 0.0048 (squares), and 0.0021 μl/s (triangles). The inset shows the histogram of the normalized flux for all vessels, after the long-term decrease in flux (~5% per h) is removed. (710 KB PDF). Click here for additional data file. Figure S5 Epi-Fluorescence Images Overlaid on Wide-Field Images Showing Vascular Retention of Circulating Fluorescein-Dextran (A) The brain surface was irradiated with 1 mW of 532-nm irradiation for 1 min after an intravenous injection of rose bengal. The laser focus was deliberately located in a region where there were no surface vessels; therefore no surface vessel was clotted. The vascular retention and parenchymal extravasagation of the fluorescein-conjugated dextran used for in vivo imaging is somewhat more extensive than that observed in Figure 3. This is likely because a surface target vessel was not present to absorb and scatter the incident laser light, leading to a higher fluence incident on the sub-surface capillaries, and increasing the extent of the photochemical damage. (B) A surface vessel was irradiated at 1 mW for 1 min without any rose bengal present. No retention of the fluorescein-dextran is evident in the sub-surface capillaries, indicating no photochemical damage. (2.4 MB PDF). Click here for additional data file. Figure S6 MAP2 Immunohistology of Coronal Brain Slices beneath a Photothrombotic Clot of a Surface Arteriole (Ischemic Side) and from the Corresponding Location on the Contralateral Side (A, B, E, and F) Clotted side. (C, D, G, and H) Contralateral side. The boxes in panels (A), (C), (E), and (G) indicate the locations of the images in panels (B), (D), (F), and (H), respectively. The arrows in (B) and (F) indicate cells showing minor neuropathology (single arrow indicates cork-screw dendrites; double arrow indicates shrunken cells with eccentric nuclei. Most cells in panels (B) and (F) exhibit no pathology. The example on the left is the same as that shown, in part, in Figure 3. In the example on the right, some trapped RBCs are visible in capillaries beneath the photothrombotically clotted vessel (E), although the capillaries were still flowing after clot formation, based on in vivo TPLSM. (6.8 MB PDF). Click here for additional data file. Figure S7 Control Experiment Showing that Photothrombotic Clot Formation Requires Both Rose Bengal and Green Laser Irradiation (A) Baseline TPLSM image of same vessel shown in Figure 2. The green circle indicates the region that will be irradiated with 532-nm light. (B) After 2 min irradiation with 0.5 mW of 532-nm laser light before intravenous injection of rose bengal. No clot formation is evident. (C) After an additional 2-min irradiation after intravenous injection of rose bengal. Forming clot indicated by green arrow. (386 KB PDF). Click here for additional data file. Protocol S1 TPLSM for Photothrombotic Clotting and Characterization of RBC Flux A locally designed two-photon laser scanning microscope was modified to allow real-time photothrombotic clotting of individual cortical surface blood vessels. TPLSM line-scans were used to collect space–time data of RBC motion in individual cortical blood vessels, and an algorithm based on singular value decomposition was used to find the slope of the streaks formed by moving RBCs in this space–time data, thereby determining the RBC speed. Control experiments established that systematic flux changes in vessels due to intravenous injections or to shedding of clot material during the formation of a localized photothrombotic clot are less than 10%. (419 KB PDF). Click here for additional data file. Video S1 Two-Photon Fluorescence Image Sequence Showing the Formation of a Localized Photothrombotic Clot in a Surface Arteriole This is the same example shown in Figure 2. Images are displayed at a rate that is sped up by a factor of 15. The fluctuating, streaked appearance of the vessels is due to the motion of RBCs, and indicates flow. The initial direction of flow in the targeted arteriole (vertically centered in image) is right to left, branching into three vessels on the left of the frame. A saturated white strip at the top of the frame indicates irradiation with 523-nm light. Clot material is formed just downstream from the irradiated region of the vessel, and some sheds off and is carried away down one of the downstream branches. Note that after a complete occlusion is formed, the streaked appearance is maintained in the downstream branches, indicating they are still flowing. (7.3 MB MOV). Click here for additional data file. We thank Pablo Blinder, Eshel Ben-Jacob, Herbert Levine, and Stephen Segal for useful discussions, Earl Dolnick for assistance with the electronics, Donald Pizzo and Leon Thal for use of their photomicroscope, and Coherent, Inc., for the loan of equipment. This work was funded by the David and Lucile Packard Foundation (DK), the Veteran's Affairs Medical Research Department (PDL), the National Institute of Health grants NS/041096 (DK), EB/003832 (DK), NS/025907 (FFE), and NS/043300 (PDL), a La Jolla Interfaces in Science Postdoctoral Fellowship (CBS), and the National Science Foundation Graduate Fellowship Program (NN). Competing interests. The authors have declared that no competing interests exist. Author contributions. CBS, BF, NN, LFS, PDL, and DK conceived and designed the experiments. CBS, BF, NN, LFS, PST, and FFE performed the experiments. CBS, BF, NN, PDL, and DK analyzed the data. PDL and DK contributed reagents/materials/analysis tools. CBS, BF, and DK wrote the paper. Citation: Schaffer CB, Friedman B, Nishimura N, Schroeder LF, Tsai PS, et al. (2006) Two-photon imaging of cortical surface microvessels reveals a robust redistribution in blood flow after vascular occlusion. PLoS Biol 4(2): e22. Abbreviations MAP2microtubule associated protein 2 MCAmiddle cerebral artery NAnumerical aperture RBCred blood cell TPLSMtwo-photon laser scanning microscopy ==== Refs References Scremin OU Paxinos G Cerebral vascular system The rat nervous system 1995 San Diego (California) Academic Press 3 35 Mchedlishvili G Arterial behavior and blood circulation in the brain 1963 New York Consultants Bureau 338 Liebeskind DS Collateral circulation Stroke 2003 34 2279 2284 12881609 Nonaka H Akima M Nagayama T Hatori T Zhang Z Microvasculature of the human cerebral meninges Neuropathology 2003 23 129 135 12777101 Brozici M van der Zwain A Hillen B Anatomy and functionality of leptomeningeal anastomoses: A review Stroke 2003 34 2750 2762 14576375 van der Eecken HM Adams RD The anatomy and functional significance of the meningeal arterial anastomoses of the human brain J Neuropathol Exp Neurol 1953 12 132 157 13053234 Moody DM Bell MA Challa VR Features of the cerebral vascular pattern that predict vulnerability to perfusion or oxygenation deficiency: An anatomic study AJNR Am J Neuroradiol 1990 11 431 439 2112304 Wei L Rovainen CM Woolsey TA Ministrokes in rat barrel cortex Stroke 1995 26 1459 1462 7631353 Lo EH Dalkara T Moskowitz MA Mechanisms, challenges and opportunities in stroke Nat Rev Neurosci 2003 4 399 415 12728267 Iadecola C Neurovascular regulation in the normal brain and in Alzheimer's disease Nat Rev Neurosci 2004 5 347 360 15100718 Zipfel WR Williams RM Webb WW Nonlinear magic: Multiphoton microscopy in the biosciences Nat Biotechnol 2003 21 1369 1377 14595365 Denk W Strickler JH Webb WW Two-photon laser scanning fluorescence microscopy Science 1990 248 73 76 2321027 Zhang S Boyd J Delaney KR Murphy TH Rapid reversible changes in dendritic spine structure in vivo gated by the degree of ischemia J Neurosci 2005 25 5333 5228 15930381 Kleinfeld D Mitra PP Helmchen F Denk W Fluctuations and stimulus-induced changes in blood flow observed in individual capillaries in layers 2 through 4 of rat neocortex Proc Natl Acad Sci U S A 1998 95 15741 15746 9861040 Chaigneau E Oheim M Audinat E Charpak S Two-photon imaging of capillary blood flow in olfactory bulb glomeruli Proc Natl Acad Sci U S A 2003 100 13081 13086 14569029 Krammer B Vascular effects of photodynamic therapy Anticancer Res 2001 21 4271 4277 11908681 Watson BD Dietrich WD Busto R Wachtel MS Ginsberg MD Induction of reproducible brain infarction by photochemically initiated thrombosis Ann Neurol 1985 17 497 504 4004172 Nakase H Kakizaki T Miyamoto K Hiramatsu K Sakaki T Use of local cerebral blood flow monitoring to predict brain damage after disturbance to the venous circulation: Cortical vein occlusion model by photochemical dye Neurosurgery 1995 37 280 285 7477780 Markgraf CG Kraydieh S Prado R Watson BD Dietrich WD Comparative histopathologic consequences of photothrombotic occlusion of the distal middle cerebral artery in Sprague-Dawley and Wistar rats Stroke 1993 24 286 293 8421830 Ishikawa M Sekizuka E Oshio C Sato S Yamaguchi N Platelet adhesion and arteriolar dilation in the photothrombosis: Observation with the rat closed cranial and spinal windows J Neurol Sci 2002 194 59 69 11809168 Zea Longa E Weinstein PR Carlson S Cummins R Reversible middle cerebral artery occlusion without craniectomy in rats Stroke 1989 20 84 91 2643202 Heiss WD Graf R Lottgen J Ohta K Fujita T Repeat positron emission tomographic studies in transient middle cerebral artery occlusion in cats: Residual perfusion and efficacy of postischemic reperfusion J Cereb Blood Flow Metab 1997 17 388 400 9143221 Dirnagl U Villringer A Einhaupl KM In-vivo confocal scanning laser microscopy of the cerebral microcirculation J Microsc 1992 165 147 157 1552568 Hirase H Creso J Buzsaki G Capillary level imaging of local cerebral blood flow in bicuculline-induced epileptic foci Neuroscience 2004 128 209 216 15450368 Fung YC Biodynamics: Circulation 1984 New York Springer-Verlag 404 Pries AR Secomb TW Gaehtgens P Biophysical aspects of blood flow in the microvasculature Cardiovasc Res 1996 32 654 667 8915184 Villringer A Them A Lindauer U Einhaupl K Dirnagl U Capillary perfusion of the rat brain cortex: An in vivo confocal microscopy study Circ Res 1994 75 55 62 8013082 Chapman JD Franko AJ Sharplin J A marker for hypoxic cells in tumours with potential clinical applicability Br J Cancer 1981 43 546 550 7236492 Haseldonckx M van Bedaf D van de Ven M van Reempts J Borgers M Vasogenic oedema and brain infarction in an experimental penumbra model Acta Neurochir Suppl 2000 76 105 109 11449986 Latov N Nilaver G Zimmerman EA Johnson WG Silverman AJ Fibrillary astrocytes proliferate in response to brain injury Dev Biol 1979 72 381 384 389711 Dawson DA Hallenbeck JM Acute focal ischemia-induced alterations in MAP2 immunostaining: Description of temporal changes and utilization as a marker for volumetric assessment of acute brain injury J Cereb Blood Flow Metab 1996 16 170 174 8530550 Wei L Craven K Erinjeri J Liang GE Bereczki D Local cerebral blood flow during the first hour following acute ligation of multiple arterioles in rat whisker barrel cortex Neurobiol Dis 1998 5 142 150 9848087 Morita Y Fukuuchi Y Koto A Suzuki N Isozumi K Rapid changes in pial arterial diameter and cerebral blood flow caused by ipsilateral carotid artery occlusion in rats Keio J Med 1997 46 120 127 9339639 Watson BD Prado R Veloso A Brunschwig JP Dietrich WD Cerebral blood flow restoration and reperfusion injury after ultraviolet laser–facilitated middle cerebral artery recanalization in rat thrombotic stroke Stroke 2002 33 428 434 11823647 Wei L Erinjeri JP Rovainen CM Woolsey TA Collateral growth and angiogenesis around cortical stroke Stroke 2001 32 2179 2184 11546914 Coyle P Diameter and length changes in the cerebral collaterals after middle cerebral artery occlusion in the young rat Anat Rec 1984 210 357 364 6507901 Greenberg DA Angiogenesis and stroke Drug News Perspect 1998 11 265 270 15616645 Serrador JM Picot PA Rutt BK Shoemaker JK Bondar RL MRI measures of middle cerebral artery diameter in conscious humans during simulated orthostasis Stroke 2000 31 1672 1678 10884472 Djurberg HG Seed RF Evans DA Brohi FA Pyper DL Lack of effect of CO2 on cerebral arterial diameter in man J Clin Anesth 1998 10 646 651 9873965 Valdueza JM Draganski B Hoffmann O Dirnagl U Einhaupl KM Analysis of CO2 vasomotor reactivity and vessel diameter changes by simultaneous venous and arterial Doppler recordings Stroke 1999 30 81 86 9880393 Baron JC Perfusion thresholds in human cerebral ischemia: Historical perspective and therapeutic implications Cerebrovasc Dis 2001 11 Suppl 1 2 8 Hossmann KA Viability thresholds and the penumbra of focal ischemia Ann Neurol 1994 36 557 565 7944288 Zhao W Belayev L Ginsberg MD Transient middle cerebral artery occlusion by intraluminal suture: II. Neurological deficits, and pixel-based correlation of histopathology with local blood flow and glucose utilization J Cereb Blood Flow Metab 1997 17 1281 1290 9397027 Bederson JB Pitts LH Tsuji M Nishimura MC Davis RL Rat middle cerebral artery occlusion: Evaluation of the model and development of a neurologic examination Stroke 1986 17 472 476 3715945 Cox SB Woolsey TA Rovainen CM Localized dynamic changes in cortical blood flow with whisker stimulation corresponds to matched vascular and neuronal architecture of rat barrels J Cereb Blood Flow Metab 1993 13 899 913 8408316 Reina-de la Torre F Rodriguez-Baeza A Sahuquillo-Barris J Morphological characteristics and distribution pattern of the arterial vessels in human cerebral cortex: A scanning electron microscope study Anat Rec 1998 251 87 96 9605225 Feekes JA Hsu S-W Chaloupka JC Cassell MD Tertiary microvascular territories define lacunar infarcts in the basal ganglia Ann Neur 2005 58 18 30 Tsai PS Friedman B Squier JA Kleinfeld D Ultrashort pulsed laser light: A cool tool for ultraprecise cutting of tissue and cells Opt Photonic News 2004 14 24 29 Nishimura N Schaffer CB Friedman B Tsai PS Lyden PD Targeted insult to individual subsurface cortical blood vessels using ultrashort laser pulses: Three models of stroke Nat Methods 2006 In press Lammie GA Pathology of small vessel stroke Br Med Bull 2000 56 296 306 11092081 Wardlaw JM Sandercock PA Dennis MS Starr J Is breakdown of the blood-brain barrier responsible for lacunar stroke, leukoaraiosis, and dementia? Stroke 2003 34 806 812 Arauz A Murillo L Cantu C Barinagarrementeria F Higuera J Prospective study of single and multiple lacunar infarcts using magnetic resonance imaging: Risk factors, recurrence, and outcome in 175 consecutive cases Stroke 2003 2453 2458 14500936 Lee SH Bae HJ Kwon SJ Kim H Kim YH Cerebral microbleeds are regionally associated with intracerebral hemorrhage Neurology 2004 62 72 76 14718700 Fazekas F Kleinert R Roob G Kleinert G Kapeller P Histopathologic analysis of foci of signal loss on gradient-echo T2*-weighted MR images in patients with spontaneous intracerebral hemorrhage: Evidence of microangiopathy-related microbleeds AJNR Am J Neuroradiol 1999 20 637 642 10319975 Fisher CM The arterial lesions underlying lacunes Acta Neuropathol (Berl) 1969 12 1 15 Vermeer SE Prins ND den Heijer T Hofman A Koudstaal PJ Silent brain infarcts and the risk of dementia and cognitive decline N Engl J Med 2003 348 1215 1222 12660385 Vermeer SE Koudstaal PJ Oudkerk M Hofman A Breteler MMB Prevalence and risk factors of silent brain infarcts in the population-based Rotterdam Scan Study Stroke 2002 33 21 25 11779883 O'Brien JT Erkinjuntti T Reisberg B Roman G Sawada T Vascular cognitive impairment Lancet Neurol 2003 2 89 98 12849265 Tuor UI Farrar JK Pial vessel caliber and cerebral blood flow during hemorrhage and hypercapnia in the rabbit Am J Physiol Heart Circ Physiol 1984 247 H40 H51 Kleinfeld D Denk W Yuste R Lanni F Konnerth A Two-photon imaging of neocortical microcirculation Imaging neurons: A laboratory manual 2000 Cold Spring Harbor (New York) Cold Spring Harbor Laboratory Press 23.1 23.15 Kleinfeld D Delaney KR Distributed representation of vibrissa movement in the upper layers of somatosensory cortex revealed with voltage sensitive dyes J Comp Neurol 1996 375 89 108 8913895 Tsai PS Nishimura N Yoder EJ Dolnick EM White GA Frostig RD Principles, design, and construction of a two photon laser scanning microscope for in vitro and in vivo brain imaging In vivo optical imaging of brain function 2002 Boca Raton (Florida) CRC Press 113 171 Pooler JP Valenzeno DP Dye-sensitized photodynamic inactivation of cells Med Phys 1981 8 614 628 7027011	16379497	PMC1324794	CC BY	2021-01-05 08:21:47	no	PLoS Biol. 2006 Feb 3; 4(2):e22	utf-8	PLoS Biol	2,006	10.1371/journal.pbio.0040022	oa_comm
==== Front PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 2007652910.1371/journal.pbio.0040043SynopsisBiophysicsNeuroscienceNeurology/NeurosurgeryRattus (Rat)Redundancy in Cortical Surface Vessels Supports Persistent Blood Flow SynopsisGross Liza 2 2006 3 1 2006 3 1 2006 4 2 e43Copyright: © 2006 Public Library of Science.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Two-Photon Imaging of Cortical Surface Microvessels Reveals a Robust Redistribution in Blood Flow after Vascular Occlusion ==== Body While examining the brain of a man who had died of a mesenteric tumor, Sir Thomas Willis was shocked to find a nearly totally blocked right carotid artery, the brain's main blood source. In his 1664 book, Anatomy of the Brain, the Oxford physician and neuroanatomist reported that given the “influx of blood being denied by this route, it seemed remarkable that this person had not previously died of an apoplexy.” A stroke was averted, Willis explained, because the circular arrangement of the arteries on the base of the brain—later named the circle of Willis—created redundant connections in the circulatory network. Scientists are still elucidating the details of vascular resilience—and how stroke can result when it is compromised. Much as the circle of Willis protects large-scale blood flow, it's been a matter of conjecture if a similar mechanism ensures microvessel flow to local brain regions. Until recently, scientists have lacked the appropriate methods to test this hypothesis. In a new study, Chris Schaffer, David Kleinfeld, and their physics and medical school colleagues take advantage of cutting-edge techniques in microscopy and laser-induced clotting (photothrombosis) to directly measure the resilience and dynamics of cortical blood flow. By manipulating and monitoring blood flow in the surface vasculature of rat parietal cortex, they show that the arteriole network rapidly reestablishes blood flow following a targeted occlusion to a surface vessel. The surface arteriole communicating network in the rat To analyze flow dynamics among the interconnected arterioles on the cortical surface, Schaffer et al. labeled blood plasma with a dye that allowed them to map the vessel architecture and also measure the flow of individual blood cells, using two-photon microscopy. Candidate vessels were selected for photothrombotic clotting, a technique that uses photosensitive molecules and focal laser beams to produce free radicals and damage the vessel wall, ultimately triggering an occluding blood clot. This enabled the authors to pinpoint a single vessel for occlusion without harming any neighboring vessels. They further used two-photon microscopy to determine the direction and speed of red blood cells by repetitively scanning the axis of each vessel, which allowed them to monitor the pattern of blood flow throughout the surface vascular network in real time. Schaffer et al. first tested the brain's vascular resilience by inducing targeted photothrombotic clots in an individual surface arteriole upstream of a branch point. Despite the blockage, blood flow was maintained in both downstream branches because of a reversal in the direction of flow through one of the two branches—just one second after occlusion. Of a total 47 clots in 34 rats, all showed the same result. The redistribution in flow was sufficiently robust so that little change in flow occurred in vessels farther downstream from the occlusion. A second test of the brain's vascular resilience involved a more traditional method to block flow that uses a fine filament threaded through the carotid to partially obstruct the middle cerebral artery, the major source of blood to the parietal cortex. While the flow is reduced throughout the entire surface network of communicating arterioles, a pattern of reversal in flows was also seen. Thus, reversals are a common feature in the redistribution of blood across the cortex. Schaffer et al. show that the architecture of the cortical surface arterioles, with redundant connections between branches of the middle cerebral artery, ensures persistent blood flow and protects against localized occlusions. This extends the concept of redundant connections from a single loop in the circle of Willis, at the base of the brain, to the network of communicating arterioles on the cortical surface. Since humans and rats share a similar surface vasculature, these results could help identify potential links between vascular topology and stroke vulnerability in different regions of the brain.	20076529	PMC1324795	CC BY	2021-01-05 08:28:16	no	PLoS Biol. 2006 Feb 3; 4(2):e43	utf-8	PLoS Biol	2,006	10.1371/journal.pbio.0040043	oa_comm
==== Front PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1637949810.1371/journal.pmed.0030020Research ArticleEvolutionGenetics/Genomics/Gene TherapyImmunologyInfectious DiseasesMicrobiologyEpidemiology/Public HealthRespiratory MedicineInfectious DiseasesTuberculosisGeneticsEpidemiologyPromoter Variation in the DC-SIGN–Encoding Gene CD209 Is Associated with Tuberculosis DC-SIGN Variation and TuberculosisBarreiro Luis B 1 2 Neyrolles Olivier 2 Babb Chantal L 3 Tailleux Ludovic 2 Quach Hélène 1 McElreavey Ken 4 van Helden Paul D. 3 Hoal Eileen G 3 Gicquel Brigitte 2 Quintana-Murci Lluis 1 1CNRS FRE2849, Unit of Molecular Prevention and Therapy of Human Diseases, Institut Pasteur, Paris, France2Unité de Génétique Mycobactérienne, Institut Pasteur, Paris, France3Faculty of Health Sciences, Stellenbosch University, Tygerberg, South Africa4Reproduction, Fertilité et Populations, Institut Pasteur, Paris, France To whom correspondence should be addressed. E-mail: [email protected] Competing Interests: The authors have declared that no competing interests exist. Author Contributions: LBB, ON, BG, and LQ designed the study. LBB, ON, LT, KM, BG, and LQ analyzed the data. LBB, CLB, and HQ performed the sequencing and genotyping analyses. PDH and EGH coordinated the collection, clinical information, and DNA extraction of the cohort samples. LBB, ON, CLB, LT, HQ, KM, PDH, EGH, BG, and LQ contributed to writing the paper. 2 2006 3 1 2006 3 2 e208 7 2005 19 10 2005 Copyright: © 2006 Barreiro et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Variants in the DC-SIGN−Encoding Gene CD209 and Susceptibility to Tuberculosis Background Tuberculosis, which is caused by Mycobacterium tuberculosis, remains one of the leading causes of mortality worldwide. The C-type lectin DC-SIGN is known to be the major M. tuberculosis receptor on human dendritic cells. We reasoned that if DC-SIGN interacts with M. tuberculosis, as well as with other pathogens, variation in this gene might have a broad range of influence in the pathogenesis of a number of infectious diseases, including tuberculosis. Methods and Findings We tested whether polymorphisms in CD209, the gene encoding DC-SIGN, are associated with susceptibility to tuberculosis through sequencing and genotyping analyses in a South African cohort. After exclusion of significant population stratification in our cohort, we observed an association between two CD209 promoter variants (−871G and −336A) and decreased risk of developing tuberculosis. By looking at the geographical distribution of these variants, we observed that their allelic combination is mainly confined to Eurasian populations. Conclusions Our observations suggest that the two −871G and −336A variants confer protection against tuberculosis. In addition, the geographic distribution of these two alleles, together with their phylogenetic status, suggest that they may have increased in frequency in non-African populations as a result of host genetic adaptation to a longer history of exposure to tuberculosis. Further characterization of the biological consequences of DC-SIGN variation in tuberculosis will be crucial to better appreciate the role of this lectin in interactions between the host immune system and the tubercle bacillus as well as other pathogens. Two variants in the gene that encodes DC-SIGN, the major receptor for Mycobacterium tuberculosis on dendritic cells, are associated with susceptibility to TB ==== Body Introduction One-third of the world's population is estimated to be infected with Mycobacterium tuberculosis, the etiological agent of tuberculosis (TB). This disease tops the World Health Organization list of deaths due to a single infectious agent, with the death toll between 2 and 3 million people per year [1]. A perplexing, and yet unsolved, feature of TB is that less than 10% of infected individuals develop the disease. Substantial epidemiological evidence supports that host-related factors, such as sex, age, HIV infection, malnutrition, and BCG (bacille Calmette-Guérin) vaccination, influence the balance between the tubercle bacilli and host immune defences [2,3]. In addition, there is increasing evidence that host genetic factors determine differences in host susceptibility to mycobacterial infection and might contribute therefore to the pattern of clinical disease [4–7]. From a host perspective, the innate immunity system acts as the first line of host defense against microbial pathogens [8]. Initial recognition of pathogens by the innate immunity system is mediated by phagocytic cells, such as dendritic cells (DCs) or macrophages, through germline-encoded receptors, known as pattern recognition receptors [9]. DCs bear a range of pattern recognition receptors, such as C-type lectins and Toll-like receptors, involved both in recognition of conserved products of microbial metabolism and in the induction of adaptive immunity [8,10–12]. In particular, C-type lectins detect pathogens by their characteristic carbohydrate structures and internalise them for further antigen processing and presentation [13]. We have recently shown that a prototypic C-type lectin, DC-SIGN (dendritic cell–specific ICAM-3 grabbing nonintegrin), is the major Mycobacterium tuberculosis receptor on human DCs [14]. DC-SIGN is specifically, though not exclusively, expressed on DCs and functions both as a cell adhesion and as a pathogen recognition receptor [15]. As an adhesion receptor, it plays an important role in many DC functions, such as DC-T cell interaction and DC migration [16,17]. Besides its cellular recognition role, DC-SIGN serves as pathogen uptake receptor and mediates interactions with a plethora of pathogens other than M. tuberculosis [18]. Indeed, it has been shown that DC-SIGN allows DCs to capture other bacteria such as Helicobacter pylori and certain Klebsiella pneumonia strains, but also viruses such as HIV-1, Ebola, cytomegalovirus, hepatitis-C, dengue, and SARS-coV, and parasites like Leishmania pifanoi and Schistosoma mansoni [19–27]. In addition, recent data suggest that DC-SIGN may mediate intracellular signalling events leading to cytokine secretion and, on this basis, it has been proposed that the lectin could be used by pathogens, including M. tuberculosis, as a part of an immune evasion strategy to their own advantage [28,29]. In light of the ability of DC-SIGN to interact with M. tuberculosis and other pathogens, it is plausible that variation in its gene may influence the pathogenesis of a number of infectious diseases, including TB. We have therefore explored the relationship between CD209 polymorphisms and susceptibility to TB by determining CD209 sequence variation in a cohort of South African Coloured origin. Methods Patients and Methods The study was conducted in a cohort of 711 individuals, including 351 TB patients and 360 healthy controls, living in the Cape Town area. Certain suburbs of metropolitan Cape Town have some of the highest reported incidence rates of TB in the world, despite extensive BCG vaccination. Indeed, our study population comes from two suburbs that have been extensively studied because of their uniform ethnicity (known as South African Coloured) and socio-economic status as well as high incidence of TB and low prevalence of HIV [30]. In addition, our study group represents a present-day homogenous population [31] that previously received genetic input from Khoisan, Malaysian, Bantu, and European descent populations [32]. Thus, it represents a community originating from populations with different susceptibilities to TB and offers a unique opportunity to dissect the contributing genetic variants and their probable geographic/ethnic origins. TB patients were bacteriologically-confirmed (smear-positive and/or culture-positive) to present pulmonary tuberculosis. Their mean age (± standard deviation) was 36.7 ± 10.9 y, and 51.8% were male. Controls were unrelated healthy individuals from the same community, with the same socio-economic status, access to health facilities, and chance of diagnosis, and with neither signs nor previous history of TB (mean age 34.6 ± 12.5 y, 22% male). The annual risk of infection in this suburb was estimated at 2.5% in 1987 and at 2.8%–3.5% in 1999, and it is therefore highly likely that, in such an environment, the vast majority of controls have been exposed to M. tuberculosis [33,34]. All subjects were HIV-negative and older than 18 y. Informed consent was obtained from all participants, and the study was approved by the ethics committee of the Faculty of Health Sciences, Stellenbosch University (South Africa). Laboratory Procedures and Statistical Analysis To identify informative CD209 single nucleotide polymorphisms (SNPs) and to avoid ascertainment bias in the choice of markers to be tested, we first sequenced the whole CD209 genomic region (seven coding exons, flanking intronic regions, and 1,000 base pairs situated 5′ of the start codon) in 28 randomly chosen individuals (56 chromosomes). Using polymorphisms with a minimum allele frequency of 0.05, unphased genotypic data were converted into haplotypes using the accelerated EM (Expectation Maximization) algorithm implemented in Haploview v3.1 [35]. To evaluate the accuracy of the EM algorithm, haplotype reconstruction was performed in parallel using the Bayesian statistical method [36] implemented in Phase v.2.1.1. Equivalent results were obtained using both methods, with all haplotypes presenting high levels of statistical support. In order to define a minimal number of SNPs explaining most haplotypic diversity, we used the BEST v1.0 software [37]. Eight haplotype-tagging SNPs were then selected to genotype the entire panel of 711 individuals. Further, potential population stratification between cases and controls was tested by genotyping 25 unlinked SNP markers in the entire study cohort. DNA samples were genotyped by either fluorescence polarization (VICTOR-2TM technology; PerkinElmer, Wellesley, California, United States) or TaqMan (ABI Prism-7000 Sequence Detection System; Applied Biosystems, Foster City, California, United States) assays. Statistical testing for genotypic and haplotypic associations were performed using S TATA 8.2 and Haploview v3.1, respectively. The haplotype frequencies were obtained by summing the fractional likelihood of each haplotype for each individual (i.e., if a particular individual has been determined to have a 40% likelihood of haplotype A and 60% likelihood of haplotype B, 0.4 and 0.6 would be added to the counts for A and B, respectively) [35]. Results and Discussion Two variants located in the CD209 promoter region (−871 A/G and −336 A/G) exhibited a frequency distribution significantly distorted between TB patients and controls, as indicated by a Chi-square test (Table 1). For the −871 variant, genotypes GG and GA were less frequently observed in cases (16.8%) compared to the control group (27.2%) (p = 8.2 × 10−4). For the −336 variant, genotypes GG and GA were more frequent in cases (70.6%) than in controls (61.9%) (p = 0.01). These observations suggest that the alleles −871A (odds ratio [OR]: 1.85; 95% CI: 1.29–2.66) and −336G (OR: 1.48; 95% CI: 1.08–2.02) increase the risk of developing TB in our South African cohort. At the haplotype level (Table 2), a Chi-square test first revealed that the global distribution of haplotype frequencies was significantly different between cases and controls (p = 1.2 × 10−3). One haplotype (H3) turned out to be the main haplotype responsible for such a distorted frequency distribution (Table 2). This haplotype, which contains both −871G and −336A, was found to be strongly associated with the control group (p = 1.6 × 10−3; OR: 1.7; 95% CI: 1.22–2.38). The associations with this haplotype, and with −871, remained highly significant (p = 1.3 × 10−2 and 6.6 × 10−3 respectively), even after the conservative Bonferroni correction for multiple testing. Table 1 DC-SIGN Genotype Distributions in Patients with Tuberculosis and in Healthy Controls Table 2 Haplotype Distributions in Patients with Tuberculosis and in Healthy Controls Although our cohort is considered a present-day homogeneous community that has received genetic contribution from different populations multiple generations ago [31,32], population stratification between cases and controls can be a confounding factor leading to a spurious positive association. Indeed, the use of admixed populations in association-mapping studies can be very useful to identify disease-causing genetic variants that differ in frequency across parental populations. However, when the admixture event is too recent, allelic frequencies can differ coincidentally among cases and controls, reflecting a nonuniform genetic contribution from the parental populations to each subpopulation (i.e., cases and controls), rather than a genuine association between a given genetic variant and the phenotype under study. In this case, the study-cohort is said to present population stratification. To formally test and quantify the levels of background genetic differences [38], if any, between cases and controls, we genotyped the entire cohort for a panel of 25 independent SNPs markers which are (1) not in linkage disequilibrium with the candidate CD209 locus and with any other known gene, (2) randomly distributed along the genome, and (3) polymorphic among the major ethnic groups (Table 3). The mean χ2 statistic among the 25 SNPs for the comparison of allele frequencies between cases and controls, which represents the levels of stratification (μ) between the two groups [39], was 1.25 (p = 0.26), implying that the two groups were not significantly stratified. As an additional correction for stratification, we divided the χ2 values obtained for our candidate gene CD209 by the level of stratification detected (1.25) [39]. Even after such a conservative correction, the associations observed with −336 and −871 as well as with H3 remained significant (−336 p = 2.8 × 10−2; −871 p = 2.7 × 10−3; H3 p = 4.8 × 10−3). These observations support therefore the idea that the −871G and −336A variants are indeed genuinely associated with a protective role against TB. Table 3 Frequency Distribution in the Study Cohort of the 25 SNPs Used to Test for Population Stratification In order to gain insights into the frequency distribution of these two SNPs, we genotyped them in 254 human chromosomes from sub-Saharan Africa, Europe, and East Asia as well as in eight chimpanzee chromosomes. We observed that the −871G and −336A forms, which we propose as offering protection against TB, corresponded to the derived allele in humans; we also observed that these forms are present at higher frequencies in Eurasians as compared to Africans (Table 4). Indeed, the −871G is absent in African populations whereas it reaches high frequencies (20%–40%) in European and Asian populations. Given the absence of the haplotypic combination of −871G and −336A among sub-Saharan Africans, its presence among South African Coloureds suggests that it was introduced through the historically well-known admixture with Europeans and Asians [31]. This observation highlights the power of using admixed populations to better understand historical issues associated with the geographic/ethnic origin of disease-affecting alleles, provided that their prevalence varies in the ancestors of the admixed population (i.e., different frequency of H3 in Africans versus non-Africans; Table 4). Table 4 Frequency Distribution of the Eight CD209 SNPs Genotyped in the Multi-Ethnic Panel of 127 Individuals (n = 254 Chromosomes) as Well as in the 711 Individuals of the South African Cohort (n = 1,422 Chromosomes) In the context of TB, it has been suggested that present-day susceptibility to TB is determined by previous history of exposure [40]. There is fairly convincing evidence that TB has been endemic in Europe for several hundred years, whereas in Africa it has probably been rare before contact was initiated with Europeans [41–43]. It is expected therefore that M. tuberculosis has exerted stronger selective pressures on European than African populations [42]. Our results lend support to this hypothesis and suggest that the protective alleles −871G and −336A increased in frequency in non-African populations as a result of genetic adaptation to a longer period of TB exposure. The potential impact of tuberculosis on the frequency of resistant alleles in European populations has been recently addressed using epidemiological data and statistical modeling [44]. The authors have sought to evaluate the expected changes in resistant allele frequencies, during the 300-y period corresponding to the peak epidemics of TB in Europe. They concluded that if a given resistant allele was at a low frequency in the beginning of an epidemic, selection by M. tuberculosis alone would increase the frequency of this allele, but not enough to bring it to epidemiologically significant levels. In this context, since DC-SIGN is known to interact with a vast range of pathogens, it is indeed likely that the increased frequencies observed today for both −871G and −336A in non-African populations (specially for −871G which is absent in sub-Saharan Africans) may have been driven, not only by the selective pressures imposed by M. tuberculosis, but also by other infectious agents. Indeed, two independent studies have recently reported a genetic association between the −336A variant and protection against parenteral HIV infection [45] and severity of dengue pathogenesis [46]. Although HIV infection, for example, is too recent to have left any signature of selection on CD209, these observations emphasize the possible action of other pathogens in shaping the patterns of variability of this gene. From a functional point of view, the −336A allele has been shown to affect an Sp1-like binding site and to modulate transcriptional activity in vitro by increasing the levels of expression [46]. In the context of TB, increased DC-SIGN expression levels by DCs may result in better capture and processing of mycobacterial antigens, leading to a stronger and wider T-cell response. In addition, we have recently shown that DC-SIGN expression is markedly induced in alveolar macrophages in active TB patients and that M. tuberculosis is preferentially phagocytosed by DC-SIGN–expressing macrophages in these individuals [47]. Thus, the higher prevalence observed among healthy individuals of the −336A variant, which is associated with increased DC-SIGN expression, may underlie an increased efficiency of host phagocytes, such as DCs and macrophages, to control the infection. In addition to the −336A variant, our genetic data showed a strong association of the −871G allele with healthy controls, suggesting also a functional consequence of this variant that, either alone or in combination with −336A, remains to be defined. In conclusion, the significant association found for the CD209 promoter variants together with their phylogenetic status and frequency distribution strongly suggests that the −871G and −336A alleles may reduce the risk of developing TB. More generally, our results, together with those reporting association of CD209 promoter variants with both HIV susceptibility and dengue pathogenesis [45,46] suggest that variation in this lectin may be of crucial importance in the outcome of a number of infections due to DC-SIGN–interacting pathogens. Detailed in vitro and in vivo studies assessing the functional consequences of CD209 variants on the quality of the host immune response against pathogens, including M. tuberculosis, are now required to eventually develop knowledge-based and effective pathway-targeted treatments. Patient Summary Background Tuberculosis (TB) is an infectious disease caused by the bacterium Mycobacterium tuberculosis. The disease kills between 2 million and 3 million people each year, and almost one-third of the world's population, or 2 billion people, are thought to be infected with the bacterium. However, only about 10% of infected individuals become sick. In the other cases, the body's immune system is capable of controlling the growth of the bacteria. Whether a person's immune system is strong enough to keep the bacterium in check depends on a variety of factors, such as age, how well nourished the person is, other infections, and genetic make-up. Why Was This Study Done? Understanding the genetic factors that influence whether an infected person is more or less susceptible to getting sick with TB should help doctors to fight the disease. A number of research groups are studying these genetic factors, and some genes have previously been identified that seem to increase a person's risk of TB. Some of the authors of this study work on a molecule that is part of the immune system. The molecule is called DC-SIGN, and it plays a role in infection of cells in the lung by the Mycobacterium. Together with colleagues who are genetics experts, they wanted to see whether variation in the gene (called CD209) that codes for DC-SIGN influences the risk of someone infected with M. tuberculosis getting sick. What Did the Researchers Do and Find? They worked with a group of South African participants from an area close to Cape Town where TB is very common. Half of the participants had TB; the other half did not (even though, because TB is so common in that area, it was likely that most if not all of them had been infected with M. tuberculosis at some point). The researchers compared make-up of the gene for DC-SIGN in all participants and found that some variants of the gene were more common in the group of healthy individuals. In other words, having this particular type of DC-SIGN-encoding gene seemed to protect individuals from getting TB. What Does This Mean? DC-SIGN is a central part of the immune system, and others scientists have reported links between variants in the gene for DC-SIGN and the risk of picking up other infectious diseases, including dengue fever and HIV/AIDS. The present study lends more support to the notion that DC-SIGN is a key player in the control of infectious diseases. Understanding more about DC-SIGN could help to develop better treatments for these infections. Where Can I Find More Information Online? The following Web sites provide information about tuberculosis. World Health Organization pages on TB: http://www.who.int/mediacentre/factsheets/fs104/en/index.html OMNI pages on TB: http://omni.ac.uk/browse/mesh/D014376.html Medline PLUS pages on TB: http://medlineplus.nlm.nih.gov/medlineplus/tuberculosis.html European TB Vaccine Cluster: http://www.tb-vac.org LBB was supported by a “Fundação para a Ciência e a Tecnologia” fellowship (SFRH/BD/18580/2004). This research project has been co-financed by the European Commission, within the 6th Framework Programme (contract no. LSHP-CT-2003–503367). The text represents the authors' views and does not necessarily represent a position of the Commission who will not be liable for the use made of such information. We thank M. Kennedy for help with sample collection, and the South African Medical Research Council and Department of Science and Technology/National Research Foundation (DST/NRF) Centres of Excellence for financial assistance. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Citation: Barreiro LB, Neyrolles O, Babb CL, Tailleux L, Quach H, et al. (2006) Promoter variation in the DC-SIGN–encoding gene CD209 is associated with tuberculosis. PLoS Med 3(2): e20. Abbreviations DCdendritic cell ORodds ratio SNPsingle nucleotide polymorphism TBtuberculosis ==== Refs References Frieden TR Sterling TR Munsiff SS Watt CJ Dye C Tuberculosis Lancet 2003 362 887 899 13678977 Enarson DA Rouillon A Davies PD The epidemiological basis of tuberculosis control Clinical tuberculosis 1994 London Chapman and Hall 19 32 Lienhardt C From exposure to disease: The role of environmental factors in susceptibility to and development of tuberculosis Epidemiol Rev 2001 23 288 301 12192738 Bellamy R Susceptibility to mycobacterial infections: The importance of host genetics Genes Immun 2003 4 4 11 12595896 Blackwell JM Genetics and genomics in infectious disease susceptibility Trends Mol Med 2001 7 521 526 11689339 Cooke GS Hill AV Genetics of susceptibility to human infectious disease Nat Rev Genet 2001 2 967 977 11733749 Casanova JL Abel L Genetic dissection of immunity to mycobacteria: The human model Annu Rev Immunol 2002 20 581 620 11861613 Janeway CA Medzhitov R Innate immune recognition Annu Rev Immunol 2002 20 197 216 11861602 Kimbrell DA Beutler B The evolution and genetics of innate immunity Nat Rev Genet 2001 2 256 267 11283698 Taylor PR Martinez-Pomares L Stacey M Lin HH Brown GD Macrophage receptors and immune recognition Annu Rev Immunol 2005 23 901 944 15771589 Medzhitov R Toll-like receptors and innate immunity Nat Rev Immunol 2001 1 135 145 11905821 Fujita T Evolution of the lectin-complement pathway and its role in innate immunity Nat Rev Immunol 2002 2 346 353 12033740 McGreal EP Miller JL Gordon S Ligand recognition by antigen-presenting cell C-type lectin receptors Curr Opin Immunol 2005 17 18 24 15653305 Tailleux L Schwartz O Herrmann J Pivert E Jackson M DC-SIGN is the major Mycobacterium tuberculosis receptor on human dendritic cells J Exp Med 2003 197 121 127 12515819 Soilleux EJ Morris LS Leslie G Chehimi J Luo Q Constitutive and induced expression of DC-SIGN on dendritic cell and macrophage subpopulations in situ and in vitro J Leukoc Biol 2002 71 445 457 11867682 Geijtenbeek TB Torensma R van Vliet SJ van Duijnhoven GC Adema GJ Identification of DC-SIGN, a novel dendritic cell-specific ICAM-3 receptor that supports primary immune responses Cell 2000 100 575 585 10721994 Geijtenbeek TB Krooshoop DJ Bleijs DA van Vliet SJ van Duijnhoven GC DC-SIGN-ICAM-2 interaction mediates dendritic cell trafficking Nat Immunol 2000 1 353 357 11017109 Geijtenbeek TB van Vliet SJ Engering A ‘t Hart BA van Kooyk Y Self- and nonself-recognition by C-type lectins on dendritic cells Annu Rev Immunol 2004 22 33 54 15032573 Geijtenbeek TB Kwon DS Torensma R van Vliet SJ van Duijnhoven GC DC-SIGN, a dendritic cell-specific HIV-1-binding protein that enhances trans-infection of T cells Cell 2000 100 587 597 10721995 Alvarez CP Lasala F Carrillo J Muniz O Corbi AL C-type lectins DC-SIGN and L-SIGN mediate cellular entry by Ebola virus in cis and in trans J Virol 2002 76 6841 6844 12050398 Halary F Amara A Lortat-Jacob H Messerle M Delaunay T Human cytomegalovirus binding to DC-SIGN is required for dendritic cell infection and target cell trans-infection Immunity 2002 17 653 664 12433371 Lozach PY Lortat-Jacob H de Lacroix de Lavalette A Staropoli I Foung S DC-SIGN and L-SIGN are high-affinity binding receptors for hepatitis C virus glycoprotein E2 J Biol Chem 2003 278 20358 20366 12609975 Tassaneetrithep B Burgess TH Granelli-Piperno A Trumpfheller C Finke J DC-SIGN (CD209) mediates dengue virus infection of human dendritic cells J Exp Med 2003 197 823 829 12682107 Marzi A Gramberg T Simmons G Moller P Rennekamp AJ DC-SIGN and DC-SIGNR interact with the glycoprotein of Marburg virus and the S protein of severe acute respiratory syndrome coronavirus J Virol 2004 78 12090 12095 15479853 Bergman MP Engering A Smits HH van Vliet SJ van Bodegraven AA Helicobacter pylori modulates the T helper cell 1/T helper cell 2 balance through phase-variable interaction between lipopolysaccharide and DC-SIGN J Exp Med 2004 200 979 990 15492123 Appelmelk BJ van Die I van Vliet SJ Vandenbroucke-Grauls CM Geijtenbeek TB Carbohydrate profiling identifies new pathogens that interact with dendritic cell-specific ICAM-3-grabbing nonintegrin on dendritic cells J Immunol 2003 170 1635 1639 12574325 Colmenares M Puig-Kroger A Pello OM Corbi AL Rivas L Dendritic cell (DC)-specific intercellular adhesion molecule 3 (ICAM-3)- grabbing nonintegrin (DC-SIGN, CD209), a C-type surface lectin in human DCs, is a receptor for Leishmania amastigotes J Biol Chem 2002 277 36766 36769 12122001 Geijtenbeek T van Vliet S Koppel E Sanchez-Hernandez M Vandenbroucke-Grauls C Mycobacteria target DC-SIGN to suppress dendritic cell function J Exp Med 2003 197 7 17 12515809 Tailleux L Maeda N Nigou J Gicquel B Neyrolles O How is the phagocyte lectin keyboard played? Master class lesson by Mycobacterium tuberculosis Trends Microbiol 2003 11 259 263 12823942 Beyers N Gie RP Zietsman HL Kunneke M Hauman J The use of a geographical information system (GIS) to evaluate the distribution of tuberculosis in a high-incidence community S Afr Med J 1996 86 40 41, 44 8685780 van der Ross RE 100 questions about Coloured South Africans 1993 Cape Town (South Africa) UWC Printing Department 36 Nurse GT Weiner JS Jenkins T The peoples of Southern Africa and their affinities 1985 Oxford (United Kingdom) Clarendon Press 410 Munch Z Van Lill SW Booysen CN Zietsman HL Enarson DA Tuberculosis transmission patterns in a high-incidence area: A spatial analysis Int J Tuberc Lung Dis 2003 7 271 277 12661843 Beyers N Michaelis I Gie R Schaaf HS Richardson M Transmission of tuberculosis (TB) to children in a high incidence area Int J Tuberc Lung Dis 2001 5 Suppl 1 S185 Barrett JC Fry B Maller J Daly MJ Haploview: Analysis and visualization of LD and haplotype maps Bioinformatics 2005 21 263 265 15297300 Stephens M Donnelly P A comparison of Bayesian methods for haplotype reconstruction from population genotype data Am J Hum Genet 2003 73 1162 1169 14574645 Sebastiani P Lazarus R Weiss ST Kunkel LM Kohane IS Minimal haplotype tagging Proc Natl Acad Sci U S A 2003 100 9900 9905 12900503 Pritchard JK Rosenberg NA Use of unlinked genetic markers to detect population stratification in association studies Am J Hum Genet 1999 65 220 228 10364535 Reich DE Goldstein DB Detecting association in a case-control study while correcting for population stratification Genet Epidemiol 2001 20 4 16 11119293 O'Brien SJ Ghetto legacy Curr Biol 1991 1 209 211 15336121 Stead WW Senner JW Reddick WT Lofgren JP Racial differences in susceptibility to infection by Mycobacterium tuberculosis N Engl J Med 1990 322 422 427 2300105 Stead WW Genetics and resistance to tuberculosis. Could resistance be enhanced by genetic engineering? Ann Intern Med 1992 116 937 941 1580452 Daniel TM Bates JH Downes KA Bloom BR History of tuberculosis Tuberculosis: Pathogenesis, protection, control 1994 Washington, DC ASM Press 13 24 Lipsitch M Sousa AO Historical intensity of natural selection for resistance to tuberculosis Genetics 2002 161 1599 1607 12196403 Martin MP Lederman MM Hutcheson HB Goedert JJ Nelson GW Association of DC-SIGN promoter polymorphism with increased risk for parenteral, but not mucosal, acquisition of human immunodeficiency virus type 1 infection J Virol 2004 78 14053 14056 15564514 Sakuntabhai A Turbpaiboon C Casademont I Chuansumrit A Lowhnoo T A variant in the CD209 promoter is associated with severity of dengue disease Nat Genet 2005 37 507 513 15838506 Tailleux L Pham-Thi N Bergeron-Lafaurie N Herrmann JL Charles P DC-SIGN induction in alveolar macrophages defines a privileged target host cell population for mycobacteria in patients with tuberculosis PLoS Med 2005 2 12 e381 16279841	16379498	PMC1324949	CC BY	2021-01-05 11:22:50	no	PLoS Med. 2006 Feb 3; 3(2):e20	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030020	oa_comm
==== Front PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1638159810.1371/journal.pmed.0030021Research ArticleVirologyEpidemiology/Public HealthHealth EconomicsHealth PolicyMedical EthicsMedical LawHealth EconomicsHealth PolicyScreeningInfectious DiseasesPublic HealthCost-Effectiveness of Alternative Blood-Screening Strategies for West Nile Virus in the United States Alternative Blood-Screening StrategiesKorves Caroline T 1 2 Goldie Sue J 3 Murray Megan B 1 4 1Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts, United States of America2The Earth Institute, Columbia University, New York, New York, United States of America3The Harvard Center for Risk Analysis, Department of Health Policy and Management, Harvard School of Public Health, Boston, Massachusetts, United States of America4Infectious Disease Unit, Massachusetts General Hospital, Boston, Massachusetts, United States of AmericaMostashari Farzad Academic EditorNew York City Department of Health and Mental HygieneUnited States of America To whom correspondence should be addressed. E-mail: [email protected] Competing Interests: The authors have declared that no competing interests exist. Author Contributions: CTK, SJG, and MBM designed the study and contributed to writing the paper. CTK and MBM analyzed the data. SJG provided oversight of the model-development process and model simulations. 2 2006 24 1 2006 3 2 e214 8 2005 20 10 2005 Copyright: © 2006 Korves et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. The Cost of Screening Blood Donations for West Nile Virus Background West Nile virus (WNV) is endemic in the US, varying seasonally and by geographic region. WNV can be transmitted by blood transfusion, and mandatory screening of blood for WNV was recently introduced throughout the US. Guidelines for selecting cost-effective strategies for screening blood for WNV do not exist. Methods and Findings We conducted a cost-effectiveness analysis for screening blood for WNV using a computer-based mathematical model, and using data from prospective studies, retrospective studies, and published literature. For three geographic areas with varying WNV-transmission intensity and length of transmission season, the model was used to estimate lifetime costs, quality-adjusted life expectancy, and incremental cost-effectiveness ratios associated with alternative screening strategies in a target population of blood-transfusion recipients. We compared the status quo (baseline screening using a donor questionnaire) to several strategies which differed by nucleic acid testing of either pooled or individual samples, universal versus targeted screening of donations designated for immunocompromised patients, and seasonal versus year-long screening. In low-transmission areas with short WNV seasons, screening by questionnaire alone was the most cost-effective strategy. In areas with high levels of WNV transmission, seasonal screening of individual samples and restricting screening to blood donations designated for immunocompromised recipients was the most cost-effective strategy. Seasonal screening of the entire recipient pool added minimal clinical benefit, with incremental cost-effectiveness ratios exceeding US$1.7 million per quality-adjusted life-year gained. Year-round screening offered no additional benefit compared to seasonal screening in any of the transmission settings. Conclusions In areas with high levels of WNV transmission, seasonal screening of individual samples and restricting screening to blood donations designated for immunocompromised recipients is cost saving. In areas with low levels of infection, a status-quo strategy using a standard questionnaire is cost-effective. Screening all blood donations in all States to avoid transmission to blood transfusion recipients is not cost-effective. ==== Body Introduction The first US-based case of West Nile virus (WNV) disease was reported in New York in 1999 [1]; since then, the virus has spread across the country, leading to 16,637 detected cases of WNV-associated illness and 647 WNV-associated deaths from 1999 to 2004, inclusive (http://www.cdc.gov/). WNV is a neuropathic flavivirus, transmitted from birds to humans via the mosquito vector. Naturally acquired infection is most often asymptomatic; about 20% of those infected develop a flu-like illness characterized by fever, while a smaller proportion (less than 1%) develop neuroinvasive disease (NI). WNV-associated NI can result in death or in serious sequelae; these poor outcomes occur most commonly in elderly and immunocompromised patients [2]. The Centers for Disease Control and Prevention (CDC) recently reported on the first 23 patients infected through blood transfusion, six of whom died [3]. Although they are a very small minority of the WNV cases reported, transfusion-acquired infections are potentially avoidable, and public health authorities moved quickly to try to safeguard the blood supply from the virus. The Food and Drug Administration (FDA) accelerated the approval of two investigational nucleic acid-based blood-safety tests, and screening of donations was initiated by July 2003 [4]. WNV was subsequently detected in 818 blood donations [3], leading public health officials to conclude that the screening program had prevented at least some WNV transmission. Nonetheless, at least six cases of transfusion-associated WNV have been reported since the initiation of the testing, raising concern that current testing strategies may be inadequate. The choice of optimal screening strategy will likely depend on the prevalence of WNV infection among donors, the duration of the epidemic season, the dilution of the pooled samples, and the underlying health status of blood-transfusion recipients. Although screening of blood by nucleic acid test (NAT) is currently mandated by the FDA, actual implementation of screening strategies is largely at the discretion of the individual states and the blood-collection agencies. For example, the FDA specifies that testing must be performed in all states from May until the end of October, but allows year-long testing if states deem it necessary [4]. Similarly, NAT can be performed on individual blood donations or on pooled samples. Compared to individual sample testing, the pooling of samples lowers screening costs at the expense of diminished assay sensitivity [5]. Furthermore, current policy mandates screening blood for all transfusion recipients. A subset of transfusion recipients may have an impaired immune response owing to malignancy, HIV, or to the use of immunosuppressive drugs. Targeted screening of blood designated for immunocompromised patients, who are most at risk of developing severe consequences following WNV infection, may be an alternative to universal screening. This approach has been used to prevent transfusion-associated cytomegalovirus infection in immunocompromised patients [6]. To identify the most medically effective and cost-effective screening strategies under a range of different circumstances, we conducted a cost-effectiveness analysis of alternative strategies for WNV blood screening and considered the effects of variable assay characteristics, transfusion outcomes, and pricing that may affect current and future policy decisions. Methods Analytic Overview We developed a state-transition model to simulate the natural history of WNV infection transmitted from blood donors to transfusion recipients. The model was used to estimate lifetime costs, life expectancy, and quality-adjusted life expectancy for different blood-screening strategies in a target population of people receiving blood transfusions. Since the cost-effectiveness of screening depends on the prevalence of infectious cases among blood donors, the analyses were performed for three different populations of donors residing in states in which the incidence of WNV varies in intensity and duration of the natural transmission season. Parameters describing the natural history of transfusion-associated WNV were derived from the literature. The implications of alternative parameter values were evaluated in sensitivity analyses. We chose the 95% confidence interval of estimated probabilities to establish clinical parameter ranges, and we consulted experts in the field to establish price and assay ranges. We adopted a societal perspective and followed the reference-case recommendations of the Panel on Cost-Effectiveness in Health and Medicine [8]. Future costs and QALYs saved were discounted at an annual rate of 3%. Alternative screening strategies were compared using the ICER. We assessed the internal consistency and face validity of our model by predicting the number of transfusion cases for each state for 2002 and comparing these outputs with the number of cases reported to state health departments for that year (R. J. Powell, K. Signs, R. Timperi, K. A. Winpisinger, personal communications). We conducted extensive sensitivity analyses to evaluate the stability of our conclusions over a wide range of parameter estimates and assumptions. Strategies We considered four main WNV-screening strategies to be added to baseline screening of blood donors for general infectious diseases. Baseline screening involves the distribution of a donor questionnaire that elicits information about any recent history of fever and is designed to ensure that the blood of individuals with active infections is excluded from the blood supply. Blood centers design their own questionnaires about general health and well being based on donor-eligibility rules established by the FDA (http://americanredcrossblood.org/). An abbreviated physical examination includes checking for abnormalities in blood pressure, pulse, and temperature (http://www.aabb.org/). The four WNV-screening strategies to be superimposed on the questionnaire include: (1) nucleic acid testing of minipools of 16 samples (MP16-NAT) followed by testing the samples in a reactive pool by individual nucleic acid test (ID-NAT); (2) nucleic acid testing of minipools of six samples (MP6-NAT) followed by individual testing of the samples in a reactive pool by ID-NAT; (3) nucleic acid testing of individual samples by ID-NAT; and (4) individual nucleic acid testing on blood donations designated for immunocompromised recipients only. We also evaluated the implications of “seasonal targeting,” meaning that supplemental strategies would be brought into operation only during the “high incidence” WNV season from May through to the end of October in contrast to year-round screening. Model Structure and Assumptions We constructed a Markov model for each of the three WNV-transmission scenarios described in Table 1. Markov models depict the natural history of disease as an evolving sequence of health states, defined to capture important clinical outcomes, each of which is associated with specific costs. The time horizon of the analysis incorporates a transfusion recipient's lifetime and is divided into equal weekly increments during which transitions between health states occur. The following five mutually exclusive health states included in this model describe the various WNV infection and disease statuses of people after they have received a blood transfusion: (1) no WNV infection or asymptomatic infection; (2) WNV infection leading to febrile illness only; (3) WNV infection leading to NI with no long-term sequelae; (4) WNV infection leading to NI associated with long-term sequelae leading to either institutionalized or home care; or (5) death (Figure 1). Individuals entering the model are assumed to be uninfected at the time of transfusion and are assigned gender, age, age-specific post-transfusion life expectancy, and immune status based on the distribution of these characteristics in previously studied transfused populations [9]. Figure 1 Post-Transfusion Health States Five post-transfusion health states were identified. Following transfusion, individuals entered the uninfected or asymptomatic infection state, febrile-illness state, or NI state and progressed to other health states in the direction of the arrows. Individuals were followed until death. Table 1 WNV Disease Characteristics in General Population Giving Rise to Risk in Hypothetical Transfusion Cohorts Transition probabilities derived from a review of the literature were used to move transfusion recipients through different health states over time until all the members of the transfused cohort had died. For example, upon transfusion, each individual faced a risk of transfusion-acquired WNV infection. The risk of infection varied depending on the sensitivity and specificity of the specific screening strategies being analyzed. We assumed that a WNV-positive blood donation always resulted in infection but that the risk of disease in those who were infected was higher in elderly and immunocompromised patients. Once infected, individuals progressed through a week-long incubation period after which they could either remain asymptomatic or make the transition to a health state characterized by febrile illness only or by NI, with or without long-term sequelae or death. Patients who moved into the febrile-illness-only state either recovered or died within 1 wk from causes unrelated to WNV infection, while those in the NI state faced a weekly probability of recovery, death from WNV, or death from unrelated causes. Natural History of Transfusion-Acquired WNV We estimated the weekly probability that a unit of transfused blood would be infected with WNV in each of the three transmission scenarios using a method developed by Biggerstaff et al. [10,11]. Following this approach, we used data on the dates of onset of detected neuroinvasive cases, estimated distributions of WNV incubation, symptomatic and viremic periods, and the ratio of detected to undetected infections to estimate the number of potential blood donors with WN viremia at each time point over a 1-y period. Dates of onset of naturally acquired WNV-associated NI for each of the three scenarios, which were based on corresponding WNV-associated NI case data, were obtained from the CDC Arbonet surveillance Team for 2002. Based on data from previous serological studies, we estimated that, for every reported case of WNV-associated NI, 140 cases of WNV infection had occurred that were either asymptomatic or had presented with fever only [12]. We then used the dates of symptom onset of the detected cases as “anchor times” from which we generated, for each non-NI case, a date of infection, onset of viremia, onset of symptoms, resolution of symptoms, and resolution of viremia. By dividing the number of viremic individuals by the state population, we determined the probability of WNV infection among the general population for a given week for each scenario. Figure 2 illustrates the potential viremic donor times. Without detection by NAT, asymptomatic individuals could donate on any day. Durations of the incubation, viremic, and symptomatic periods were derived from the literature; the ranges of these values are given in Table 2. Once the time line of infection was generated for each case, we counted the number of potential donors who were viremic on each day during the course of a year, and generated a curve representing the prevalence of viremic blood donors over the time course of the simulation. We repeated the simulation 1,000 times to obtain Monte Carlo averages of the prevalence of viremic donors, and these averages were then used to estimate the probability that a blood-transfusion recipient would be transfused with WNV-infected blood. By designing our model to output the weekly probability of a viremic donation, we were able to evaluate the impact of implementing supplemental assay screening for selected weeks within the year. Figure 2 Potential Donor Time for an Infected Individual A potential blood donor who develops WNV symptoms is viremic and eligible to donate for a longer period of time as the latent period decreases and the incubation period increases. A potential blood donor who recovers from symptoms in a short amount of time may be eligible to donate before viremia ends. Table 2 Model Variables: Baseline Values and Values Used in Sensitivity Analysis Table 2 Continued Clinical Data Clinical parameters [13–20] used in the model are shown in Table 2. Since there are few data on death from NI due to WNV, we used national hospital data on death from NI due to multiple causes collected by the Healthcare Cost and Utilization Project (http://www.ahrq.gov/hcupnet/). We utilized unpublished follow-up data on 36 hospitalized patients with WNV to estimate long-term recovery from WNV-associated NI (D. Nash and A. K. Labowitz, personal communication). Distributions of age, sex, and immune status for transfusion recipients were based on transfusion look-back studies [14]. Since there are no population-based studies that reported the probability of developing febrile illness or NI after acquiring transfusion-associated WNV infection, we relied on data from an experimental study of deliberate WNV inoculation of humans with cancer [13]. Age and sex-specific background mortality for the blood recipients was based on a transfusion-cohort study [21]. Screening Tests Table 2 summarizes baseline estimates of test sensitivity and specificity for ID-NAT [22]. ID-NAT sensitivity was derived from a study [22] which compared methods of virus detection in macaques experimentally inoculated with WNV. The sensitivities of MP6-NAT and MP16-NAT relative to ID-NAT were estimated from data on the proportion of ID-NAT–positive samples that were identified by minipool tests; FDA meeting transcripts (http://www.fda.gov/ohrms/dockets/ac/03/transcripts/4014T1.htm) and a publicly available Web site (http://www.innovations-report.de/html/berichte/medizin_gesundheit/bericht-24048.html) provided these data. In the absence of definitive data, we assumed that specificity of these tests was high. Health-Related Quality of Life We used a quality-of-life well-being index to assign quality weights to uninfected and asymptomatically infected individuals based on their age and sex [23]. Quality weights for the NI state were based on a study of herpes simplex infection of the central nervous system [24]. The quality weights for neurological sequelae states requiring institutionalized care and home care were based on a study of neurological sequelae resulting from Haemophilus influenzae vaccination [25]. Costs The cost estimates [26–28] used in the base case are shown in Table 2. Direct costs for screening blood donors using WNV assays included screening kit and reagents, laboratory technician fees, and the costs of a discarded false positive, donor notification, and retrieval of a test-positive sample. Since WNV assays have not yet been priced, we estimated their cost from studies of similar assays for other viruses. Other screening costs were obtained from state laboratories (R. Timperi, personal communication) and could be corroborated with data from a screening study [26] for transfusion-acquired malaria in Canada. Direct medical costs for individuals with NI and neurological sequelae from WNV were derived from published studies on other arboviral infections that lead to similar clinical outcomes. These studies included detailed estimates of resource utilization, including hospitalization, outpatient visits, and laboratory tests. Data from the US Bureau of Labor Statistics were used to assign a cost for the time required by an individual to care for a homebound patient on a full-time basis. To account for inflation, all costs were converted to 2003 US dollars by use of the Medical Care Component of the Consumer Price Index (http://www.bls.gov/). Results Face Validity of the Model For the year 2002, the model predicts plausible numbers of transfusion-acquired WNV-associated NI cases. Assuming events are Poisson-distributed, these case predictions fall within the 95% tolerance intervals constructed by the observed numbers of transfusion-acquired WNV-associated NI cases by the relevant state health departments (Table 3). Table 3 Observed and Predicted Number of Transfusion-Acquired WNV-Associated NI Cases in 2002 with Status-Quo Screening (Questionnaire) Projected Clinical and Economic Outcomes The clinical outcomes of the different screening strategies in three different regions of varying transmission intensity are presented in Table 4. In the absence of specific screening for WNV, we projected, per 2 million transfusions, a total of 277 cases of WNV infection plus 50 cases of NI for the high-infection/short-duration epidemic scenario, and 205 cases of WNV infection plus 46 cases of NI for the high-infection/long-duration epidemic scenario. For the low-infection/short-duration epidemic scenario, we projected eight cases of WNV infection and one case of NI per 2 million transfusions. Table 4 Predicted Clinical Outcomes Associated with Supplemental NAT Blood Screening Relative to Status-Quo Screening (Questionnaire) Screening year-round did not prevent a greater number of cases than did seasonal screening from May to the end of October, even in areas with a long transmission season. The introduction of screening by ID-NAT had no impact in the low-intensity setting on the expected case numbers, but reduced the expected number of infections and cases of NI in the two higher-intensity settings. Screening with the MP6-NAT and MP16-NAT also reduced the number of infections and cases of NI, but was less effective than screening with ID-NAT for the higher-transmission-intensity settings. While the strategy of restricting screening to blood designated for the immunocompromised population alone prevented fewer infections than screening the entire population of blood donors with ID-NAT, there was little impact on the number of cases of NI. In the low-infection/short-duration scenario, the baseline strategy of screening through questionnaire alone was the least-costly and most cost-effective alternative; supplemental screening strategies increased the total cost significantly, but did not reduce the number of cases or increase quality-adjusted life expectancy in this setting (Table 5). In contrast, several of the screening strategies in the higher-intensity scenarios were less expensive and more cost-effective than the questionnaire alone, since the incremental costs of implementing these strategies were outweighed by the averted direct medical costs. In the high-intensity/short-duration setting, seasonal ID-NAT screening of blood designated for transfusion of immunocompromised patients was the most cost-effective of these strategies. Although seasonal screening of the entire donor pool with ID-NAT was nominally more effective, this was associated with an incremental quality-adjusted life expectancy benefit of less than 1 min. Comparable results were obtained for the high-intensity/long-duration scenario. In this setting again, seasonal screening with ID-NAT of blood designated for transfusion of immunocompromised patients dominated the other strategies, providing 3.6 min more of quality-adjusted life expectancy than the questionnaire alone and at less cost. Screening the entire donor pool with ID-NAT was slightly more effective at the cost of US$1.7 million/QALY gained. Table 5 Predicted QALYs, Costs, Clinical Outcomes, and ICERs Associated with NAT Blood Screening for WNV in a Cohort of 2,000,000 Sensitivity Analyses Within areas of low infection /short duration of WNV, the questionnaire alone remained the least-costly strategy across a wide range of sensitivity analyses. When we assumed that assay sensitivity was as high as other nucleic acid tests, seasonally screening the entire donor pool by MP16-NAT was also on the efficiency curve; however, the ICER was US$1.2 million/QALY gained. Similarly, when we assumed an estimate from the high end of the plausible range for the probability of severe disease, ID-NAT added less than 1 min to the average quality-adjusted life expectancy at a cost of US$1 million/QALY. Further details are shown in Tables S1–S6. For areas with high infection/short duration, the rank order of strategies was sensitive to variations in test sensitivity and the risk of developing severe disease. When we assumed the high assay sensitivity, seasonally screening the entire donor pool by MP6-NAT was the least-costly strategy; seasonally screening the entire donor pool by ID-NAT was also on the efficiency curve but exceeded US$7 million/QALY. When we assumed a high risk of severe disease, unrestricted seasonal screening by ID-NAT was the only non-dominated strategy. We also evaluated the effect of shortened seasonal screening from mid-July to mid-October versus full seasonal screening from May to the end of October for this transmission area. Shortened seasonal screening offered the same clinical benefit at lower cost than full seasonal screening; targeted screening of blood designated for transfusion of immunocompromised patients was the least-costly strategy. The ICER for universal screening by ID-NAT versus targeted screening of blood designated for transfusion of immunocompromised patients remained too high for universal screening to be a cost-effective strategy, even with shortened seasonal screening. For the high-infection/long-duration scenario, our results were sensitive to both improved assay sensitivity and changes in assumptions about the risk of developing severe disease. When we assumed high assay sensitivity, seasonal screening by MP6-NAT was the only non-dominated strategy. When we assumed an estimate from the low end of the plausible range for risk of severe disease, the questionnaire strategy was least costly, although seasonal screening by ID-NAT of blood designated for transfusion of immunocompromised patients offered additional clinical benefit for an ICER of US$56,000/QALY gained. Discussion The recent emergence of WNV in the US has led to a perceived need to safeguard the blood supply from viremic blood donations. Strategies for screening blood for emerging viral infections such as WNV are often put into place without systematic evaluation of their costs, benefits, and cost-effectiveness. In this study, we conducted a cost-effectiveness analysis of alternative strategies for blood screening and considered the efficacy of these strategies in areas with varying epidemic intensity, exploring the effect of variable assay characteristics, transfusion outcomes, and pricing that may affect current and future policy decisions. Our analyses demonstrated that in areas with high infection rates, in the order of those seen in Mississippi and Nebraska in 2002, seasonal screening of blood designated for immunocompromised recipients prolongs quality-adjusted life expectancy compared with implementing a baseline questionnaire alone. Although other strategies, such as screening pooled samples designated for all donors, provided some benefit compared to a questionnaire alone, they were more costly and either less effective or only marginally more effective than restricted seasonal screening. In areas with low infection and seasonal transmission, none of the NAT strategies offered additional clinical benefit given current test-sensitivity estimates, although they were associated with substantial costs. These results suggest that the general screening of blood for WNV may not be as attractive a public health strategy as it first appeared to be, and that more restricted screening strategies may be preferable to currently mandated policies. The finding that blood-screening strategies for WNV may be outside the usually accepted cost-effectiveness thresholds is consistent with previous cost-effectiveness analyses for blood screening for infectious agents [21,29–32]. A recent analysis of NAT screening for hepatitis B and C and HIV compared to serological testing alone showed that the ICER exceeded US$1.5 million/QALY gained, well beyond the US$50,000–100,000 threshold commonly used as an indicator of willingness to pay for a health-care intervention [21]. AuBuchon et al. [29] have previously enumerated some of the reasons why cost-effectiveness estimates of blood-screening tests are so unattractive; risks are relatively low, transfusion recipients often have a reduced quality-adjusted life expectancy, and costs are incurred for all donations, few of which are infectious. Despite this, blood-screening tests are often implemented for reasons that are not captured in a cost-effectiveness analysis. There is a perception that blood recipients cannot be held responsible for avoiding risk, and therefore the system must protect them at any cost. Individuals are willing to pay more to avoid a catastrophic outcome, even when the risk is low compared to other outcomes. Furthermore, policy makers are more likely to apply an intervention to a small and defined group such as blood recipients rather than to a less-visible group [32]. Nonetheless, in an era of major cuts in public health expenditure and increasingly limited resources available for health care, it is worthwhile reconsidering the economic implications of this priority; resources spent preventing the rare case of transfusion-associated WNV might be better utilized in a host of other interventions against infectious disease, including those focused on reducing WNV transmission through mosquito vectors. If such an approach were successful, it might obviate the need for screening blood for the virus in many areas. Our analysis has a number of limitations as the ecology of WNV in the US, and the clinical course and sequelae of transfusion-acquired WNV infection, have not been clearly defined. Transmission intensities have varied over the years within some geographic regions since the emergence of WNV. Choosing the most cost-effective approach to screening within a specific area will depend on the ability to predict transmission intensity for the current season. Recently, a risk equation based on mosquito abundance, infectivity, vector competence, and host feeding behavior was developed to predict short-term future human WNV infections in an area [33]. The utility of this index to predict human infections is under investigation. Methods to validate, and improve upon, current prediction tools for WNV infection would enhance our ability to select the most cost-effective screening strategies. Improved methods to both measure and efficiently monitor the mosquito population parameters that determine virus transmission to humans would allow us to shift policy in response to important temporal changes in transmission patterns. Lacking other data, we estimated the risk of developing NI after an infected transfusion from a single study in which patients with cancer were deliberately inoculated with WNV [13]. These data may overestimate the true risk of disease, since the patients studied may have been more susceptible to severe disease than healthier blood recipients. Such a bias would exaggerate the benefits of screening in our analyses. Otherwise, if the potentially higher dose of virus from a transfusion-associated infection results in an even higher risk of developing NI than we estimated, the benefits of screening are underestimated in our analyses [10,11]. In addition, in the absence of large-cohort data, we assumed that data from small studies on the long-term clinical consequences of WNV-associated NI patients represent the expected sequelae of infection. Among the least well-defined parameters used in this analysis were those reflecting the performance of the newly introduced nucleic acid–screening tests. These were FDA-approved prior to establishing their sensitivity and specificity, and these characteristics have yet to be published. Given the imprecision of these estimates together with our expectation that the assay will improve with further product development, we repeated our analyses assuming that NAT for WNV was highly sensitive and specific. However, this analysis assumed that an improved test bore no additional cost. While various measures to enhance detection of low levels of viremia have been proposed, these would add further steps to screening rather than replace existing approaches, possibly adding substantially to expense. New methods that achieve small boosts in sensitivity (such as IgM antibody testing as an adjunct method to detect positive samples that have escaped NAT detection) are unlikely to be cost-effective under the assumptions made in this analysis. Custer et al. [34] demonstrated that while continuous ID-NAT screening would overburden blood-testing laboratories, ID-NAT screening during select times of the transmission season is needed currently, since minipool assays fail to detect 23% of the viremic samples detected by ID-NAT. In conclusion, we found that NAT screening of blood donations for WNV improved clinical outcomes only in those areas where the incidence of WNV is high, and that limiting screening to high-intensity transmission seasons and to blood donations designated for immunocompromised patients reduced costs without decreasing quality-adjusted life expectancy in most scenarios. We recommend that states adopt screening policies based on the intensity and duration of their WNV epidemics. Regional data, in conjunction with the results of this analysis and consideration of societal risk attitudes and preferences, may collectively point to a relaxation of the current federally mandated NAT screening of all donations in low-intensity areas. When high rates of natural infection indicate that NAT screening is appropriate, we recommend use of ID-NAT rather than minipool screening. States should consider the restricted screening of blood designated for immunocompromised patients alone. Finally, we suggest that blood-screening policies be carefully scrutinized for cost-effectiveness and that their relative contribution to safeguarding public health be considered in the making of policy decisions. Supporting Information Table S1 Predicted QALYs, Costs, Clinical Outcomes, and ICERs Associated with NAT Blood Screening for WNV in a Cohort of 2,000,000 High-sensitivity estimates. (82 KB DOC). Click here for additional data file. Table S2 Predicted QALYs, Costs, Clinical Outcomes, and ICERs Associated with NAT Blood Screening for WNV in a Cohort of 2,000,000 Low-severity estimates. (76 KB DOC). Click here for additional data file. Table S3 Predicted QALYs, Costs, Clinical Outcomes, and ICERs Associated with NAT Blood Screening for WNV in a Cohort of 2,000,000 High-severity estimates. (76 KB DOC). Click here for additional data file. Table S4 Predicted QALYs, Costs, Clinical Outcomes, and ICERs Associated with NAT Blood Screening for WNV in a Cohort of 2,000,000 Low-cost estimates. (76 KB DOC). Click here for additional data file. Table S5 Predicted QALYs, Costs, Clinical Outcomes, and ICERs Associated with NAT Blood Screening for WNV in a Cohort of 2,000,000 High-cost estimates. (76 KB DOC). Click here for additional data file. Table S6 Predicted QALYs, Costs, Clinical Outcomes, and ICERs Associated with NAT Blood Screening for WNV in a Cohort of 2,000,000 Shortened seasonal screening for high-infection/short-duration transmission area. (43 KB DOC). Click here for additional data file. Patient Summary Background West Nile virus (WNV) was first isolated from a sick woman in the West Nile region of Uganda in 1937. The virus has subsequently been found to be widespread in Africa and Eurasia, and sporadic outbreaks have been reported throughout these regions. WNV was first detected in the US in 1999, in a sick woman in New York. The disease has since spread to most states in the continental US, making thousands of people ill and causing several hundred deaths. Wild birds are the principal host of WNV, and the virus is transmitted to humans mainly by mosquitoes that bite both birds and humans. Most of the people who get infected by a mosquito bite do not get sick at all, but about 20% develop a flu-like illness. In a small number of cases—especially among the elderly and people with a weakened immune system—the infection spreads to the nervous system and can cause death or long-term disability. Like other blood-borne diseases, WNV can be transferred by blood transfusion with contaminated blood. Such cases have occurred in the US and have killed few people. Why Was this Study Done? WNV can be detected in blood samples by recently developed and approved tests. These tests detect most, but not all, cases of contamination with the virus. This means the WNV deaths resulting from transfusion of contaminated blood are potentially avoidable by screening donated blood. As a consequence, the US Food and Drug Administration (FDA) has mandated screening of donated blood samples. However, the FDA has not prescribed specific screening strategies, and the decision on how to best screen blood samples has been left to the individual states and the blood-collection agencies. The researchers who carried out this study wanted to determine which screening strategies would be cost-effective—that is, which strategies would prevent infections through contaminated blood for a reasonable price. In an ideal world, cost would not matter when it comes to protecting human life and health, but in reality there is limited money available for public health measures. Studies such as this one are therefore essential to help politicians decide how to spend the money. What Did the Researchers Do and Find? They calculated the costs of screening and the number of prevented infections through blood transfusion for a number of different scenarios. They found that in states with low WNV infection rates, the risk of an infected person donating blood was so low that screening was unlikely to prevent cases of serious illness from WNV, despite substantial costs. In states where WNV is common, screening throughout the year is likely to prevent cases of serious illness, but at a substantial cost. In states where WNV is common, screening blood only from May to the end of October (the months when mosquitoes are around and people get infected from them), however, was as effective at identifying contaminated blood samples as screening throughout the year. One way to reduce costs substantially was to create a separate blood pool that is reserved for transfusions to people with a weakened immune system and to screen only those samples. Because those are the people most at risk for severe WNV illness, this strategy would still prevent most of those cases. What Do These Findings Mean? It is not clear whether the current policy to screen all blood samples in all states makes sense from a health economy point of view. Restricting screening to states where WNV is common and to samples designated for people at higher risk for severe WNV illness would reduce costs significantly without putting the recipients of blood transfusions at a substantially higher risk of serious illness caused by WNV. Where Can I Get More Information Online? The following Web sites provide information on WNV. Pages from the US Centers for Disease Control and Prevention: http://www.cdc.gov/ncidod/dvbid/westnile/ Pages from the US National Biological Information Infrastructure: http://westnilevirus.nbii.gov/ MedlinePlus pages: http://www.nlm.nih.gov/medlineplus/westnilevirus.html US Food and Drug Administration pages: http://www.fda.gov/oc/opacom/hottopics/westnile.html US Geological Survey pages: http://westnilemaps.usgs.gov/ Funding for CTK was provided by the National Institutes of Health grants T32 AI007535 and R01 AI052284–02 and by Harvard School of Public Health, Department of Epidemiology. We would like to acknowledge Dr. Anne Labowitz and Dr. Denis Nash, both formerly of the New York City Department of Health and Mental Hygiene, for sharing their data on long-term follow-up of patients with West Nile virus. We would also like to thank the editors and independent reviewers (Dr. Brad Biggerstaff and Dr. Bruce Lee) for helpful comments. The funding agencies had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Citation: Korves CT, Goldie SJ, Murray MB (2006) Cost-effectiveness of alternative blood-screening strategies for West Nile virus in the United States. PLoS Med 3(2): e21. Abbreviations CDCCenters for Disease Control and Prevention FDAFood and Drug Administration ICERincremental cost-effectiveness ratio ID-NATindividual nucleic acid test MP6-NATnucleic acid testing of minipools of six samples MP16-NATnucleic acid testing of minipools of 16 samples NATnucleic acid test NIneuroinvasive disease QALYquality-adjusted life-year WNVWest Nile virus ==== Refs References Nash D Mostashari F Fine A Miller J O'Leary D The outbreak of West Nile virus infection in the New York City area in 1999 N Engl J Med 2001 344 1807 1814 11407341 Huhn GD Sejvar JJ Montgomery SP Dworkin MS West Nile virus in the United States: An update on an emerging infectious disease Am Fam Physician 2003 68 653 660 12952382 Centers for Disease Control and Prevention Update: West Nile virus screening of blood donations and transfusion-associated transmission—United States, 2003 MMWR Morb Mortal Wkly Rep 2004 53 281 284 15071426 Bren L West Nile virus: Reducing the risk FDA Consum 2003 37 20 27 12625302 Busch MP Dodd RY NAT and blood safety: What is the paradigm? Transfusion 2000 40 1157 1160 11061848 Roback JD CMV and blood transfusions Rev Med Virol 2002 12 211 219 12125013 Sachs JD Macroeconomics and health: Investing in health for economic development. Report of the Commission on Macroeconomics and Health 2001 Geneva World Health Organization 213 Available: http://www.un.oeg/esa/coordination/ecosoc/docs/RT.K.MacroeconomicsHealth.pdf . Accessed 27 December 2005. Gold ML Siegel JE Russel LB Weinstein MC Cost-effectiveness in health and medicine: The Report of the Panel on Cost-Effectiveness in Health and Medicine 1996 New York Oxford University Press 425 Vamvakas EC Taswell HF Long-term survival after blood transfusion Transfusion 1994 34 471 477 8023387 Biggerstaff BJ Petersen LR Estimated risk of West Nile virus transmission through blood transfusion during an epidemic in Queens, New York City Transfusion 2002 42 1019 1026 12385413 Biggerstaff BJ Petersen LR Estimated risk of transmission of the West Nile virus through blood transfusion in the US, 2002 Transfusion 2003 43 1007 1017 12869104 Mostashari F Bunning ML Kitsutani PT Singer DA Nash D Epidemic West Nile encephalitis, New York, 1999: Results of a household-based seroepidemiological survey Lancet 2001 358 261 264 11498211 Southam CM Moore AE Induced virus infections in man by the Egypt isolates of West Nile virus Am J Trop Med Hyg 1954 3 19 50 13114588 Vamvakas EC Goldstein R Four-year survival of transfusion recipients identified by hepatitis C lookback Transfusion 2002 42 691 697 12147020 Jeha LE Sila CA Lederman RJ Prayson RA Isada CM West Nile virus infection: A new acute paralytic illness Neurology 2003 61 55 59 12847156 Weiss D Carr D Kellachan J Tan C Phillips M Clinical findings of West Nile virus infection in hospitalized patients, New York and New Jersey, 2000 Emerg Infect Dis 2001 7 654 658 11589170 Sejvar JJ Haddad MB Tierney BC Campbell GL Marfin AA Neurologic manifestations and outcome of West Nile virus infection JAMA 2003 290 511 515 12876094 Pealer LN Marfin AA Petersen LR Lanciotti RS Page PL Transmission of West Nile virus through blood transfusion in the United States in 2002 N Engl J Med 2003 349 1236 1245 14500806 Pepperell C Rau N Krajden S Kern R Humar A West Nile virus infection in 2002: Morbidity and mortality among patients admitted to hospital in southcentral Ontario CMAJ 2003 168 1399 1405 12771068 Klee AL Maidin B Edwin B Poshni I Mostashari F Long-term prognosis for clinical West Nile virus infection Emerg Infect Dis 2004 10 1405 1411 15496241 Marshall DA Kleinman SH Wong JB AuBuchon JP Grima DT Cost-effectiveness of nucleic acid test screening of volunteer blood donations for hepatitis B, hepatitis C and human immunodeficiency virus in the United States Vox Sang 2004 86 28 40 14984557 Ratterree MS Gutierrez RA Travassos da Rosa AP Dille BJ Beasley DW Experimental infection of rhesus macaques with West Nile virus: Level and duration of viremia and kinetics of the antibody response after infection J Infect Dis 2004 189 669 676 14767821 Fryback DG Dasbach EJ Klein R Klein BE Dorn N The Beaver Dam Health Outcomes Study: Initial catalog of health-state quality factors Med Decis Making 1993 13 89 102 8483408 Institute of Medicine Vaccines for 21st Century: A tool for setting priorities 2001 Washington (DC) National Academies Press 460 Livartowski A Boucher J Detournay B Reinert P Cost-effectiveness evaluation of vaccination against Haemophilus influenzae invasive diseases in France Vaccine 1996 14 495 500 8782346 Shehata N Kohli M Detsky A The cost-effectiveness of screening blood donors for malaria by PCR Transfusion 2004 44 217 228 14962313 Villari P Spielman A Komar N McDowell M Timperi RJ The economic burden imposed by a residual case of eastern encephalitis Am J Trop Med Hyg 1995 52 8 13 7856830 Utz JT Apperson CS MacCormack JN Salyers M Dietz EJ Economic and social impacts of La Crosse encephalitis in western North Carolina Am J Trop Med Hyg 2003 69 509 518 14695088 AuBuchon JP Birkmeyer JD Busch MP Cost-effectiveness of expanded human immunodeficiency virus-testing protocols for donated blood Transfusion 1997 37 45 51 9024489 Busch MP Dodd RY Lackritz EM AuBuchon JP Birkmeyer JD Value and cost-effectiveness of screening blood donors for antibody to hepatitis B core antigen as a way of detecting window-phase human immunodeficiency virus type 1 infections. The HIV Blood Donor Study Group Transfusion 1997 37 1003 1011 9354817 Jackson BR Busch MP Stramer SL AuBuchon JP The cost-effectiveness of NAT for HIV, HCV, and HBV in whole-blood donations Transfusion 2003 43 721 729 12757522 AuBuchon JP Birkmeyer JD Busch MP Safety of the blood supply in the United States: Opportunities and controversies Ann Intern Med 1997 127 904 909 9382369 Kilpatrick AM Kramer LD Campbell SR Alleyne EO Dobson AP West Nile virus risk assessment and the bridge vector paradigm Emerg Infect Dis 2005 11 425 429 15757558 Custer B Tomasulo PA Murphy EL Caglioti S Harpool D Triggers for switching from minipool testing by nucleic acid technology to individual-donation nucleic acid testing for West Nile virus: Analysis of 2003 data to inform 2004 decision making Transfusion 2004 44 1547 1554 15504158 Goldblum N Jasinska-Klingberg W Klingberg MA Marberg K Sterk VV The natural history of West Nile Fever. I. Clinical observations during an epidemic in Israel Am J Hyg 1956 64 259 269 13372519 Hannoun C Corniou B Causse G Panthier R Development of serum antibodies in 4 cases of West Nile virus infection Ann Inst Pasteur (Paris) 1967 113 29 36 6076185	16381598	PMC1324950	CC BY	2021-01-05 11:14:05	no	PLoS Med. 2006 Feb 24; 3(2):e21	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030021	oa_comm
==== Front PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1637949910.1371/journal.pmed.0030027Research ArticleInfectious DiseasesMicrobiologyVirologyPathologyRespiratory MedicineInfectious DiseasesLung functionPathologyRespiratory MedicineTravel MedicineTime Course and Cellular Localization of SARS-CoV Nucleoprotein and RNA in Lungs from Fatal Cases of SARS SARS Time Course and LocalizationNicholls John M 1 Butany Jagdish 2 Poon Leo L. M 3 Chan Kwok H 3 Beh Swan Lip 1 Poutanen Susan 4 Peiris J. S. Malik 3 Wong Maria 1 1Department of Pathology, The University of Hong Kong, Pok Fu Lam, Hong Kong SAR, China2Department of Pathology, University of Toronto, Toronto General Hospital, Toronto, Ontario, Canada3Department of Microbiology, The University of Hong Kong, Pok Fu Lam, Hong Kong SAR, China4Department of Microbiology, Toronto Medical Laboratories and Mount Sinai Hospital, University of Toronto, Toronto, Ontario, CanadaZaki Sherif Academic EditorCenters for Disease ControlUnited States of America* To whom correspondence should be addressed. E-mail: [email protected] (JMN); E-mail: [email protected] (MW) Competing Interests: The authors have declared that no competing interests exist. Author Contributions: JMN designed the study. JB analyzed the data. KHC contributed to writing the paper. JMN, MW, and JSMP are the principal investigators, with overall responsibility for the design of the study and writing of the report. SLB coordinated the autopsy organization and collection of autopsy materials of SARS deaths in Hong Kong. LLMP designed the protein expression constructs and transfection studies. LLMP and JSMP were involved in the development of the RT-PCR assay. JMN and KHC were involved in the development of the SARS-CoV IHC assay. SP was involved in the treatment of the Canadian SARS cases and the coordination of obtaining autopsy specimens from those who died. All authors were involved in the correlative interpretation of the clinical, pathological, and molecular data. 2 2006 3 1 2006 3 2 e2723 8 2005 24 10 2005 Copyright: © 2006 Nicholls et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Pathological Clues to How the SARS Virus Kills Background Cellular localization of severe acute respiratory syndrome coronavirus (SARS-CoV) in the lungs of patients with SARS is important in confirming the etiological association of the virus with disease as well as in understanding the pathogenesis of the disease. To our knowledge, there have been no comprehensive studies investigating viral infection at the cellular level in humans. Methods and Findings We collected the largest series of fatal cases of SARS with autopsy material to date by merging the pathological material from two regions involved in the 2003 worldwide SARS outbreak in Hong Kong, China, and Toronto, Canada. We developed a monoclonal antibody against the SARS-CoV nucleoprotein and used it together with in situ hybridization (ISH) to analyze the autopsy lung tissues of 32 patients with SARS from Hong Kong and Toronto. We compared the results of these assays with the pulmonary pathologies and the clinical course of illness for each patient. SARS-CoV nucleoprotein and RNA were detected by immunohistochemistry and ISH, respectively, primarily in alveolar pneumocytes and, less frequently, in macrophages. Such localization was detected in four of the seven patients who died within two weeks of illness onset, and in none of the 25 patients who died later than two weeks after symptom onset. Conclusions The pulmonary alveolar epithelium is the chief target of SARS-CoV, with macrophages infected subsequently. Viral replication appears to be limited to the first two weeks after symptom onset, with little evidence of continued widespread replication after this period. If antiviral therapy is considered for future treatment, it should be focused on this two-week period of acute clinical disease. The SARS coronavirus targets primarily the pulmonary alveolar epithelium. Viral replication seems limited to the first two weeks after symptom onset and restricted to the lungs. ==== Body Introduction Severe acute respiratory syndrome (SARS) is a new disease that originated in the Guangdong province of China in November 2002, and subsequently spread to Hong Kong in February 2003 [1,2]. Facilitated by busy international traffic, it soon caused major disease outbreaks in multiple cities around the world over the following five months [3,4]. It had an alarming tendency to spread to health care workers, and unlike most respiratory viral infections, predominantly involved the lower respiratory tract. The disease was associated with an unusually high morbidity and mortality; 20%–30% of patients developed severe respiratory symptoms requiring intensive care support [2,3,5]. Worldwide, the average case mortality rate was 10.0%, but in Hong Kong and Toronto, the case mortality rate was 17.0%, with older patients and those with chronic illnesses having a worse outcome [1,2,5]. The causative agent has now been determined to be a novel coronavirus (SARS-CoV) that is genetically distinct from any previously identified coronavirus known to cause disease in animals or humans [1,6–8]. In a primate model system, experimental infection of cynomolgus macaques with SARS-CoV alone was sufficient to cause a disease complex similar to that observed in humans [9,10]. Genetically, the coronavirus species closest to SARS-CoV are those isolated from small animals traded for culinary purposes, such as the Himalayan palm civet and raccoon-dog, which show 99.8% sequence homology but differ by the insertion of a 29-nucleotide fragment [11]. It is postulated that both coronaviruses have probably arisen from a common natural reservoir, with facilitated interspecies transmission and possibly genetic alteration to SARS-CoV during the close handling and trading of the animals in the same environment. The pulmonary pathology observed in limited postmortem material has been reported together with the clinical symptomatology of SARS [2,8,12–15]. The detection of the coronavirus in relation to pulmonary changes is a critical factor in the understanding of the pathogenesis of the disease. However, this has been hampered by the lack of readily available monoclonal anti-SARS antibodies, as well as the impracticability of extensive electron microscopic examination. Furthermore, although investigation of the time course of SARS-CoV infection with respect to the detection of SARS-CoV RNA has been done by PCR methods, there have been no systematic, large-scale investigations of the temporal or cellular detection of SARS-CoV in lung tissues. Additionally, the extent of organ distribution has not been examined in great detail. Indeed, a recent single-case publication on the immunohistochemical detection of SARS highlighted “…the need for more cases of SARS to be studied to determine the temporal relationship between the duration of illness and viral clearance in human lung tissue…” [15]. This study therefore had three purposes. The first was to investigate the presence of SARS-CoV by immunohistochemistry (IHC) and in situ hybridization (ISH) in lung autopsy specimens from patients who died at different time points of the disease. The second purpose was to determine the cellular distribution of SARS within the lung, and the third was to examine extrapulmonary tissues of patients who died of SARS to determine the extent of systemic distribution. Methods Patient Profiles and Material A total of 32 patients who died with SARS-CoV infection confirmed by PCR, serological, or viral culture tests were included in the study. The patients came from two geographical regions affected by SARS (Hong Kong [HK] and Toronto, Canada [TO]). For HK cases 1 and 10–13, whole lung specimens were perfused at autopsy with 4% neutral buffered formalin at 20–30 cm H2O pressure until the lungs were fully expanded, before immersion in formalin fixative. Multiple blocks were extensively sampled from all lobes of the lungs for HK cases 1, 6, 7, and 9–13; limited representative blocks were sampled from HK cases 2–5 and 8. For all Toronto cases, approximately half of each lung lobe was removed at autopsy, and a portion was snap-frozen for molecular studies. Multiple blocks of tissue were processed from each lobe for light microscopic examination. Stains for microorganisms and for collagen, and immunohistochemical stains for microorganisms, were performed as necessary. The clinical profiles were retrieved from the patients' records. HK case 1 was a patient who died in the quarantine period for SARS exposure during hospitalization for congestive heart failure. PCR of a throat swab was positive on the fifth day after symptom onset, and she developed fever and respiratory symptoms thereafter. HK case 13 was a patient who died unexpectedly during convalescence from SARS. All other patients died of SARS with progressive deterioration in lung function. All HK patients except cases 1 and 13 had received ventilatory support during life. The pulmonary histopathological features of six of the HK cases (2–6 and 8) have been previously reported [12]. In addition to the autopsy tissue from all HK cases, HK case 5 also had an open lung biopsy, which was also included in this study. Descriptions of the procurement and clinical courses of the TO cases have been described in previous reports [14,16]. Development of a Monoclonal Antibody against SARS-CoV Nucleoprotein BALB/c mice were immunized intraperitoneally with 0.1 ml of heat-killed SARS-CoV HKU39849-infected FRhK4 cell lysate (107 TCID50/ml). Injections of similar doses were repeated biweekly for 2 mo. Four days after the last booster, 108 spleen cells from an immunized mouse were fused with 107 of NSI myeloma cells with polyethylene glycol (PEG, molecular weight 4,000; BDH, Poole, United Kingdom) as the fusing agent. Hybridomas were screened for production of antibodies against SARS-CoV HKU39849-infected cells and recombinant SARS-CoV nucleocapsid protein by ELISA. Those that produced SARS-CoV nucleoprotein-specific antibodies were cloned twice by limiting dilution. Purified hybridomas were then injected intraperitoneally into mineral oil-primed mice for the production of ascitic fluid. Monoclonal antibodies (4D11 and 3E4) were purified from the ascitic fluid by precipitation with 50% ammonium sulfate and subcloned to ensure monoclonality. Development of Transfection Controls The open reading frames of spike (S), envelope (E), and N genes of SARS-CoV were cloned into the AgeI site of a protein expression vector, pcDNA3A [17]. To express the above viral proteins in eukaryotic cells, these protein expression plasmids were transfected into 293T cells individually or in appropriate combinations. Briefly, 1 μg of each of the plasmids was transfected into 293T cells by Lipofectamine 2000 (Invitrogen, Carlsbad, California, United States) as instructed by the manufacturer. At the indicated posttransfection time points, transfected cells were harvested and washed with PBS, then centrifuged at 1,600 rpm for 10 min. Pelleted cells were fixed in PBS with 0.1% glutaraldehyde and 2% formaldehyde, and stored at 4 °C. SARS-CoV S and E gene-transfected cells were used as negative controls, and SARS-CoV N gene-transfected cells were used as positive controls for IHC to confirm the specificity of the monoclonal antibodies against SARS-CoV nucleoprotein. Immunohistochemistry The primary antibodies used included those against cytokeratin (1:50, clone AE1/AE3 [Dako, Glostrup, Denmark]), CD68 (1:50; clone KP1, Dako), EMA (clone E29, Dako), thyroid transcription factor 1 (TTF1; clone 8G7G3/1, Zymed/Invitrogen), chromogranin (clone LK2H10 [Ventana, Tucson, Arizona, United States]), DC-SIGN (a gift from Professor J-L Virilizier, Pasteur Institute, Paris, France), LCA (clone T29/33, Dako), and SARS-CoV N (as described above) (1:400, clone 4D11). Antigen retrieval was performed by microwaving sections in 10 mM citrate buffer (pH 6.0) for 15 min and incubating with 1:200 4D11 antibody at 4 °C overnight. Secondary labeling was performed with biotinylated rabbit anti-mouse (Dako #E-0354) at 1:100 for 30 min at room temperature and was followed by incubation with streptavidin-ABC complex (Dako #K-0377) at 1:100 for 30 min at room temperature and color development by the 3-amino-9-ethylcarbazole (AEC) substrate kit (Vector Laboratories, Burlingame, California, United States; #SK-4200) at room temperature (15–30 min). For double labeling of lung tissue sections, the 4D11 antibody was labeled with FITC, and a TRITC anti-mouse secondary antibody was used. Sections were microwaved in 10 mM citrate buffer (pH 6.0) for 15 min, blocked with 10% normal donkey serum for 10 min at room temperature, incubated with the non-SARS monoclonal overnight, then incubated with TRITC-conjugated donkey anti-mouse antibody at 1:100 for 1 h at room temperature. The FITC-conjugated 4D11 antibody was incubated at 1:100 for 1 h at room temperature, followed by counterstaining of the nuclei with DAPI for 4 min and mounting with DAKO fluorescence mount (Dako #S-3023). Examination was performed with a Nikon Eclipse E-800 fluorescent microscope with a dual FITC/rhodamine filter. In Situ Hybridization We produced digoxigenin-labeled antisense riboprobes specific for the N, E, and S genes of SARS-CoV, and these were pooled and hybridized with 30 μg/ml proteinase K-treated lung sections for 30 min at room temperature. The hybridization buffer (250 μg/ml salmon sperm DNA, 125 μg/ml rat total RNA, 200 mg/ml yeast tRNA, 50% deionized formamide, 10% dextran sulfate, 1× Denhardt's solution, 1 mM EDTA [pH 8.0], 0.01% sodium pyrophosphate, 0.3 M sodium chloride, and 10 mM Tris-HCl [pH 8.0]) was mixed with the probes and hybridized overnight at 50°C. The hybridized probes were detected with mouse anti-digoxigenin at 1:100 in 10% normal rabbit serum for 1 h at room temperature, then biotinylated rabbit anti-mouse (Dako, E-0354) at 1/100 for 30 min. at room temperature, followed by detection with AEC substrate kit (Vector, SK-4200) at room temperature (up to 30 min). Prior to the tests, the sensitivity of the SARS-CoV probes were verified semiquantitatively by hybridization to separate cell block preparations of a SARS-CoV infected FRHK4 cell line harvested at 8 and 24 h. RT-PCR SARS-CoV RT-PCR was completed on fresh (HK cases 1 and 10–13), fixed (HK cases 2–9), and snap-frozen (all TO cases) lung tissues from all patients. For the HK cases, total RNA was extracted by standard methods and reverse transcribed to cDNA with reverse transcriptase. PCR was performed with initial denaturation at 94 °C for 8 min followed by 40 cycles of 94 °C for 1 min, 50 °C for 1 min, and 72 °C for 1 min using primers as previously described [1,5,12]. For the TO cases, RT-PCR was completed using RealArt HPA-Coronavirus LC-RT assay (Artus, Hamburg, Germany) as previously described [16,18]. Results The HK and TO cases' clinical profile, treatment history, and SARS-CoV IHC, ISH, and RT-PCR results are summarized in Table 1. Table 1 Profile of Cases Listed According to Duration of Disease from Symptom Onset for Each Location Lung Pathology HK patient 1, who died 5 d after symptom onset during quarantine for SARS contact, showed moderate interstitial and alveolar edema with occasional epithelial desquamation and regeneration. A moderate number of macrophages with abundant foamy cytoplasm had accumulated in the alveolar spaces, but giant cells were not observed. In addition, features of aspiration pneumonia were present, involving small bronchi and adjacent parenchyma in all lobes of the lung. HK patients 2–12 showed established changes indicating diffuse alveolar damage (DAD) at different phases of disease progression that affected the lungs to varying extents. Lung pathology of the TO cases has been described previously [14]. Although most of the advanced histological changes revealing DAD can be attributed to ventilation, it should be pointed out that there were also changes of DAD present in five of the TO cases in which no intubation was carried out; thus, viral damage is therefore the most likely mechanism of these changes. SARS-CoV IHC, ISH, and RT-PCR Results The transfected cells showed strong cytoplasmic staining for transfectants containing the N gene with no staining on the M or S gene transfectants (Figure 1), confirming the specificity of the monoclonal antibody for the N protein. The control sections of FRhK4 cells showed cytoplasmic staining that was focal at 8 h and diffuse at 24 h. Figure 1 Confirmation of the Specificity of the Monoclonal Antibody 4D11 for the SARS-CoV Nucleocapsid Protein SARS-CoV immunohistochemical staining of 293T cells transfected with spike (S), nucleocapsid (N), and spike, nucleocapsid, membrane, and envelope (SEMN) genes of SARS-CoV. Untransfected control is shown in the lower right photomicrograph. The monoclonal antibody 4D11 identifies the N protein. AEC stain with hematoxylin counterstain; magnification 100×. Four cases showed positive staining for N protein by IHC (HK case 1 and TO cases 1, 4, and 5) (Figure 2). All four of these cases died within 14 d of symptom onset. No staining of the bronchial epithelium was present (Figure 2A), and there was only focal staining of the bronchiolar epithelium (Figure 2B), which showed no evidence of necrosis or regeneration. This observation was confirmed by electron microscopy with viral particles identified in ciliated cells. There was cytoplasmic staining of flattened type 1 pneumocytes as well as alveolar macrophages (Figure 2C and 2D). Adjacent hilar lymph nodes were available for examination in one case (HK case 1), and, of six nodes sampled from this patient, only one isolated mononuclear cell was noted that stained positive for N protein but was negative for the N gene by ISH. Although the staining of the pneumocytes was diffusely cytoplasmic, the staining of the alveolar macrophages was coarsely granular (Figure 2F). For all patients who died beyond 14 d after symptom onset, only scattered single positive cells were seen in lung sections (Figure 2G and 2H), and these were in flattened pneumocytes as well as in mononuclear cells. In one of the Toronto cases (TO case 1) a thrombus was identified (Figure 2E), and positive elongated cells as well as rounded mononuclear cells were identified in this thrombus. This thrombus was present in only one section, and further labeling to confirm the etiology of the infected cells could not be performed. None of the multinucleated epithelial cells seen in any of the patients were positive by ISH or IHC. Figure 2 Examples of SARS-CoV IHC/ISH Results No staining for N protein is present in the bronchus of HK case 1 (A), but there is focal positive staining of bronchiolar epithelium in this case (B). Positive staining is seen in epithelial cells and detached cells in the alveoli in TO case 1 (C), which on higher magnification morphologically resemble type 1 pneumocytes (D). In TO case 2, there is a thrombus present which contains mononuclear and spindle-shaped cells (E). Carbon-containing macrophages are also positive (F). Ten days after symptom onset, positive staining was reduced and is noted mainly in the exudates (G) with only very focal positive mononuclear cells seen (H). AEC stain with hematoxylin counterstain; magnification 100×. In the four cases that stained positive for N protein by IHC, double labeling with the directly conjugated FITC SARS monoclonal antibody and antibodies for macrophages and epithelial cells showed that the infected cells were positive for EMA (anti-epithelial) and CD68 (anti-macrophage) (Figure 3A and 3B). The number of infected cells was greatest in TO case 1 and HK case 1, both of whom died 5 d after symptom onset. Of interest, TO case 1 showed positive staining of epithelial cells but no staining of macrophages (Figure 3C and 3D). Macrophages showed positive coarse granular staining in HK case 1 and TO cases 4 and 5, who died 11 and 14 d after symptom onset, respectively (Figure 3E). No staining was seen in chromogranin-positive or DC-SIGN-positive cells, and no colocalization with TTF1 (a marker for type 2 pneumocytes) was seen. No staining of lymphocytes was identified. TO case 5, who died on the day 14 after symptom onset, had only scattered single positive cells, and these were in flattened epithelial cells and mononuclear cells. Figure 3 Colocalization of SARS-CoV Infection of Human Lung Tissues Shown are tissue samples stained with SARS-CoV monoclonal nucleocapsid FITC and epithelial membrane antigen TRITC. At day 5 postinfection, there is positive staining of type 1 pneumocytes (A), with epithelial cell debris present in the alveolar lumen (B). In one case (TO case 1), there is positive staining for SARS-CoV of the type 1 pneumocytes with no staining of CD68 positive macrophages (C and D). HK case 1, however, shows FITC staining within the cytoplasm of CD68-positive cells (E). The same four cases that showed positive staining for N protein by IHC (HK case 1 and TO cases 1, 4, and 5) were also positive by ISH for the pooled S, E, and N SARS-CoV genes. Multiple lung samples from HK case 1 showed strong cytoplasmic signals in alveolar cells, including desquamating pneumocytes and alveolar macrophages. Weak signals were detected in occasional macrophages scattered along the bronchial epithelium. No signals were detected in endothelial or stromal cells. In addition to the four cases that stained positive for SARS-CoV by IHN and ISH, a few SARS-CoV infected cells were also detected by ISH and IHC in an open lung biopsy taken 10 d after symptom onset from HK case 5; however, both ISH and IHC were negative on the autopsy lung sample taken from this patient after death (20 d after symptom onset). Results of RT-PCR are listed in Table 1. Autopsy lung samples from all but three HK cases were positive for SARS-CoV by RT-PCR. When IHC and ISH staining was compared with RT-PCR viral load and disease duration, positive cells by IHC and ISH were identified mainly in cases who had died less than 14 d from onset of illness and in whom lung tissues were associated with viral loads greater than 105 copies per gram of tissue. In two of the cases that were positive for IHC in the lung, other organs were examined for IHC. In HK case 1, when there was strong positive IHC and ISH staining of the lung identified, no IHC- or ISH-positive cells were identified in other organs examined (adrenal gland, kidney, spleen, liver, and heart). In TO case 4, which also showed positive lung staining, no other organs were positive by IHC or ISH (liver, spleen, and kidney). Examination of three other TO cases that stained negative in the lung also showed no positive extrapulmonary staining. Discussion SARS-CoV nucleocapsid protein and RNA were detected in lung autopsy samples from four of the seven patients who died within 2 wk after illness onset, all of whom had high (greater than 105 copies per gram) RT-PCR viral loads in the autopsy lung tissue. SARS-CoV nucleocapsid protein and RNA were also detected in an open lung biopsy from a patient on day 5 after symptom onset (before death), but not in the same patient's autopsy lung specimen collected on day 20 after symptom onset. Among those lung samples staining positive by ISH or IHC, the signals localized mainly to the alveolar epithelial cells, with lower intensity in alveolar macrophages, indicating that the former are the chief target cells of the virus. Only scattered positive cells were identified in the bronchiolar epithelium, and no significant staining of the bronchial epithelium was observed. No distinctive spread to regional lymph nodes was seen, and we were not able to determine colocalization with DC-SIGN–positive cells. Using an oligonucleotide probe with signal amplification, Nakajima et al. demonstrated SARS-CoV genomes in the alveolar epithelium and macrophages of a 46-year-old woman with SARS, but details of the clinical course of the disease were not stated [19]. The preferential localization of the virus to alveolar cell components relative to the bronchial epithelium seen at the time of death suggests that the alveolar environment might be more permissive to viral replication, and could also account for the high incidence of lower respiratory tract involvement. It should be noted, however, that there are no studies of SARS in humans in which bronchial or tracheal biopsy material from nonfatal cases has been examined. Therefore, whether there is SARS-CoV replication in these parts of the respiratory tract, as is the case in nonhuman primates [10], remains unanswered. Among the 25 patients who died later than two weeks after symptom onset, no ISH or IHC signals were detected in postmortem lung tissues, even in lesions that were morphologically earlier, such as alveolar edema and hyaline membrane or in lungs showing heterogeneous morphological progression, where relatively uninvolved areas were interspersed among more advanced lesions. These findings suggest that there is a decrease of SARS-CoV replication and, consequently, a decrease in intracellular viral copy numbers in lung tissue after the first two weeks of disease, with the terminal event not dependent on continued widespread viral replication. It has also been noted that viral loads detected by RT-PCR or viral culture from nasopharyngeal aspirates, urine, and stool samples start to decrease 10–15 d after symptom onset, correlating with the time taken for the development of specific anti-SARS antibodies, which starts approximately 10 d from disease onset [5,20]. The identification of the cells infected by SARS-CoV in humans has yielded contradictory findings. To et al., using ISH combined with IHC, found dual labeling of only epithelial cells using AE1/AE3 and negative staining for macrophages (antibody CD68) [21]. In contrast, Chen and Hsiao, using ISH, demonstrated a positive signal in carbon-containing macrophages and also within vascular lumina [22]. A later report, also from Taiwan and using ISH, demonstrated positive signal in pneumocytes and not in macrophages, but it is not clear if the same patient material (from a 36-year-old woman) was used in these last two publications [23]. The discrepancies between these findings may be partially explained by three factors: the choice of probes, the age of the patient, and the disease progression. To et al. [21] used the membrane (M) gene, whilst Chow et al. [23] used a mixture of probes (N, M, and replicase [REP]) and found the greatest intensity with the N and M gene and minimal staining with the REP. Our findings support those of both To et al. and Chow et al., in that there was mainly epithelial staining but also definite macrophage (CD68) dual staining, suggesting that, although epithelial cells are the main target of SARS-CoV, the presence of viral RNA in macrophages represents either phagocytosis or low-level replication in these cells. None of these previous ISH-based studies have demonstrated ISH-positive lymphocytes, as reported by Gu et al. [24]. Our data also suggest that pneumocytes may be infected first, followed by macrophages; TO case 1, who died 5 d after symptom onset, had staining of pneumocytes but no other cells, whereas all other cases, who died five or more days after symptom onset had both pneumocyte and macrophage staining. It is possible that the material used for the To et al. study came from early cases. Shieh et al. describe finding type 2 pneumocytes positive for SARS-CoV in one patient [15]. However, in this study, staining appeared to be restricted to morphological type 1 pneumocytes, in keeping with the findings of Haagmans et al. [25]. However, Haagmans et al. were unable to demonstrate macrophage uptake of SARS-CoV [25]. We also found occasional bronchiolar epithelial cells that were positive for SARS-CoV by ISH, IHC, and electron microscopy, which has not been noted in other studies. To this extent, the immunohistochemical findings are similar to those seen with respiratory syncytial virus (RSV) infections, in which there is staining of bronchial epithelial cells, pneumocytes, and macrophages [26]. The age of the patient is also important, as SARS has been associated with greater mortality in the elderly than in the younger age group, and laboratory studies have shown that “aged” mice infected with SARS-CoV had more severe disease and greater cytokine production than younger mice [27]. Thus, differences in age may also be responsible for some of the differences noted between published results. The spectrum of pulmonary lesions in the cases described in this study is characteristic of the acute and resolving stages of acute lung injuries, including DAD due to intrapulmonary or extrapulmonary causes such as bacterial and viral infection, trauma, shock, etc. Unlike the findings of Gu et al., who demonstrated ISH-positive cells in giant cells [24], we repeatedly found no evidence of positive staining by ISH or IHC in the giant cells in our material. As giant cells were seen mainly in patients after 14 d of disease duration, whether they have the same mechanism of formation as seen in animal models needs further investigation. Interestingly, multinucleated epithelial cells have been noted in lung infections due to other viruses such as measles, parainfluenza, and respiratory syncytial viruses. In fact, none of the changes observed in the present cases is unique for SARS-CoV infection. The pathophysiological mechanisms underlying DAD and subsequent pulmonary fibrosis have been investigated in clinical and experimental models [28,29]. It has been shown that proinflammatory and fibrogenic cytokine pathways are activated within the first 24–48 h following pulmonary insult [30,31]. High initial levels of these activities are associated with persistent pulmonary damage and increased risks of subsequent pulmonary fibrosis and poor outcome in DAD [32–34]. Previous studies have published viral loads per gram of tissue using quantitative PCR. This quantitative PCR method has been shown to be of value in determining viral load of viruses in peripheral blood, such as Epstein-Barr virus; but unlike blood, lung tissue is not homogeneous and, in an assessment of viral copies per gram of tissue, whether the tissue sample contains bronchus, scar tissue, edema fluid, or exudates is crucial. In addition, blood in tissues may lead to false-positive PCR results due to viremia rather than viral replication in the tissue. In this study, persons with high SARS-CoV viral loads in their lung tissue had good correlation with the IHC and ISH results [18], but lower lung and organ tissue viral loads showed negative results. These lower viral loads may reflect a small amount of residual genome from a previous infection comparable to RSV infections reported in both stable as well as acute exacerbation of chronic obstructive pulmonary disease patients [35]. This is a possible explanation for the low level of RT-PCR positivity in the lung tissues and in other tissues in SARS patients in which we could not detect SARS-CoV by IHC or ISH. Furthermore, in the fatal cases of SARS most patients had seroconverted at the time of death (unpublished data). However, we cannot exclude the possibility of low-level replication in these tissues that was not detectable by IHC and ISH, due to possible limitations of the sensitivity of these assays. This study has shown that alveolar epithelium and macrophages are the chief targets of SARS-CoV infection in the human lung, in agreement with the findings reported in a primate model of the disease [10]. The predominant infiltrating cell in the lungs were macrophages, and this agrees with finding of high levels of macrophage-trophic chemokines in vivo, and in response to SARS-CoV infection in vitro [36]. However, the relative contributions of direct viral damage versus immunopathology in the pathogenesis of SARS remains unclear. Detectable levels of intracellular viral RNA and proteins in lung tissue are present in only the earliest phase of the disease, and they disappear by 2 wk after the onset of symptoms. Although we found only four positive cases by ISH and IHC, and all these were in patients who died within the first 2-wk period, it is likely that many of the negative postmortem cases would have been positive had a lung biopsy been performed earlier. Lack of IHC and ISH staining of extrapulmonary tissues in fatal cases of early SARS suggest that viral involvement of major organs may not be as widespread as data based on RT-PCR results have previously suggested. Our findings have important implications on the clinical and therapeutic management of SARS should it return in the future, in that if antiviral therapy is to be instituted, there may be a broader window of opportunity for intervention with antiviral therapy than there is, for example, with influenza, in which treatment within 48 h is required for discernible clinical impact. Patient Summary Background The first SARS (severe acute respiratory syndrome) outbreak started in November 2002 in China, and over the next five months spread to a number of other cities around the world. During that time, 20%–30% of the infected people became seriously ill, and 10% died. SARS is caused by the SARS-CoV, which infects the lungs of patients and leads to breathing problems. Since the outbreak, scientists around the world have been working hard to understand the SARS virus and how it causes disease, and to develop drug treatments and vaccines. Why Was This Study Done? For this study, researchers from Hong Kong and Toronto (both cities hit by the SARS outbreak) joined forces for a large and detailed study of the bodies of patients who had died from SARS. They hoped that this would help them to understand how the virus had killed the patients, and when and how certain treatments might be most effective. The researchers wanted to see whether there were differences between individuals who died from SARS at different times after they first fell ill. They were also interested in the distribution of SARS-infected cells in the lungs of these patients, and whether they could detect the SARS virus outside of the lung in other parts of the body. What Did the Researchers Do and Find? They studied the organs of 32 patients who had died from SARS. The researchers had detailed clinical information on each of them. Some had died less than a week after they became ill, and others had died after more than a month. The researchers used a variety of modern scientific tools to visualize exactly where in the bodies the SARS virus had reached. They detected the virus in the lungs of most patients who died within two weeks of falling ill, but not in the lungs of any of the patients who died after more than two weeks of illness. Where detectable, the virus was mostly found in an area of the lung called the alveolar epithelium. Each of your lungs is made up of a network of breathing tubes, and each tube ends in a tiny pouch called an alveolus, which is surrounded by blood vessels. When you breathe in oxygen, it eventually passes into your bloodstream across the lining of these pouches, and this lining is the alveolar epithelium. The researchers also examined other organs (including heart, kidney, and liver) from two of the patients whose lungs had tested positive for the virus and three others who had shown no sign of virus in the lungs. They did not detect the virus in any of these samples. What Does This Mean? These results suggest that the SARS virus causes illness and death by attacking cells in the alveolar epithelium. The virus multiplies mostly within this group of cells, and only for a limited time after infection. Antiviral drugs are likely to be only useful during this initial time window after infection. After less than two weeks, the body's immune system seems to be able to fight back and prevent the virus from further multiplying, but this does not lead to recovery in all patients. In fact, most of the patients in this study (25 out of 32) died more than two weeks after they became ill and without signs that the SARS virus was still multiplying in their body. This suggests that in patients who die, the initial damage to the lungs is so severe that they cannot recover, and this is not dependent on continued virus replication. The study also suggests that death from SARS is not due to the virus multiplying outside of the lungs. Where Can I Find More Information Online? The following Web sites provide information on SARS. Pages from the Government of Hong Kong: http://www.info.gov.hk/info/sars/eindex.htm Health Canada pages: http://www.phac-aspc.gc.ca/sars-sras/index.html Pages from the US Centers for Disease Control and Prevention: http://www.cdc.gov/ncidod/sars/ World Health Organization pages: http://www.who.int/csr/sars/en/ MedlinePlus pages: http://www.nlm.nih.gov/medlineplus/severeacuterespiratorysyndrome.html We thank Kevin Fung for expert immunohistochemical and hybridization technical support. We also acknowledge staff of the Queen Mary Hospital, United Christian Hospital, Kwong Wah Hospital, Princess Margaret Hospital, Queen Elizabeth Hospital, and Tuen Mun Hospital, Hong Kong, for facilitating case collection in Hong Kong. We also acknowledge the Office of the Chief Coroner of Ontario, Canada for facilitating case collection in Toronto, Canada, and we acknowledge G. Farcas, T. Mazzulli, B. Willey, S. Asa, P. Faure, P. Akhavan, D. E. Low, and K. C. Kain for their help in coordinating and completing SARS-CoV RT-PCR on the Canadian SARS patients' tissues; and the Canadian SARS Research Network for facilitating clinical chart reviews. This work was supported by the Vice Chancellor's Fund for SARS Research 2003 granted by The University of Hong Kong (21395059 and 21395052) and the National Institute of Allergy and Infectious Diseases, United States (public health research grant A195357). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Citation: Nicholls JM, Butany J, Poon LLM, Chan KH, Beh SL et al. (2006) Time course and cellular localization of SARS-CoV nucleoprotein and RNA in lungs from fatal cases of SARS. PLoS Med 3(2) e:27. Abbreviations CoVcoronavirus DADdiffuse alveolar damage IHCimmunohistochemistry ISHin situ hybridization SARSsevere acute respiratory syndrome SARS-CoVSARS coronavirus ==== Refs References Peiris JSM Lai ST Poon LL Guan Y Lam LY Coronavirus as a possible cause of severe acute respiratory syndrome Lancet 2003 361 1319 1325 12711465 Lee N Hui D Wu A Chan P Cameron P A major outbreak of severe acute respiratory syndrome in Hong Kong N Engl J Med 2003 348 1986 1994 12682352 Poutanen SM Low DE Henry B Finkelstein S Rose D Identification of severe acute respiratory syndrome in Canada N Engl J Med 2003 348 1995 2005 12671061 Hsu LY Lee CC Green JA Ang B Paton NI Severe acute respiratory syndrome (SARS) in Singapore: Clinical features of index patient and initial contacts Emerg Infect Dis 2003 9 713 717 12781012 Peiris JS Chu CM Cheng VC Chan KS Hung IF Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: A prospective study Lancet 2003 361 1767 1772 12781535 Holmes KV SARS coronavirus: A new challenge for prevention and therapy J Clin Invest 2003 111 1605 1609 12782660 Zeng FY Chan CW Chan MN Chen JD Chow KY The complete genome sequence of severe acute respiratory syndrome coronavirus strain HKU-39849 (HK-39) Exp Biol Med 2003 228 866 873 Ksiazek TG Erdman D Goldsmith CS Zaki SR Peret T A novel coronavirus associated with severe acute respiratory syndrome N Engl J Med 2003 348 1953 1966 12690092 Kuiken T Fouchier RA Schutten M Rimmelzwaan GF van Amerongen G Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome Lancet 2003 362 263 270 12892955 McAuliffe J Vogel L Roberts A Fahle G Fischer S Replication of SARS coronavirus administered into the respiratory tract of African Green, rhesus and cynomolgus monkeys Virology 2004 330 8 15 15527829 Guan Y Zheng BJ He YQ Liu XL Zhuang ZX Isolation and characterization of viruses related to the SARS coronavirus from animals in Southern China Science 2003 302 276 278 12958366 Nicholls JM Poon LL Lee KC Ng WF Lai ST Lung pathology of fatal severe acute respiratory syndrome Lancet 2003 361 1773 1778 12781536 Franks TJ Chong PY Chui P Galvin JR Lourens RM Lung pathology of severe acute respiratory syndrome (SARS): A study of 8 autopsy cases from Singapore Hum Pathol 2003 34 743 748 14506633 Hwang DM Chamberlain DW Poutanen SM Low DE Asa SL Pulmonary pathology of severe acute respiratory syndrome in Toronto Mod Pathol 2005 18 1 10 15272286 Shieh WJ Hsiao CH Paddock CD Guarner J Goldsmith CJ Immunohistochemical, in situ hybridization, and ultrastructural localization of SARS-associated coronavirus in lung of a fatal case of severe acute respiratory syndrome in Taiwan Hum Pathol 2005 36 303 309 15791576 Farcas GA Poutanen SM Mazzulli T Willey BM Butany J Fatal severe acute respiratory syndrome is associated with multiorgan involvement by coronavirus J Infect Dis 2005 191 193 197 15609228 Fodor E Crow M Mingay LJ Deng T Brownlee GG A single amino acid mutation in the PA subunit of the influenza virus RNA polymerase inhibits endonucleolytic cleavage of capped RNAs J Virol 2002 76 8989 9001 12186883 Mazzulli T Farcas GA Poutanen SM Willey BM Low DE Severe acute respiratory syndrome-associated coronavirus in lung tissue Emerg Infect Dis 2004 10 20 24 15078592 Nakajima N Asahi-Ozaki Y Nagata N Sato Y Dizon F SARS coronavirus-infected cells in lung detected by new in situ hybridization technique Jpn J Infect Dis 2003 56 139 141 12944688 Li G Chen X Xu A Profile of specific antibodies to the SARS-associated coronavirus N Engl J Med 2003 349 508 509 12890855 To KF Tong JM Chan PKS Au FW Chim SS Tissue and cellular tropism of the coronavirus associated with severe acute respiratory syndrome: An in-situ hybridization study of fatal cases J Pathol 2004 202 157 163 14743497 Chen PC Hsiao CH Re: To KF, Tong JH, Chan PK, et al. Tissue and cellular tropism of the coronavirus associated with severe acute respiratory syndrome: An in-situ hybriodization study of fatal cases J Pathol 2004 203 729 731 15141389 Chow KC Hsiao CH Lin TY Chen CL Chiou SH Detection of severe acute respiratory syndrome-associated coronavirus in pneumocytes of the lung Am J Clin Pathol 2004 121 574 580 15080310 Gu J Gong E Zhang B Zheng J Gao Z Multiple organ infection and the pathogenesis of SARS J Exp Med 2005 202 415 424 16043521 Haagmans BL Kuiken T Martina BE Fouchier RA Rimmelzwaan GF Pegylated interferon-alpha protects type 1 pneumocytes against SARS coronavirus infection in macaques Nat Med 2004 10 290 203 14981511 Wright C Oliver KC Fenwick FI Smith NM Toms GL A monoclonal antibody pool for routine immunohistochemical detection of human respiratory syncytial virus antigens in formalin-fixed, paraffin-embedded tissue J Pathol 1997 182 238 244 9274537 Roberts A Paddock C Vogel L Butler E Zaki S Aged BALB/c mice as a model for increased severity of severe acute respiratory syndrome in elderly humans J Virol 2005 79 5833 5838 15827197 Matthay MA Zimmerman GA Esmon C Bhattacharya J Coller B Future research directions in acute lung injury: Summary of a National Heart, Lung, and Blood Institute working group Am J Respir Crit Care Med 2003 167 1027 1035 12663342 Ware LB Matthay MA The acute respiratory distress syndrome N Engl J Med 2000 342 1334 1349 10793167 Ingbar DH Mechanisms of repair and remodeling following acute lung injury Clin Chest Med 2000 21 589 616 11019729 Pugin J Verghese G Widmer MC Matthay MA The alveolar space is the site of intense inflammatory and profibrotic reactions in the early phase of acute respiratory distress syndrome Crit Care Med 1999 27 304 312 10075054 Nash JR McLaughlin PJ Hoyle C Roberts D Immunolocalization of tumour necrosis factor alpha in lung tissue from patients dying with adult respiratory distress syndrome Histopathology 1991 19 395 402 1757078 Strieter RM Mechanisms of pulmonary fibrosis: Conference summary Chest 2001 120 Suppl 1 S77 S85 Chesnutt AN Matthay MA Tibayan FA Clark JG Early detection of type III procollagen peptide in acute lung injury. Pathogenetic and prognostic significance Am J Respir Crit Care Med 1997 156 840 845 9310002 Borg I Rohde G Loseke S Bittscheidt J Schultze-Werninghaus G Evaluation of a quantitative real-time PCR for the detection of respiratory syncytial virus in pulmonary diseases Eur Respir J 2003 21 944 951 12797486 Wong CK Lam CW Wu AK Ip WK Lee NL Plasma inflammatory cytokines and chemokines in severe acute respiratory syndrome Clin Exp Immunol 2004 136 95 103 15030519	16379499	PMC1324951	CC BY	2021-01-05 11:22:50	no	PLoS Med. 2006 Feb 3; 3(2):e27	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030027	oa_comm
==== Front PLoS MedPLoS MedpbioplosbiolPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0030059SynopsisInfectious DiseasesMicrobiologyVirologyPathologyRespiratory MedicineInfectious DiseasesPathologyRespiratory MedicineLung functionTravel MedicinePathological Clues to How the SARS Virus Kills Synopsis2 2006 3 1 2006 3 2 e59Copyright: © 2006 PLoS Medicine.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Time Course and Cellular Localization of SARS-CoV Nucleoprotein and RNA in Lungs from Fatal Cases of SARS ==== Body Severe acute respiratory syndrome (SARS) first emerged in Guangdong Province, China, in November 2002. At the end of February 2003, an infected doctor from the province inadvertently took the illness to Hong Kong. From there, a woman staying in the same hotel contracted the disease and took it back with her when she returned to Toronto, Canada. SARS, with its ability to spread by close person-to-person contact and with its 10% death rate, was on the move and was threatening to cause a worldwide epidemic. The World Health Organization responded rapidly to this threat by issuing a global alert, and warning against unnecessary travel to affected areas. This and rigorous local containment efforts meant that only 8,098 people became ill, and only 774 people died in this first SARS epidemic. The last case of the epidemic was reported in Taiwan in June 2003, and since then there have only been a few isolated cases. But the coronavirus responsible for SARS (SARS-CoV) could cause another epidemic at any time. Consequently, scientists continue to study SARS intensively, including researchers from Hong Kong and Toronto, who have joined forces to examine the organs of people who died from SARS in these two cities and now report their results. As John Nicholls, the leader of the international team, explains, understanding how the SARS virus kills people should help in the treatment of SARS if it re-emerges. Nicholls and colleagues collected post-mortem material from 32 fatal cases of SARS (the largest such collection to date), and asked three questions about the pathogenesis of SARS. First, was SARS-CoV present in the lungs of these patients throughout their illness? Second, which cells in the lungs contained the virus? Third, did any other tissues contain SARS-CoV? The researchers used three molecular techniques to look for SARS-CoV in their specimens: immunohistochemistry, which detects virus-specific proteins; in situ hybridization, which detects the viral genome; and reverse-transcriptase polymerase chain reaction (RT-PCR), which measures viral load. The researchers found that SARS-CoV was present only in the lungs of patients who died within two weeks of becoming ill (four out of seven patients). In 25 patients who died more than two weeks after the onset of symptoms (generally high temperature and lower respiratory tract symptoms, followed by pneumonia), there was no SARS-CoV in post-mortem lung tissue, although in one case the virus had been present in an open lung biopsy taken five days after disease onset. The researchers found no virus in tissues other than the lung in any of the patients, but in the four patients whose lungs contained SARS-CoV, the virus was found in pneumocytes—cells that line the alveoli, the terminal air spaces where gas exchange occurs—and sometimes in alveolar macrophages, a type of immune cell. No SARS-CoV was found in the cells lining the tubes leading to the alveoli, which explains why patients with SARS only have lower respiratory tract symptoms. SARS coronavirus (green) in lung cells These results indicate that the human immune system can stop SARS-CoV replicating within two weeks of infection. By that time, however, the damage to the lungs in some patients appears to be so great that they die even without continued viral replication. This time course of events indicates that antiviral drugs are likely to be useful only during the early stages of SARS. In addition, the absence of virus outside the lungs suggests that death is the result of SARS-CoV replicating in the lungs alone. Whether SARS-CoV fatally damages lung tissue directly or whether macrophages recruited to the lungs in response to infection with SARS-CoV cause fatal immunopathological changes remains an open question.	0	PMC1324952	CC BY	2021-01-05 11:14:05	no	PLoS Med. 2006 Feb 3; 3(2):e59	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030059	oa_comm
==== Front PLoS MedPLoS MedpbioplosbiolPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0030061SynopsisEvolutionGenetics/Genomics/Gene TherapyImmunologyInfectious DiseasesMicrobiologyEpidemiology/Public HealthRespiratory MedicineEpidemiologyGeneticsInfectious DiseasesTuberculosisVariants in the DC-SIGN–Encoding Gene CD209 and Susceptibility to Tuberculosis 2 2006 3 1 2006 3 2 e61Copyright: © 2006 PLoS Medicine.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Promoter Variation in the DC-SIGN-Encoding Gene CD209 Is Associated with Tuberculosis ==== Body One-third of the world's population is thought to be infected with Mycobacterium tuberculosis, the bacterium that causes tuberculosis. Yet, even though 2–3 million people die from tuberculosis every year, only 5%–10% of individuals infected with the causative agent ever become ill. Many factors affect how well a person's immune system fights off an attack by M. tuberculosis, including HIV infection, age, and malnutrition. It is also becoming clear that, as with other infectious diseases, host genetic factors affect susceptibility to tuberculosis. One of the first genes to be associated with susceptibility to tuberculosis was NRAMP1 (natural resistance-associated macrophage protein 1, now renamed SLC11A1). Others include the genes for the vitamin D receptor and the mannose receptor, and major histocompatibility complex (MHC) class II alleles. Variants of these genes, all of which encode proteins involved in the immune response, may affect how well the immune system deals with M. tuberculosis. Luis Barreiro, Lluis Quintana-Murci, and colleagues now report that two variants in the gene encoding another immune system molecule—DC-SIGN, which stands for dendritic cell–specific ICAM-3-grabbing nonintegrin—are linked to tuberculosis susceptibility. The first rapid line of defense against M. tuberculosis and other pathogens is the innate immune system, where germline-encoded receptors recognize general features on pathogens. DC-SIGN, a C-type lectin that recognizes specific carbohydrate side chains present on the surface of pathogens, is the major receptor of this type for M. tuberculosis on human dendritic cells (immune system cells that process pathogens for presentation to the acquired immune system). Consequently, CD209—the gene encoding DC-SIGN—is a good place to search for genetic variants associated with tuberculosis susceptibility. False-color image of Mycobacterium tuberculosis The researchers looked for CD209 gene variants in 351 individuals with tuberculosis and in 360 healthy controls living in the Cape Town area of South Africa. People living there are known as South African Colored and have a present-day uniform ethnicity. However, they derive genetically from populations of different ethnic backgrounds with different susceptibilities to tuberculosis. Due to the very high local incidence of tuberculosis, everyone living in this region, even healthy controls, is likely to have been exposed to M. tuberculosis. None of the participants were HIV-positive. Barreiro and colleagues discovered two single-nucleotide polymorphisms (SNPs) or variants in CD209 whose frequency differed significantly between healthy controls and people with tuberculosis. Both were in the promoter of CD209 (promoters are noncoding regions that enable genes to be transcribed and proteins to be made). Variants –871G (a guanine 871 nucleotides upstream from the CD209 coding region) and –336A (an adenosine 336 nucleotides upstream from the coding region), either alone or in combination, were associated with decreased risk of developing tuberculosis in the study population. To make sure that these associations reflected differences in disease susceptibility and not ethnic differences between the control and the diseased groups (a problem called population stratification), the researchers showed that 25 other SNPs located genomewide had similar frequencies in both groups. The researchers also report that the “protective” variants of CD209 are more prevalent in Eurasians than in Africans—their presence in the South African Colored population likely reflects the mixing of the two ethnic populations. They suggest that –871G and –336A may have been selected for in Europeans but not in Africans because tuberculosis has been endemic in Europe for several centuries, whereas in Africa it is thought to have arrived much more recently. However, because DC-SIGN is a receptor for many other pathogens, this geographical pattern of variants may reflect exposure to other infectious agents besides M. tuberculosis. Indeed, genetic variation at position –336 of the CD209 promoter is associated with both the susceptibility to HIV and the severity of dengue fever. Whether it is the polymorphisms at positions –871 and –336 that directly affect host susceptibility (as opposed to other genetic variants of CD209 that these polymorphisms are closely linked to) is not entirely clear; nor is it known how the variants affect host susceptibility to infectious disease. One possibility is that they alter how much DC-SIGN is made (and results by others suggest that this is the case for at least one of the variants), which in turn could change the effectiveness of the host immune response. Knowing the functional consequences of the CD209 variants and the mechanisms by which they affect disease susceptibility, suggest the authors, should ultimately help in the development of new knowledge-based treatments for tuberculosis and other infectious diseases.	0	PMC1324953	CC BY	2021-01-05 11:14:06	no	PLoS Med. 2006 Feb 3; 3(2):e61	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030061	oa_comm
==== Front PLoS MedPLoS MedpbioplosbiolPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0030062SynopsisVirologyEpidemiology/Public HealthHealth EconomicsHealth PolicyMedical EthicsMedical LawHealth EconomicsHealth PolicyInfectious DiseasesPublic HealthScreeningThe Cost of Screening Blood Donations for West Nile Virus 2 2006 24 1 2006 3 2 e62Copyright: © 2006 PLoS Medicine.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Cost-effectiveness of Alternative Blood-Screening Strategies for West Nile Virus in the United States ==== Body West Nile virus (WNV) was first isolated in 1937 from a sick woman living in the West Nile District of Uganda. Since then, it has been found in other parts of Africa, Europe, the Middle East, and central Asia, as well as around the Pacific. Then in 1999, it arrived in the Queens borough of New York City. During this first season, it caused 62 cases of encephalitis and seven deaths. WNV has since then spread across the United States, and can now be found in most states. WNV is a flavivirus, a type of RNA virus. Like many other flaviviruses, including the dengue and yellow fever viruses, WNV is a blood-borne virus that is passed to people through mosquito bites; in the case of WNV, the mosquitoes acquire the virus predominantly by biting infected birds. Most people infected with WNV have no symptoms, but 20% of them develop West Nile fever, a flu-like disease. About 1% of infected individuals—usually older people or those with a weak immune system—develop severe neuroinvasive disease, either encephalitis or meningitis, which can cause long-term health problems or death. In 2005, about 100 people in the US died after infection with WNV. Virtually all WNV transmission is through mosquito bites, but a few cases of WNV infection, some of them fatal, have been linked to contaminated blood transfusions. Consequently, in 2004, the US Food and Drug Administration (FDA) mandated that blood donations must be screened for WNV. The FDA did not recommend a specific screening method. In an ideal world, screening would aim to reduce the risk of contracting WNV from a blood transfusion as much as possible, at any cost. In the real world, however, the health benefits of any screening methodology (lives saved and improvements in the quality of life) have to be balanced against the costs of screening. To find out where this balance lies for WNV, Caroline Korves, Sue Goldie, and Megan Murray have estimated the cost-effectiveness of different strategies for screening blood donations for WNV and now report their results. A number of mosquito species can transmit WNV, including Aedes japonicus, depicted here The researchers used a computer-based mathematical model to compare different screening strategies. The baseline strategy was a donor questionnaire—blood donors reporting a recent fever cannot donate blood. The other strategies tested were nucleic acid testing of pools or individual samples of blood for WNV, universal screening versus screening restricted to donations destined for immunocompromised recipients, and seasonal screening versus screening throughout the year (WNV transmission peaks in late summer/early fall). The researchers modeled the cost-effectiveness of these strategies in areas with high levels of WNV transmission over a long season (as occurred in Mississippi in 2002), high transmission over a short season (Nebraska 2002), and low transmission over a short season (Massachusetts 2002). Korves and colleagues found that in low-transmission areas with a short season, screening by questionnaire alone was the most cost-effective strategy—any other strategy was unlikely to prevent any cases of serious illness despite greatly increasing costs. In areas with high transmission, the best approach was seasonal screening by nucleic acid testing of individual donations earmarked for immunocompromised recipients. The researchers also discovered that seasonal screening of all donations provided little additional clinical benefit and was prohibitively expensive, and that screening throughout the year provided no additional benefit in any setting. These results indicate that the currently mandated policy of screening donated blood for WNV may not be the best public health strategy. More restricted screening strategies may be preferable, suggest the investigators, with individual states adopting screening strategies that reflect the intensity and duration of their West Nile epidemics. The researchers also note that their estimates of the cost of WNV blood screening strategies were greater than generally accepted cost-effectiveness thresholds for health interventions. Thus, the resources spent in preventing rare cases of WNV infection arising from blood transfusion might be better used to reduce WNV transmission through controlling mosquito vectors. If such an approach were successful, they suggest, it might obviate the need for screening blood for WNV in many areas.	0	PMC1324954	CC BY	2021-01-05 11:22:50	no	PLoS Med. 2006 Feb 24; 3(2):e62	utf-8	PLoS Med	2,006	10.1371/journal.pmed.0030062	oa_comm
==== Front BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-1291633667810.1186/1471-2458-5-129Research ArticlePublic perception of drinking water from private water supplies: focus group analyses Jones Andria Q [email protected] Catherine E [email protected]é Kathryn [email protected] Shannon E [email protected] Scott A [email protected] David [email protected] Spencer J [email protected] Eric [email protected] Division of Community Health, Faculty of Medicine, The Health Sciences Centre, Memorial University of Newfoundland, St. John's, Newfoundland, A1B 3V6, Canada2 Department of Population Medicine, University of Guelph, 50 Stone Road East, Guelph, Ontario, N1G 2W1, Canada3 Foodborne, Waterborne and Zoonotic Infections Division, Public Health Agency of Canada, 160 Research Lane, Suite 206 Guelph, Ontario, N1G 5B2, Canada4 Department of Agricultural Economics and Business, University of Guelph, 50 Stone Road East, Guelph, Ontario, N1G 2W1, Canada5 City of Hamilton Health Protection Branch, Public Health and Community Services, 1 Hughson Street North, Hamilton, Ontario, L8R 3L5, Canada2005 9 12 2005 5 129 129 11 7 2005 9 12 2005 Copyright © 2005 Jones et al; licensee BioMed Central Ltd.2005Jones 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 Over four million Canadians receive their drinking water from private water supplies, and numerous studies report that these supplies often exceed the minimal acceptable standards for contamination. Canadians in rural areas test their water intermittently, if at all, and treatment of water from private supplies is not common. Understanding the perceptions of drinking water among residents served by private systems will enable public health professionals to better target education and outreach activities, and to address the needs and concerns of residents in their jurisdictions. The purpose of this study was to explore the drinking water perceptions and self-described behaviours and needs of participants served by private water systems in the City of Hamilton, Ontario (Canada). Methods In September 2003, three focus group discussions were conducted; two with men and women aged 36–65 years, and one with men and women 20–35 years of age. Results Overall, participants had positive perceptions of their private water supplies, particularly in the older age group. Concerns included bacterial and chemical contamination from agricultural sources. Testing of water from private supplies was minimal and was done less frequently than recommended by the provincial government. Barriers to water testing included the inconvenience of the testing process, acceptable test results in the past, resident complacency and lack of knowledge. The younger participants greatly emphasized their need for more information on private water supplies. Participants from all groups wanted more information on water testing, and various media for information dissemination were discussed. Conclusion While most participants were confident in the safety of their private water supply, the factual basis for these opinions is uncertain. Improved dissemination of information pertaining to private water supplies in this population is needed. Observed differences in the concerns expressed by users of different water systems and age groups may suggest the need for targeted public education strategies. These focus groups provided significant insight into the public perception of private water supplies and the need for public health outreach activities; however, to obtain a more representative understanding of the perceptions in this population, it is important that a larger scale investigation be performed. ==== Body Background Over four million Canadians receive their drinking water from private water supplies, predominantly from groundwater wells [1]. Numerous studies report that Canadian private water supplies often exceed the minimal acceptable standards for microbial and chemical contamination [1-5], and it is estimated that 45 percent of all waterborne disease outbreaks in Canada involve non-municipal systems, largely in rural or remote areas [1]. Further, several studies report that Canadians in rural areas test their water only intermittently, if at all, and that treatment of water from private supplies is not common [1-3]. Understanding the perceptions of drinking water among residents served by private water systems will enable public health professionals to better target public education and outreach activities, as well as address the needs and concerns of residents in their jurisdictions. Several surveys of drinking water consumption behaviour in North America have been performed [2,6-10], some of which explore the reasons for alternative water use, such as bottled water and water treated with in-home devices [6-9]. However, these studies mainly focus on municipally treated water, and are quantitative/semi-quantitative in nature, providing only a general understanding of residents' perceptions. Focus groups, as with other qualitative methods, are useful in generating rich, detailed, data that cannot be acquired via the use of quantitative surveys, and allow for in-depth exploration of participants' attitudes and responses [11,12]. The purpose of this study was to explore, in-depth, the drinking water perceptions and self-described behaviours and needs of participants served by private water systems in the City of Hamilton, Ontario (Canada). This included participant perceptions of drinking water, their perspectives and behaviours with respect to water testing, the reasons behind any alternative water use and their self-identified needs and desire for information pertaining to private drinking water supplies. Methods The City of Hamilton is a large urban centre surrounded by suburban and rural areas. In September 2003, we performed three focus groups with English-speaking, adult residents (20 years and older) of the City who received their household water from private (i.e. non-municipal) water supplies, including private wells and cisterns. To identify residences with private water supplies for recruitment, we linked residential addresses to digitized maps of the distribution areas served by the City's water treatment utilities within a Geographic Information System (ArcView GIS 3.1, Environmental Systems Research Institute Inc). Residences not falling within municipal water polygons were classified as having a private water source and were included in the sampling frame. These addresses were then cross-referenced with a commercial database of residential telephone numbers for the City of Hamilton. We developed recruitment criteria that were used by a professional marketing firm for telephone screening and enrollment of participants. To gather information from residents of various ages and to avoid problems with mixing distinct age groups, we stratified the focus groups by age. Two focus groups were conducted with men and women between 36 and 65 years of age, and one with men and women between 20 and 35 years of age. Exclusion criteria included being employed in the water industry and having participated in a focus group within the last calendar year. Eight or nine participants were recruited for each focus group to ensure attendance of at least six participants per group [11,12]. An exception to this was the focus group conducted with participants between 20 and 35 years of age, for which only six participants were recruited. Participants were provided a small honorarium for their participation. The Human Subjects Committee at the University of Guelph approved the study and all participants provided written, informed consent. A trained facilitator moderated the focus group discussions, which were audio-taped and professionally transcribed to maximize data capture and facilitate analyses. An assistant also recorded notes on the discussion and group interactions. A pre-tested, structured questioning route was developed according to Krueger and Casey (2000), using a combination of structured questions and pre-planned probes to improve detail and understanding. The focus group discussions were carefully moderated to gather data regarding participant perceptions of water quality, alternative water use, water testing, and their self-identified need for information pertaining to private water supplies. Systematic procedures were used to help to ensure reliability and validity in data collection, including verifying data with participants during and at the end of each focus group, a debriefing session between the moderator and assistant-moderator immediately after each group, and the use of field notes and audio-transcripts. Transcripts were checked for accuracy against the original audio recordings and field notes. Major coding categories were derived from the questioning route and sub-themes were derived from content analysis using methods described in the qualitative methods literature [11-13]. Direct quotations from participants were used for support and illustrative purposes; proper names and profane words were removed from the quotations reported herein, and portions of quotations that needed clarifying context were supplemented with additional text that was placed within square brackets. Results Participants Five people attended the first of the two focus groups conducted with people between 36 and 65 years of age, and six attended the second. Four people attended the focus group conducted with participants between 20 and 35 years of age. Each group had a roughly equal proportion of men and women and reflected a variety of income levels and educational backgrounds. All participants were Caucasian. Perceptions The facilitator asked very general questions to stimulate discussion regarding the participants' perceptions of their drinking water. Without direct influence from the investigators, participants introduced into discussion several broad themes, each of which is discussed in turn below: Sensory quality of water Participants' perceptions of the sensory quality of drinking water from their private supplies were overwhelmingly positive, particularly within the 36 to 65 year-old age group. The majority of participants reported their water to be "excellent" in taste. Common words used by participants to describe their water included "great tasting", "fresh", "very cold", "no chlorine" and "no smell". Participants reported only two distinct troubles with the sensory quality of their water. Many commented on the hardness of the water and disliked the effect it had on appliances and plumbing. A few participants also said they disliked the sulphur smell and taste of their water ("when we moved to our house it was sulphur. It was quite sulphur. I couldn't get my taste buds around it..."). However, participants generally referred to the hardness and sulphur content of their water as "inconveniences", rather than concerns. Water safety Most participants, particularly within the 36 to 65 year-old age group, were very confident in the safety of their water; all of the participants in one group even reported having "no concerns at all". Participants gave two main reasons for their belief in the safety of their water. One related to the independence of private water systems; not having to rely on others for the provision of their drinking water. Two participants illustrated this theme well when they explained why they perceived their water as safe: " [In the city], there's a lot of things you depend on, whereas... out in the country at least you have that independence. I know where my water is coming from. I know what's in it. I don't have to worry about someone monkeying with it." "... partly because if you've got a well on your property, it's your responsibility to take care of it, so you know who's looking after it. That makes it pretty darn safe." The following participant statement was also met with high agreement: "...there's a sense of that independence too and not having to depend on someone else to chlorinate the water, to make sure the system works." The well-publicized E. coli 0157:H7 outbreak, associated with the inadequate management of municipal well water in Walkerton, Ontario in 2000 [14], was often raised in support of this point. The natural filtration associated with ground water supplies was the other main reason for the participants' confidence in the safety of well water. One participant expressed this as follows: "...by the time [contamination] seeps into your well, everything has been filtered out. The water, the earth is a natural filter, and you just get cleaner water...". Despite confidence in the safety of their water supplies, several participants discussed the potential for contamination of their water with pesticides, fertilizers and fecal run-off from nearby farms. One participant said: "... one of the things that we always have our heads up with, is because we're country, and there's a huge pig farmer and a dairy, a cow farmer beside us... They have open pits and every spring they spread over the fields before they plant them, they spread and that actually at the end of the day becomes groundwater." She went on to add however: " [But] like really, it's well filtered. We have a 150 foot well too." Two participants mentioned concerns that their children would likely be the first family members affected by contamination of their water source. For instance, one commented: "But I do worry more about [contamination], probably as any parent would, like, because I have a small child, so I think when your children are small, well, I don't worry as much for myself; I probably should, but I think you get more concerned...". While participants in the 36–65 year-old age group agreed that contamination from agriculture was a possible health risk, several also said that it was inherent with country living and "just something to think about". Overall, there was a tremendous sense of pride and contentment in the quality and safety of water from private water sources in this age group. Confidence in the safety of well water was not as pronounced among participants in the 20–35 year-old group. One participant explained having concerns about the safety of her well water because of a crack in her well: "I know there's a leak in there. Whatever can get out can come in. You know, and ...I live on a cow farm, so I do not touch my [well] water." Another participant, although greatly appreciative of his well water, said: "I guess you're always kind of worried about something seeping into it, right?" When asked to explain, he replied: "Just like, stuff like Walkerton, like you hear E. coli and stuff like that and you never really know what's being dumped in the ground... like if no one's watching... dumping chemicals or something because that's going to go directly into your water." These two participants also had views regarding self-sufficiency that contrasted with those of the other participants. For instance, one participant spoke of municipal water: "At least the city has someone designated to supposedly be watching it, right? So there should be some kind of, um, supervising element to it, whereas the well water, you just, you're relying that it's good...". The other participant with concerns then agreed: "Nobody's checking it out", to which the first replied: "You're never really 100% sure." Despite this, all participants, with exception of the one group member with the crack in her well casing, considered their water to be "just as safe" or "safer" than municipal water. They reported a number of reasons for this perception, including: the "natural filtration of wells", the high probability of any contaminants being diluted to safe levels, their water not having to pass through "dirty" or failing municipal distribution systems, and the lower likelihood of their water being a target for terrorist attack. Effect of rural development on the water aquifer The most extensive concern was that nearby development and construction would have a negative effect on the water aquifers supplying the participants' wells. Two illustrative participant statements included: "The only concern we had was the new development they were putting up in behind us, whether it's going to screw up things and alter the water quality. Every time they do a major development over there you sort of check..." "I'm fortunate, I really don't have any development around me, I just have one neighbour and uh, no sign of anything going in, so I know that I'm safe for a while anyways. There's no developments going in, but you never know." Concerns pertaining to specific types of private water systems Several participants expressed concerns about the costs of replacing water pipes, especially given that these costs are incurred by themselves. One participant said that he worried about the expense and inconvenience of alternate water sources should the water from his well no longer be available. Another participant whose house was supplied by a water cistern was annoyed with insufficient quantities of water. On having a cistern she said: "... it's just a pain in the neck if you run out of water. I mean, I could be in the middle of a shower and all of a sudden, I have no water." Need for perspective Several participants commented that there was a need to maintain perspective when considering concerns with drinking water. One participant illustrated this theme well: "If you were living on a well and you're saying, my God, I wonder what that farmer next door put into my water. You know. Is it going to be safe?... You'd drive yourself crazy. So you actually have to build up some sort of confidence in whatever system you have just to avoid driving yourself insane." Hence, while some participants recognized the potential risks associated with private water supplies, most reported that they choose not to "dwell" on these concerns. Bottled water With the exception of the participant who relied solely on bottled water because of the crack in her well casing, participants had only negative comments regarding it. Two themes were evident: poor taste and skepticism as to the water source. Regarding taste, one participant said, "it tastes like plastic to me" and several participants agreed with the following participant statement: "I can't drink that bottled stuff, it's just blah... It quenches your thirst and that's about it, but other than that it's got no taste to it and if I didn't have to drink it I wouldn't." Many participants were also suspicious that commercial bottlers made fraudulent claims regarding their water sources. The participants' distrust of bottled water was well captured in one participant's comment: "... these people that buy bottled water all the time, I'm asking, do you know where that came from? [They say] 'It's all right. It's in the bottle.' [I say] what the [heck] has that got to do with it – it's got a cap on it? Don't you remember the scandal of [a prominent water company], when they found it wasn't coming from where it was supposed to be, ten years ago. They were getting it... not from the artesian well that they said they were getting it from, and they proved that [the company] had been scamming [people] for five years. You don't know what's in [bottled] water... You don't know where it's coming from..." Despite their negative opinions, participants in the 36–65 year-old age group reported occasional consumption of bottled water, with convenience being the most extensive reason for its use. The participants reported using bottled water merely because the bottle was portable, convenient and generally well-accepted in the workplace and schools. One participant said: "It's really just truly the convenience of having it in the fridge with the lid on it that I can take in the car or to work". Further, many participants reported re-filling bottles with the water from their private supply. Many also said that they only purchase bottled water when they were out and chose it as a substitute for other beverages. With the exception of one participant who said she preferred bottled distilled water for making coffee, all participants in the 36–65 year-old age group said convenience was the only reason for purchasing bottled water. In the 20–35 year-old age group however, participants reported two reasons for using bottled water. As with the older participants, there was a preference over other types of beverages while outside of the home. However, concerns were also expressed about the quality of their well water. The participant with a crack in her well reported: "When we first moved in it was brown, murky... [we] couldn't see through it and we were told that if we boiled it, it would be fine. But when you have a family, you know, it was just a lot easier to... get a cooler and just go with the spring water." Another participant reported that his family first purchased a bottled water cooler when: "they were doing quarry and mining and stuff around... And from what I understand, [they] cracked the bedrock and sulphur had leaked into the water, and then I think that's when it started smelling bad". This household had later installed a treatment device to resolve the sulphur problem, and then discontinued their in-home bottled water use. In-home water treatment devices Most participants used devices to treat the water from their private water supplies; the most common devices were water softeners. A few other systems were reported, including a water distiller and ultra-violet light and reverse osmosis devices. Of the two participants who did not use treatment devices, one was the participant who consumed only bottled water from dispensing coolers in her home. There were two major reasons for the use of treatment devices: to reduce the hardness and sulphur content of well water and to increase the safety of water from cisterns. The majority of participants reported using water softeners in order to decrease the negative effects of hard water on plumbing and appliances. Many participants shared the set-up described by one participant: "We use a water softener just because of the hardness, but then we have a three pipe system which is treated hot and cold, and an untreated cold line for drinking." Some respondents reported using treatment devices because their water had a sulphur smell and taste, which they found disagreeable. Participants who used water cisterns however, reported using ultraviolet light or reverse-osmosis devices in order to clean and decrease the contamination of their water. One cistern-owner explained their use of a treatment system: "We just thought [it was] probably safer... our eaves troughs, I mean, they're regularly cleaned, but we just thought it's probably safer for cooking and [consumption]". Another explained: "You know, just to feel more confident." Several well-owners agreed that treatment systems would be necessary for drinking water from cisterns ("Safety wise, yeah, you'd have to"). However, well owners reported that, to the contrary, they used treatment devices to decrease the hardness or sulphur content of the water, and not out of concerns for safety. Private water testing behaviours Participants were asked how frequently they tested the water from their private water source. Some participants tested it once per year and some tested every two to three months. However, many participants reported that they had never tested their water, or only tested it once every few years. A few participants were unsure as to whether their water was tested, but speculated that another member of the household might take this responsibility. When asked what tests were undertaken on their water, only few participants gave a definitive answer ("E. coli"); several were unsure because another member of the household did the testing ("All I know is that [my parents] bottle up a sample and they send it to whatever facility that tests it and we get a report back."). Even among those participants who did test their water regularly, many reported not knowing what the water was specifically tested for. One participant said: "I send it for testing, and they [at the lab] figure out what they're testing it for." Overall, the extent of testing appeared to be limited to coliforms, and a general feeling of uncertainty existed across participants with respect to water test parameters. Respondents who tested their water reported doing so for various reasons. Several said they regularly tested it because of failed tests in the past: "... when we failed the test, that's when we upped the ante a bit in terms of getting it tested." Many participants also reported testing their water because they had learned of local water problems, either from neighbours or, in some cases, because of an informative flyer being distributed to their homes. Several participants explained: "...there was a lot of talk about some of the houses right down in the village having problems, so it was close enough, hearing about local problems, we figured we'll test ours and see what it is." "Until that point we'd never tested it, since I was a kid, but once we got that flyer, we test every year now." Respondents gave numerous explanations for not testing their water regularly. Many said the inconvenience of testing prevented them from testing more often. There were many complaints of having to make several trips to the city in order to pick-up and drop-off sample bottles. The following participant statement was met with high agreement: "It's as I say, two or three trips to some place which you might not necessarily go by. You've got to make a special trip. And when we're talking country people having to get into the city, some of us avoid coming into the city." Participants also reported not testing their water regularly because their test results in the past had always been negative. The following participant statements were indicative: "You get a couple of good readings over and over and over again that you've got fantastic water, and after a while I say 'oh, well, it's fantastic water'. Why am I going to waste my time with all this?" "... we only test our water once a year. Maybe we should do it more, but because it always comes back zero-zero, we just don't do it." Many participants also reported "complacency" as a reason for not testing their water more often. For instance, several participants who reported never testing their water commented: "Oh, I'm not all that diligent in checking my water. As long as it tastes good, that's fine." " [I'm] lazy. I'm not worrying about it enough to do it. When something comes up and people start getting concerned, everyone's testing and complaining, well, [then] we'll get it tested." Some participants said that infrequent testing was also due to ignorance and not knowing that water testing was available or necessary. One participant explained: "But see, and I mean, I've lived out there for... years and I didn't realize that there were places that you could take your water to get tested, or whatnot." Another participant suspected that ignorance was especially the case for "ex-urbanites" who had recently moved to the country. He explained: "...there are a lot of people who move from the big city out into the country, and don't realize they should be testing their water... I know a couple of people who've moved in [to rural areas] recently and they've never heard of testing water, because they come from a [municipal] system: you turn the tap on, it's there." A few participants also reported not testing their water because of fear of the "government's" response to a positive result: "Once you put your water in to be tested, then they expect you to fix it if it's broken". Another participant agreed that this would deter people from testing their water: " [Especially] if it was a rental property... it's almost like you're inviting more expense to yourself if they find something." Methods to encourage private water testing The participants said that public education would help to increase water testing among people with private water supplies. In some instances, it was seen as serving to provide new information, in others it would act as valuable reminders that would "at least get us thinking about it more". An idea that garnered a lot of support in all of the groups was an educational campaign that made regular water testing habitual. Several participants likened this to the popularized North American reminder to change batteries in smoke alarms when clocks are changed for Daylight Savings time: "It's like, you always hear about it and you think, okay, change your clock, you think about your [smoke alarm] battery. It becomes, it's like advertising or something... it becomes part of your consciousness over time." Many participants thought that one-time reminders would not generate sustainable change; instead, water testing needed to become an engrained action that was part of rural-life culture. Another extensive theme to encourage water testing revolved around making the testing process more convenient. Many participants said that they would appreciate having a pick-up service, where someone, preferably students doing community work, would come to their house to collect water samples for testing. One participant illustrated his reasoning: "I would like a student to come by and test every household. [Someone] with a bottle for water, saying... 'I'm with the City of Hamilton. We're running this program. We're going to test your water and we'll send you back your results.' Because you're going to get older people. You're not going to teach an old dog new tricks. You can talk until you're blue in the face. Until someone does it for them, they won't do it." Another participant said that she thought it would be very effective because: "... we are lazy, everybody is. We've got our lives, we're all busy, you know. You go home, you just want to take off your shoes and get the converter, lay on the couch, you know. But [with such a service] you're more apt to do it, I think." Other participants agreed that the process of water testing should be made more convenient, possibly by increasing the number of pick-up and drop-off locations, particularly in the more rural areas: "I mean, in a little place like [a village within the City], if they had a drop off place where you picked up your mail, and say, they picked a day and they said, okay, the fifth of the month is water testing day. You drop off your bottle for the fifth of the month, someone from the lab comes by and picks up the carton of water from local people who want it taken, you might do it more often." Some participants also said that they would like a home test kit, which would make the process most convenient. Finally, many participants thought that having access to test results from other wells near to their home would dramatically encourage testing of water from private supplies. There was generous support for an accessible system that would share private water test results, ideally through a mapping system: "Like, the idea of a map saying the water quality has been tested here and here and here. Here are the areas with some problems, these are areas that are doing well." The participants emphasized the need for confidentiality with such a system, but also said that it would be an effective way to encourage regular water testing. As noted previously, many participants had tested their water in response to learning of water problems in neighbouring areas. Public education Participants were asked whether they wanted to receive information regarding water from private water supplies. Although most participants concurred with this, two members from the 36 to 65 year-old groups said that they did not need nor want any further information, because they were confident in the safety of their wells. Another participant said, "I wish I knew more", but had concerns about being overloaded with information. As noted above, many participants wanted a system that would share water test results from surrounding areas. The request for this information was repeatedly illustrated in all of the groups. Many participants also indicated a desire for more information pertaining to water testing. Specifically, they wanted to know what test parameters were available in the free test supplied by the City, what other tests were available and should be performed, the costs of these tests and the laboratory that performed them. Several participants said that they would like this information, but would like it specifically customized; for instance what their water should be tested for given their particular geographic location: "I would like to know... what else I could have my water tested for, and maybe should have my water tested for, depending on the area I live in." Participants also said that they would like water test results explained in clear language, specifically indicating what effects different components of their water might have: "Like, learning the different components of water and what's good, what's bad, I mean, so you have a lot of sulphur in your water. Is that going to harm you?" Several participants also wanted information pertaining to water treatment options, based on any problematic water test results they might receive. Finally, a few participants said that they would appreciate general waterborne disease information, including what pathogens may be in their water and what illnesses could result. In the 36–65 year-old age group, there was a recurring theme pertaining to education. Several participants spoke of the need for knowledge and respect for private well water systems, including their planning, construction and maintenance. They said that many negative health events associated with well water were because this mind-set was lacking. Examples included contamination associated with running stale wells or city layouts where shallow wells were located downhill from agricultural farms. One participant spoke of several areas in Southern Ontario where: "all these farms sit above and the town sits down below and you start having shallow wells. [Contamination is] what's going to happen with ground run-off. You can't get away with it." The overall impression these participants gave was the need for well owner education with respect to proper well planning, construction and maintenance. There were clear differences between the two age groups in the intensity of participants' self-identified need for information. Participants in the 36–65 year-old age group were largely content with their current level of general knowledge, but wanted more information in specific areas, like testing and test results. In the 20–35 year-old age group however, participants stated that they were deeply lacking in general knowledge, and greatly emphasized their need for information. One participant explained: "I mean, I've lived out there for... years and I didn't realize that there were places that you could take your water to get tested, or whatnot. So the awareness is definitely not there. ...If I don't want to get off my lazy [behind] and go and do [the testing], that's my choice. [But] right now, I didn't know I had a choice." Another participant said he wanted: "just generally more [information], you know", but had difficulty explaining what he specifically wanted: "See, I don't know what I need to know, so I don't know what I want to know..." Among all groups, a wide range of media was suggested for disseminating information. Common responses included local newspapers, flyers/brochures mailed to the home, a city webpage, and radio messages played on local stations during commuting hours. A few participants also suggested short, informative television "commercials". Participants also offered advice with respect to designing effective educational programs. Many said that the information should not be presented in such a way as to cause people alarm or panic: "You see, you've got to be so careful that you don't start scare-mongering." While the majority of participants supported this, one stated that some element of fear was required to motivate people to read or use the information: "...that element of fear had to happen before we even worried about our water and I think that's what's needed to kick people in the rear end to test their wells". Participants also said they wanted something direct, and to the point: " [If I take] a quick glance and it's to the point, 'people's well water, bang, see this website', that would grab my eye. It would only take me five seconds and I'd say well, that might be interesting. I better check that later." Many participants thought the flyers should include only the basic facts and should direct people to another source (for example, a website or other publication) for more information: "Don't put a lot of detail on it, just put where you could find the information that you should be looking for". With respect to information disseminated via the internet, some participants said that they wanted a direct link to information so they did not have to navigate through a larger web page. Discussion Most participants in this study expressed a strong appreciation and a high degree of confidence in the safety of their private water supplies. Many had trust in the safety of their water despite not having had it tested or using a treatment device. Several studies in Ontario and other parts of Canada however, have shown chemical and microbial contamination of private water supplies in excess of government standards for safe drinking water [1,3,4,15,16], and similar results have been observed in other developed countries [17-19]. Some participants were concerned that nearby agricultural operations and building development could contaminate their wells or otherwise have a negative effect on their aquifers. However, these participants seemed to be concerned only if these activities were occurring immediately nearby their homes. With the exception of one participant, most did not suggest that contamination of aquifers might occur some distance from the wells themselves. Further, given that some waterborne illnesses are self-limiting and/or require chronic exposure, waterborne hazards may be present without the owner's recognition, thereby posing a risk to residents of the household. Also, because residents may eventually develop immunity to a pathogen(s) present in their water, a hazard may go unrecognized and pose risk to visitors to the household. Thus, the inherent confidence of the focus group participants in the safety of their water supply may or may not be warranted; many of the participants did not have adequate information or test results on which to soundly base their opinions. If these perceptions are representative of the general population, it could indicate the need for increased dissemination of information, including possible water contaminants and their effects, the importance of regular testing of private water supplies, the fact that development or agriculture does not necessarily need to be very close in proximity to affect well water quality, and that residents relying on private water wells in Ontario have a legal responsibility for the condition of their wells. Participant concerns included bacterial and chemical contamination from agricultural sources, and in the 20 to 35 year-old group, illegal dumping of pollutants and the lack of an official monitoring system for drinking water from private water supplies. Identification of drinking water concerns in the larger population might help policy makers ensure that their activities and attention include the areas of their constituents' concerns. Bottled water use among participants in the 36 to 65 year-old age groups was not common, and participants reported using it solely for convenience or as a substitute for other beverages while not at home. Two participants in the 20 to 35 year-old age group reported choosing bottled water because of concerns regarding the safety or sensory quality of their well water. A survey of municipal water consumers in the U.S. reports that people between the ages of 18 and 34 years are more likely to believe in the safety and health benefits of bottled water compared to those over the age of 35 [9]. Generational differences in the degree of concern and interest regarding water may explain the differences between the two age groups and such differences could have implications for the targeting of public education strategies. Perceptions also differed between the two types of private water systems discussed, namely wells and cisterns. Participants who received water from cisterns reported using treatment devices because of contamination concerns. The majority of well owners however, used only water softeners, while leaving the cold water line in the kitchen untreated. This, in conjunction with the fact that water softeners alone do not remove most chemical or microbial contaminants, supports the participants' assertions that they used treatment devices solely to address the hardness and/or sulphur content of their well water, and not out of concerns for safety. Similarly, Levallois and colleagues found that only nine percent of 222 private water households in Quebec used treatment devices, with 11 percent of them doing so because of health concerns and the remainder for organoleptic reasons, like taste, odour and colour of their water [15]. The level and reasons for use of treatment devices among private water households in this study would appear to depend on the type of water system present, which could have implications for the best type of information to deliver to these different groups. Cistern owners for instance, might best benefit from information pertaining to the choice, use and maintenance of treatment devices, whereas well owners might benefit more from messages regarding the importance of water testing and what to do if a positive result is returned. The current recommendation regarding routine testing of private water supplies in Ontario is to test water for indicator bacteria at least three times yearly, in the spring, summer and fall [20]. Testing of water for E. coli and total coliforms is free in the City of Hamilton, however, other test parameters are available at a cost to the owners of private water supplies. In this study, few participants regularly tested their well water, with most testing less frequently than provincial recommendations. This is comparable with results from other Canadian studies, which report that rural residents test their private water supplies intermittently, if at all [1,3]. It is important, however, to understand why residents choose not to test their water. Such data could be used to improve current public health education programs, encourage water testing and potentially help reduce the risk of disease from private water supplies. For instance, many participants in this study did not regularly test their water because their results had always been negative in the past. This may indicate the need to disseminate information regarding intermittent contamination and the changing quality of private water supplies. Many participants also said that they did not test their water because it looked and tasted normal; hence, they may benefit from being informed that many contaminants are odourless, colourless and tasteless [21]. Others said they were not aware of the need to test their water, which suggests the need for increased awareness of private water testing in general. One of the factors most contributing to the participants' decision not to test their water was the inconvenience of the testing process. If this was representative of the overall population, it would indicate the need to emphasize the importance of testing, and/or increase the convenience of the process, perhaps via a water sample pick-up service, or increased and more convenient sample-bottle pick-up and drop-off locations. Inquiry as to residents' thoughts on how to encourage testing is also of great utility. By directly asking what would work to encourage private water owners to test their water, we are better equipped to design effective public health outreach activities. In this study, participants valued the creation of an educational program that would make water testing habitual among private water system owners. Given current Canadian recommendations regarding test frequency, a program that included public reminders with the changing of the seasons may be appropriate. Based on our results, a reminder campaign, released through a variety of public media outlets, designed to associate water testing with well-recognized dates may be appropriate. In Canada, such dates could be the first days of spring, summer and fall, or holidays like Easter (March or April), Father's Day (June) and Thanksgiving Day (October). Increasing the convenience of the process also appears to be warranted. Ideally, suitable funding would be made available to the appropriate health departments to implement such changes. Perhaps community groups or high-school community involvement curricula might organize water sample pick-up and drop-off programs. Similarly, the local health department could increase the number of service locations, particularly in the more rural areas of the City. Another idea endorsed by participants was the creation of a results-sharing program that could be used by private water supply owners to learn the test results of private water supplies within their locality. While the need for confidentiality would obviously have to be addressed, participants felt that such a system would work in promoting water testing. This is also supported by virtue of the fact that learning of local water problems was the sole motivating factor behind some participants testing their water in the past. Perhaps creation of a system that provided basic test results of very general areas (for instance, "West Hamilton") would be useful. Further investigation of the testing behaviours and opinions of the general population, as well as the logistics of such a system, is required however, before such recommendations can be made. This study highlighted areas of drinking water information that require increased dissemination in this population. Participants wanted more information on water testing, and participants in the 20–35 year-old age group said they were broadly lacking in knowledge and wanted more information regarding private water supplies in general. This suggests that the targeting of the younger generations may be warranted. Some participants also suggested a need to target residents new to rural areas, who may not otherwise be informed with respect to private water supplies; a point supported by other researchers [19,21]. In addition to their role in ensuring water quality as part of current regulations regarding home buying/selling, perhaps real estate agencies might collaborate with public health departments. For instance, they could disseminate informative flyers/other media regarding the importance of regular, on-going water testing to those purchasing homes with private water systems. Similarly, this information could be provided when property ownership changes and the deed to the home is transferred. Some of the information that participants wanted is contained within a water well Best Management Practices Guide [20], which is made available by the health department in multiple community centers within the City of Hamilton. Unfortunately, participants in this study were unaware of its existence. Further, while informative, well-organized and illustrated, the guide is approximately 90-pages long and may therefore be overwhelming for some residents. Participants in this study said they wanted information to be clear, to the point, and distributed using other forms of media, like newspapers, radio and flyers distributed to their homes. If the results of this study are representative of the larger population, it would indicate the need for changes to water safety communication within the City of Hamilton. For instance, an effective program might include flyers or short radio announcements with general information that also directs residents to other sources of information, like a webpage or the Best Management Practices guide. This would also be a more effective way of targeting residents who are not actively seeking such information. These focus groups allowed us to explore, in-depth, participants' perceptions of water from their private water supplies, including their self-identified need and desire for more information. We were able to gain this understanding with little interviewer influence, and were therefore privy to participant issues that we had not previously considered. While labour- and cost-intensive, the focus groups provided a detailed level of understanding of perceptions that we could not have been gained through quantitative survey alone. The ultimate goal of focus group research however, is to gain understanding, not to generalize. Thus, recommendations cannot be made solely on the results of this study, as the sample likely is not representative of the general population. This study has however, provided insight into the residents' perceptions of water from private supplies, including the vocabulary used by participants' to discuss the issues that were most important to them. Additionally, this information will be used to develop a survey instrument to investigate this population's perceptions on a larger scale. Our survey questions and categories are therefore more likely to reflect accurately the residents' empirical world, which will contribute to the validity of our survey instrument. Despite their benefits, the focus groups are not without their limitations. Ideally, additional focus groups would have been conducted in order to ensure saturation of the data. Unfortunately, this was not possible due to budgetary limitations. Further, we had difficulty meeting target recruitment numbers, particularly in the younger age group, which may have contributed to selection bias. There are likely two main reasons for this. First, residents on private water systems are a minority in the population of the City of Hamilton. Second, our mapping system to identify residences with private water supplies had a predictive value of just 61 percent (Jones et al., unpublished), and was more likely to misclassify private system residences as having municipal water, particularly when they were near municipal water system boundaries, than vice versa. Hence, we may have inadvertently narrowed our selection pool, and increased the relative proportion of "very rural" residences. Given that only three focus groups could be conducted, selection criteria were limited and did not include ethnicity. The fact that only Caucasians attended the groups, as well as the requirement for participants to be English-speaking may have narrowed the transferability of the results. According to Statistics Canada 2001 census data of the City of Hamilton, however, approximately 97% of the population is able to converse in English; hence selection bias in this latter regard is likely to have been minimal. Future investigations would be strengthened via the inclusion of participants from a diversity of ethnic backgrounds. It is also possible that some participants responded to questions in a socially desirable manner, modifying their true responses as a result of other people in the room. However, we made every effort to prevent this type of bias by clarifying before the groups started that there were no "right or wrong" answers, and participants were encouraged to agree or disagree with one another as appropriate. As group knowledge of socio-economic status is also known to affect group interactions, we were careful to avoid any discussion of this prior to and during, the focus groups, as recommended [11,12]. Overall, the focus groups were non-confrontational and friendly in nature suggesting that participant influence in this manner was minimal. Conclusion We performed an in-depth investigation of participants' perceptions of drinking water from private water supplies in the City of Hamilton, Ontario (Canada). Most participants, particularly those over the age of 35, were confident in the safety and quality of their water. There is some question however, as to whether these opinions are well informed. There was a high degree of uncertainty in discussing water testing, and most participants did not test their private water supplies with the frequency currently recommended. Further, where testing was done, it was for E. coli and total coliforms only, primarily via the free testing service. This may pose a public health risk, as contamination of private water supplies in various parts of Canada is well documented [1,3,4,15,16]. The results of this study show that participants want and need more information pertaining to private drinking water supplies, particularly with regard to water testing. They also suggest the need for changes to current information dissemination efforts in the population. Further, observed differences in the perceptions and needs of people on different private water systems and between different age groups may suggest the need for targeted public health strategies. If the results of this study are applicable to the general population, action on the part of public health officials, and potentially various levels of government, is necessary. A valid understanding of residents' perceptions, needs and concerns with respect to private drinking water supplies is integral to the development of effective public health strategic planning, public education programs and drinking water policy. These focus groups provided significant insight in this regard, but it is important that larger scale investigations of the perceptions of drinking water from private water supplies be performed. Competing interests The author(s) declare that they have no competing interests. Authors' contributions SJH conceived the project; AQJ designed the study, administered/moderated the focus groups, collected and analyzed the data, and drafted the manuscript; CED, KD, SEM, SAMc, DWT, SJH and EM provided critical feedback on the study design, the analyses and interpretation of results, as well as editorial comments on the manuscript. SEM also assisted with administration/moderation of the focus groups. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements The authors thank the residents of the City of Hamilton for their participation. The University of Guelph (Ontario Veterinary College and the Department of Agricultural Economics and Business), the Public Health Agency of Canada, and the City of Hamilton Health Protection Branch provided financial and in-kind support. ==== Refs Corkal D Schutzman WC Hilliard C Rural water safety from the source to the on-farm tap J Toxicol Environ Health A 2004 67 1619 1642 15371205 10.1080/15287390490491918 Levallois P Guevin N Gingras S Levesque B Weber JP Letarte R New patterns of drinking water consumption: results of a pilot study Sci Total Environ 1998 209 233 241 9514043 10.1016/S0048-9697(97)00320-3 Thompson TS General chemical water quality of private groundwater supplies in Saskatchewan, Canada Bull Environ Contam Toxicol 2003 70 447 454 12592517 10.1007/s00128-003-0007-3 Strauss B King W Ley A Hoey JR A prospective study of rural drinking water quality and acute gastrointestinal illness BMC Public Health 2001 1 8 11580869 10.1186/1471-2458-1-8 Raina PS Pollari FL Teare GF Goss MJ Barry DA Wilson JB The relationship between E. coli indicator bacteria in well-water and gastrointestinal illnesses in rural families Can J Public Health 1999 90 172 175 10401167 Auslander BA Langlois PH Toronto tap water: perception of its quality and use of alternatives Can J Public Health 1993 84 99 102 8334617 Lee S Levy D Hightower A Imhoff B EIP FoodNet Working Group Drinking water exposures and perceptions among 1998–1999 FoodNet survey respondents 2002 Atlanta, Georgia: U.S. Centers for Disease Control and Prevention Levallois P Grondin J Gingras S Evaluation of consumer attitudes on taste and tap water alternatives in Quebec Water Sci Tech 1999 40 135 139 10.1016/S0273-1223(99)00549-1 AWWARF (American Water Works Association Research Foundation) Consumer Attitude Survey on Water Quality Issues 1993 Denver, Colorado: American Water Works Association Research Foundation Jones A Dewey C Dore K Majowicz S McEwen S Waltner-Toews D Drinking water consumption patterns in a Canadian community J Water Health Krueger RA Casey MA Focus Groups: a practical guide for applied research 2000 3 Thousand Oaks, CA: Sage Publications Inc Morgan DL Krueger RA The Focus Group Kit 1997 1 Thousand Oaks, CA: Sage Publications Inc Taylor-Powell E Renner M Analyzing Qualitative Data 2003 University of Wisconsin, Cooperative Extension Bruce-Grey-Owen Sound Health Unit The Investigative Report of the Walkerton Outbreak of Waterborne Gastroenteritis 2000 Owen Sound, Ontario: Canada Levallois P Theriault M Rouffignat J Tessier S Landry R Ayotte P Girard M Gingras S Gauvin D Chiasson C Groundwater contamination by nitrates associated with intensive potato culture in Quebec Sci Total Environ 1998 217 91 101 9695174 10.1016/S0048-9697(98)00191-0 Frank R Chapman N Johnson R Survey of farm wells for nutrients and minerals, Ontario, Canada, 1986 and 1987 Bull Environ Contam Toxicol 1991 47 146 151 1932855 10.1007/BF01689466 Rutter M Nichols GL Swan A De Louvois J A survey of the microbiological quality of private water supplies in England Epidemiol Infect 2000 124 417 425 10982065 10.1017/S0950268899003970 Borchardt MA Bertz PD Spencer SK Battigelli DA Incidence of enteric viruses in groundwater from household wells in Wisconsin Appl Environ Microbiol 2003 69 1172 1180 12571044 10.1128/AEM.69.2.1172-1180.2003 Said B Wright F Nichols GL Reacher M Rutter M Outbreaks of infectious disease associated with private drinking water supplies in England and Wales 1970–2000 Epidemiol Infect 2003 130 469 479 12825731 Agriculture and Agri-Food Canada Best Management Practices: Water Wells 2003 Ontario: Agriculture and Agri-Food Canada Simpson H Promoting the management and protection of private water wells J Toxicol Environ Health A 2004 67 1679 1704 15371209 10.1080/15287390490492296	16336678	PMC1325020	CC BY	2021-01-04 16:28:58	no	BMC Public Health. 2005 Dec 9; 5:129	utf-8	BMC Public Health	2,005	10.1186/1471-2458-5-129	oa_comm
==== Front Plant MethodsPlant Methods1746-4811BioMed Central London 1746-4811-1-121635617110.1186/1746-4811-1-12CommentaryFast-track applications: The potential for direct delivery of proteins and nucleic acids to plant cells for the discovery of gene function Roberts Michael R [email protected] Department of Biological Sciences, Lancaster Environment Centre, Lancaster University, Bailrigg, Lancaster, LA1 4YQ, UK2005 15 12 2005 1 12 12 21 11 2005 15 12 2005 Copyright © 2005 Roberts; licensee BioMed Central Ltd.2005Roberts; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. In animal systems, several methods exist for the direct delivery of nucleic acids and proteins into cells for functional analysis. Until recently, these methods have not been applied to plant systems. Now, however, several preliminary reports suggest that both nucleic acids and proteins can also be delivered into plant cells by very simple, direct application. This promises to open the way for high-throughput screening for gene function in a range of plant species. ==== Body Introduction The development of assays that permit high-throughput screening for biological function is an essential goal if we are to fully exploit genome sequence information in plants. Such assays might include over-expression or gene silencing, or the determination of cellular and subcellular localisation of mRNAs and proteins. The majority of techniques that currently exist to perform such assays rely on the production of transgenic plants, or vector-based transient transformation assays. Such methods are necessarily labour intensive and time-consuming, limiting the ability of most researchers to carry out genuine 'functional genomics' projects. However, several recent publications describe systems that permit the direct delivery of nucleic acids and proteins into plant cells in a functional state, providing the potential for rapid functional assays. Discussion Delivery of macromolecules into animal cells For many years, researchers using animal cell systems have used synthetic nucleic acids to manipulate gene expression. For example, the use of antisense oligodeoxynucleotides to suppress gene expression was first reported over a quarter of a century ago [1]. Single- and double-stranded DNA and RNA molecules can be introduced into mammalian cells by simple direct application to the culture media, or assisted by various transfection reagents, resulting in antisense or siRNA-mediated suppression of gene expression. A range of different modified nucleic acids that bring different characteristics in terms of stability and binding to target sequences are now used, such as morpholinos, locked nucleic acids, peptide nucleic acids, etc. [2]. In many cases these are being developed as potential therapeutic agents [2]. More recently, proteins and other macromolecules have been delivered into cells by linking them to so-called protein transduction domains (PTDs). These are short peptide sequences that when added to the N-terminus of a recombinant protein, or conjugated to other molecules, can carry those molecules directly into cells (reviewed in [3]). The best known are found in the HIV-1 transcriptional activator Tat, and the Drosophila transcription factor, Antennapedia [3]. PTDs are generally short, polybasic peptide sequences, and artificial polycationic peptides, such as polyarginine are also effective. Importantly, the uptake of molecules tagged with these peptides does not require specific receptors, endocytosis or active transport. The ability of PTDs to carry molecules across membranes is believed to be the result of the physical characteristics of their interactions with lipid bilayers, suggesting that they should work in any system. In the past, it has generally been assumed that such delivery systems would not work in plant cells, due to the presence of the cell wall and the difficulty of delivery to multicellular, differentiated tissues. However, work in several laboratories has recently shown that in fact, both proteins and nucleic acids can be efficiently delivered into plant cells in a functional form. Delivery of macromolecules into plant cells Unnamalai et al. [4], created double stranded RNA (dsRNA) in vitro, which was then allowed to complex with a 12mer polyarginine PTD via simple electrostatic interaction. Fluorescent labelling showed uptake of complexes into suspension cultured tobacco cells, characterised by initial accumulation in the nucleus and subsequent redistribution throughout the cytoplasm within 24 h. dsRNAs targeted against the NPTII and GUS marker genes specifically and substantially reduced gene expression via siRNA-mediated post-transcriptional gene silencing for at least 3 weeks following a 1 h treatment of cells with dsRNA:PTD complexes. An even simpler, but equally effective method for gene silencing was demonstrated by Sun et al., [5]. Single-stranded DNA oligonucleotides were taken up by cells of intact barley leaves when fed through the petiole (Figure 1). Fluorescent labelling again showed accumulation first in the nucleus, and then later throughout the cell. An antisense oligonucleotide, but not a complementary sense oligo, silenced gene expression via mRNA degradation [5]. The mechanism involved appears to be hybridisation of the antisense oligo to the mRNA to form an RNA-DNA duplex. RNA-DNA duplexes act as targets for ribonuclease H (RNase H) activity that cleaves RNA around the duplex. Figure 1 Oligonucleotides are taken up by intact plant tissues and are distributed throughout the cell. Confocal microscope image of cells from intact barley leaves following application of an 18-nucleotide oligodeoxynucleotide via the transpiration stream [5]. The oligonucleotide is labelled with the fluorescent dye Alexa Fluor 488 (Molecular Probes), and appears green in the image, whilst chloroplast autofluorescence appears red. Image provided by Professor Christer Jansson, Chuanxin Sun, Anna-Stina Höglund, Helena Olsson and Elke Mangelsen, The Swedish University of Agricultural Sciences. Peptide transduction domains have also been used to deliver proteins into plant cells. Again, the technique employed was remarkably simple and effective. Chang et al., [6], produced recombinant GFP proteins in E. coli, either alone or tagged with the Tat PTD or a 9mer polyarginine peptide (R9). When these purified proteins were applied to roots of onion or tomato plants, fluorescence rapidly became visible within the nuclei and cytoplasm of cells treated with Tat-GFP and R9-GFP, but not un-tagged GFP. Uptake of PTD-tagged GFP was detectable within 1 min of application, and was maximal in 5 min. Remarkably, cells throughout the root showed fluorescence – not just those in contact with the protein solution. As in animal systems, uptake was not affected by low temperature or inhibitors of endocytosis. GFP fluorescence was maintained for at least 2 days following a 5 min application, suggesting that PTDs are able to deliver proteins that can remain functional for a significant period of time. Conclusion The direct delivery of oligonucleotides and proteins to plant tissues has a range of exciting applications for the discovery of gene function (Table 1). So far, the publications discussed above have included only limited examples of these delivery techniques in plant tissues. An important question that needs to be addressed in the future is whether such molecules can be applied to plants in ways that enables the generation of useful information in a range of biological systems. Clearly the use of suspension cultured cells is limited, and although application through cut petioles may be suitable for short-term molecular and biochemical investigation, it would not permit long-term, developmental studies. Nevertheless, the power of these tools presents an exciting opportunity for further development. They have the potential to enable systematic, high-throughput studies of gene function in a range of plant species. Table 1 Applications of oligonucleotide and protein delivery into intact tissues. Molecule Application Reference ssDNA Antisense gene silencing [2] dsRNA Post-transcriptional gene silencing [7] Peptide nucleic acids Inhibition of gene expression by chromosomal interactions [8] Proteins Functional assays [3] Sub-cellular localisation of tagged proteins Competing interests The author(s) declare that they have no competing interests. ==== Refs Zamecnik PC Stephenson ML Inhibition of Rous sarcoma virus replication and cell transformation by a specific oligodeoxynucleotide Proc Natl Acad Sci USA 1978 75 280 284 75545 Dagle JM Weeks DL Oligonucleotide-based strategies to reduce gene expression Differentiation 2001 69 75 82 11798068 10.1046/j.1432-0436.2001.690201.x Wadia JS Dowdy SF Protein transduction technology Curr Opin Biotechnol 2002 13 52 56 11849958 10.1016/S0958-1669(02)00284-7 Unnamalai N Kang BG Lee WS Cationic oligopeptide-mediated delivery of dsRNA for post-transcriptional gene silencing in plant cells FEBS Lett 2004 566 307 310 15147914 10.1016/j.febslet.2004.04.018 Sun C Höglund A-S Olsson H Mangelsen E Jansson C Antisense oligodeoxynucleotide inhibition as a potent strategy in plant biology: identification of SUSIBA2 as a transcriptional activator in plant sugar signalling Plant J 2005 44 128 138 16167901 Chang M Chou J-C Lee H-J Cellular internalization of fluorescent proteins via arginine-rich intracellular delivery peptide in plant cells Plant Cell Physiol 2005 46 482 488 15695452 10.1093/pcp/pci046 Waterhouse PM Helliwell CA Exploring plant genomes by RNA-induced gene silencing Nat Rev Genet 2003 4 29 38 12509751 10.1038/nrg982 Kaihatsu K Janowski BA Corey DA Recognition of chromosomal DNA by PNAs Chem Biol 2004 11 749 758 15217608 10.1016/j.chembiol.2003.09.014	16356171	PMC1325021	CC BY	2021-01-04 16:25:04	no	Plant Methods. 2005 Dec 15; 1:12	utf-8	Plant Methods	2,005	10.1186/1746-4811-1-12	oa_comm
==== Front Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1421632422310.1186/1465-9921-6-142ResearchHyperresponsiveness to inhaled but not intravenous methacholine during acute respiratory syncytial virus infection in mice Collins Rachel A [email protected] Rosa C [email protected] Graeme R [email protected] Constance L [email protected] Debra J [email protected] Giuseppe N [email protected] Peter D [email protected] Division of Clinical Sciences, Telethon Institute for Child Health Research, Centre for Child Health Research, The University of Western Australia, PO Box 855, West Perth WA 6872, Australia2 Department of Pharmacology, Co-Operative Research Centre (CRC) for Chronic Inflammatory Diseases, University of Melbourne, Parkville, Victoria, Australia3 Department of Pediatrics, University of Texas Health Science Center – Houston, Texas, USA2005 5 12 2005 6 1 142 142 26 8 2005 5 12 2005 Copyright © 2005 Collins et al; licensee BioMed Central Ltd.2005Collins 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 characterise the acute physiological and inflammatory changes induced by low-dose RSV infection in mice. Methods BALB/c mice were infected as adults (8 wk) or weanlings (3 wk) with 1 × 105 pfu of RSV A2 or vehicle (intranasal, 30 μl). Inflammation, cytokines and inflammatory markers in bronchoalveolar lavage fluid (BALF) and airway and tissue responses to inhaled methacholine (MCh; 0.001 – 30 mg/ml) were measured 5, 7, 10 and 21 days post infection. Responsiveness to iv MCh (6 – 96 μg/min/kg) in vivo and to electrical field stimulation (EFS) and MCh in vitro were measured at 7 d. Epithelial permeability was measured by Evans Blue dye leakage into BALF at 7 d. Respiratory mechanics were measured using low frequency forced oscillation in tracheostomised and ventilated (450 bpm, flexiVent) mice. Low frequency impedance spectra were calculated (0.5 – 20 Hz) and a model, consisting of an airway compartment [airway resistance (Raw) and inertance (Iaw)] and a constant-phase tissue compartment [coefficients of tissue damping (G) and elastance (H)] was fitted to the data. Results Inflammation in adult mouse BALF peaked at 7 d (RSV 15.6 (4.7 SE) vs. control 3.7 (0.7) × 104 cells/ml; p < 0.001), resolving by 21 d, with no increase in weanlings at any timepoint. RSV-infected mice were hyperresponsive to aerosolised MCh at 5 and 7 d (PC200 Raw adults: RSV 0.02 (0.005) vs. control 1.1 (0.41) mg/ml; p = 0.003) (PC200 Raw weanlings: RSV 0.19 (0.12) vs. control 10.2 (6.0) mg/ml MCh; p = 0.001). Increased responsiveness to aerosolised MCh was matched by elevated levels of cysLT at 5 d and elevated VEGF and PGE2 at 7 d in BALF from both adult and weanling mice. Responsiveness was not increased in response to iv MCh in vivo or EFS or MCh challenge in vitro. Increased epithelial permeability was not detected at 7 d. Conclusion Infection with 1 × 105 pfu RSV induced extreme hyperresponsiveness to aerosolised MCh during the acute phase of infection in adult and weanling mice. The route-specificity of hyperresponsiveness suggests that epithelial mechanisms were important in determining the physiological effects. Inflammatory changes were dissociated from physiological changes, particularly in weanling mice. forced oscillationairway resistancephysiology ==== Body Introduction Respiratory syncytial virus (RSV) infection is one of the most common diseases of childhood. It is estimated that RSV infects up to two-thirds of infants worldwide by one year of age, with almost all children infected at least once by the age of 2 [1-3]. Around 75% of children have IgG antibodies to RSV by 18 months of age [4]. Most RSV disease manifests as mild upper respiratory tract infection, however a small proportion of children go on to develop severe lower respiratory tract disease including bronchiolitis and pneumonia requiring hospitalisation. Primary infection occurs at an average age of 12 months, though the median age of infants requiring hospital admission is 2 to 3 months [5] and the highest morbidity of RSV disease is seen below the age of 6 months [6-9]. Severe cases place a large burden on the health-care system; acute bronchiolitis and bronchitis are the sixth most common causes of hospital admissions in Australian children [10]. Acute RSV lower respiratory tract infection is associated with wheezing, airways hyperresponsiveness, airflow obstruction and alterations in gas exchange (reviewed in [11]). Mice are commonly used as experimental models of human RSV infection [12]. While inoculation with high titres of RSV is necessary for replication to occur within the lungs due to the semi-permissive nature of RSV infection in the mouse host, clinical and pathological changes vary markedly with dose. Infection with low titres (103 – 105 plaque forming units (pfu) induces peribronchial and perivascular inflammation [13-15] but fails to induce clinical signs of illness [15]. In contrast, infection with high titres of RSV (~107 pfu) induces clinical signs of illness and weight loss [15-19] in conjunction with severe histopathological changes and pneumonia [17,20,21] that can persist for long periods of time (154 days [20,21]. Current physiological data describing the effects of RSV infection are limited, particularly due to the use of the parameter 'enhanced pause' (Penh) derived from unrestrained plethysmography [20-23]. Penh is widely regarded as being primarily related to ventilatory timing and contains little information on the physiological state of the airways [24]. Few studies have examined the physiological response to bronchoconstrictor challenge in intubated mice infected with RSV [15,18,25] and the physiological alterations that occur in response to RSV are yet to be clearly defined in terms of the site of responsiveness and baseline changes in airway and parenchymal mechanics. The aim of the present study was to assess the physiological changes occurring in the airways and parenchyma of mice infected with RSV, and to relate these alterations to the inflammatory profile induced by infection. Due to the proven success of low dose RSV models in producing inflammatory and histopathological changes, we have used a low dose (105 pfu) model of infection in order to avoid the excessive pathology and structural damage that may confound our physiological measurements. We have also sought to determine whether the physiological response to primary RSV infection differs depending on age at infection. Materials and methods Animals BALB/c mice were selected for all studies due to their availability, level of responsiveness to bronchoconstrictor challenge and permissiveness to RSV infection [12]. Mice were obtained from the Animal Resource Centre (Murdoch, Western Australia) and maintained under specific pathogen free conditions at the Telethon Institute for Child Health Research (TICHR), with food and water available ad libitum. Experimental procedures were approved by the TICHR Animal Ethics Committee and conformed to the guidelines of the National Health and Medical Research Council of Australia. Infection of mice with RSV Mice were inoculated with 1 × 105 pfu of sucrose gradient purified human RSV A2 or the equivalent concentration of sucrose buffer as weanlings (21 d; weaning) or adults (8 wk). RSV was delivered to each mouse in a 30 μl inoculum under light anaesthesia (Methoxyfluorane, Medical Developments Pty Ltd, VIC, Australia) by pipetting drops of inoculum onto one nostril until the entire volume had been aspirated. Mice were laid on their side with their mouth held closed during inoculation to prevent ingestion. Mice were housed in individually ventilated cages (IVC Sealsafe, Tecniplast, Italy) during the acute phase of infection. Low velocity HEPA filtered air was delivered to cages maintained under negative pressure. Clinical signs of illness Mice were weighed and scored for clinical signs of illness daily until 7 d post inoculation and then every 2nd or 3rd day until 21 d. Mice were scored on the basis of appearance and demeanour, according to the scale described by Graham and colleagues [26]. A score of 0 indicated no visible signs of ill health; 1 – barely ruffled fur; 2 – ruffled but active; 3 – ruffled and inactive; 4 – ruffled, inactive, hunched and gaunt; 5 – dead. Mice were killed if they fell below 70% of their original bodyweight and/or had a clinical score of ≥ 3. Lung viral titre Viral titres were assessed in lung homogenates at 5 d post inoculation by TCID50 assay on HEp-2 cells as described in [27]. Measurement of lung function Anaesthesia Mice were anesthetized by intraperitoneal injection of 0.1 ml/10 g bodyweight of a mixture of ketamine (40 mg/ml, Troy Laboratories, NSW, Australia) and xylazine (2 mg/ml, Troy Laboratories, NSW, Australia). No muscle relaxants were used. Two thirds of the dose was used to induce surgical anaesthesia and the remainder was given once the mouse was attached to the ventilator. Additional doses were given as required. Once surgical anaesthesia was established a tracheotomy was performed by insertion of a straight polyethylene cannula (internal diameter = 0.086 cm, length = 1.0 cm) into the distal trachea. Oscillatory lung mechanics Mice were ventilated with a flexiVent® small animal ventilator (SCIREQ, Montreal, PQ, Canada) at 450 breaths per minute and a tidal volume of 8 ml/kg. A positive end-expiratory pressure was set at 2 hPa. The ventilation rate was set above the normal breathing rate to suppress spontaneous breathing during measurements. Mice were allowed to stabilize on the ventilator for 5 minutes before measurements commenced. Respiratory system impedance (Zrs) was measured using a modification of the low-frequency forced oscillation technique (FOT [28] as previously described [29]. Respiratory input impedance (Zrs) was measured between 0.5 and 20 Hz by applying a composite signal containing 19 mutually prime sinusoidal waves during pauses in regular ventilation. The peak-to-peak amplitude of the oscillatory signal was 50% of tidal volume. The flexiVent ventilator was used both for regular ventilation and for delivery of the oscillatory signal without the need to disturb the mice. Measurements were excluded if coherence was < 95%. Constant phase parameter estimation The constant-phase model described by Hantos et al. [30] was used to partition Zrs into components representing the mechanical properties of the airways and parenchyma. The constant-phase model [30] was fitted as follows: Zrs = R + jωI + (G-jH)/ωα, where R is the Newtonian resistance (primarily located in the airways but containing a contribution from the chest wall), I is the inertance, G is the co-efficient of tissue damping, H is the co-efficient of tissue elastance, ω is the angular frequency and α represents the reciprocal frequency-dependent behaviour of G & H. Strictly speaking, the parameters Raw and Iaw, respectively, include the Newtonian components of tissue resistance and tissue inertance. However, measurements in intact and open-chest rats [31,32] demonstrate that the contributions of the tissues to Raw and Iaw can be neglected. We have also previously shown that the chest wall makes little contribution to Newtonian resistance in mice and thus R ≈ Raw [33]. Methacholine challenge i) Aerosol MCh challenge Following measurement of baseline lung function, mice were challenged with a saline control aerosol followed by increasing concentrations of β-methacholine chloride (MCh; Sigma-Aldrich, MO, USA; 0.001 – 30 mg/ml). Aerosols were generated with an ultrasonic nebuliser (DeVilbiss UltraNeb 2000, Somerset, PA, USA) and delivered to the inspiratory line of the flexiVent using a bias flow of medical air. Each aerosol was delivered for 2 minutes during which time regular ventilation was maintained. Five measurements were made at one-minute intervals following each aerosol. The peak response at each MCh dose was compared to the mean response to saline. Responsiveness is expressed as the provocative concentration of MCh required to induce a doubling of Raw or a 50% increase in G and H (PC200 or PC150). Responsiveness to aerosolized MCh was assessed at 5, 7, 10 and 21 d post RSV infection and 5 and 21 d post control inoculation in 6–10 mice per group. These days were chosen to coincide with peak viral titres, peak inflammatory response, viral clearance and resolution of lung disease, respectively [12,13]. ii) Intravenous MCh challenge Intravenous MCh challenge was performed at 7 d post infection (n = 6–8 per group), the time of peak responsiveness to aerosolised MCh in both adult and weanling mice. Increasing doses of MCh were administered by constant infusion (3 – 96 μg/min/kg; Stoelting syringe pump, Wood Dale, IL, USA) via a polyethylene cannula (length = 27 cm; outer diameter = 0.061 cm) inserted into the jugular vein. MCh-induced constriction was reversed by intraperitoneal injection of atropine sulfate (120 μg or ~6 mg/kg; Pharmacia & Upjohn, WA, Australia; adapted from [34] during continued infusion of MCh at the highest rate. Responsiveness of tracheal segments in vitro Tracheal smooth muscle (TSM) responsiveness was assessed in vitro by electrical field stimulation (EFS) and MCh challenge at 7 d post infection (n = 6–7 RSV, n = 5–8 control from each age group). Mice were anaesthetised as per preparation for in vivo measurement of oscillatory mechanics. Tracheal segments of approximately 0.5 cm in length were removed and cleaned of loose connective tissue and placed in 50 ml organ baths (Radnotti Glass Technology, CA, USA). The TSM segment was attached to a fixed lower support and a tri-shape tissue support connected to a force-displacement transducer (Model FT03E; Grass Instrument Co., MA, USA). The tissue was suspended between horizontal platinum wire electrodes (AD Instruments, NSW, Australia). The tissues were bathed in modified Krebs-Henseleit solution containing (in mM): 118NaCl, 25NaHCO3, 2.8CaCl2.2H2O, 1.17MgSO4, 4.7 KCl, 1.2KH2PO4 and 11.1 glucose. The baths were aerated with a 95% O2-5% CO2 gas mixture. The temperature of the baths was maintained at 37°C. Each TSM segment was equilibrated in the bath for 30 min at an optimal resting tension of 0.70 g. During this equilibration time, the tissue was challenged once with 10-4 M MCh. Tissues that did not develop a contractile response were excluded from further studies. Tissues were rinsed with fresh Krebs-Henseleit solution periodically and allowed to relax to their initial tension after reaching maximal contraction. Recordings of resting tensions and TSM contractile responses were made using a PowerLab 8/s Recorder and Chart 5.1.1 software (AD Instruments, NSW, Australia). EFS (30 V, 3 ms square wave pulses at 0.5, 1, 2, 5, 10, 20, 30, 40 Hz) were delivered via platinum electrodes by a Grass S44 stimulator connected to a stimulus isolation unit (Grass Instruments, MA, USA). The stimulus was applied until the tissue reached a maximum contraction (~10 s). The tissue was washed after every second stimulation to ensure that the relative concentrations of the ions in the Krebs-Henseleit solution were maintained. EFS responsiveness is expressed as the frequency required to induce 50% of the maximal contractile response (EC50). To assess cholinergic sensitivity of the tissues, cumulative dose-response curves to MCh were performed in half-log increments employing concentrations ranging from 10-8 to 10-4 M. Results from MCh challenge are expressed as a percentage of the maximal contractile response as well as the EC50. Tissues were washed and rested repeatedly between EFS and MCh challenge. Bronchoalveolar lavage and lung fixation Lungs were lavaged at the completion of lung function measurements and just prior to death of the animal by washing 1 ml of ice-cold lavage fluid (0.9% saline containing 0.35% lidocaine (Sigma, St Louis, MO, USA) and 0.2 % BSA (CSL Ltd, Parkville, VIC, Australia) in and out of the lungs three times. Bronchoalveolar lavage fluid (BALF) was processed for total and differential cell counts. Cytospins for differential counts were stained with Leishmans stain (BDH Laboratory Supplies, Poole, England). Lavage supernatants were stored at -80°C. Total and differential cell counts were performed on lavage samples from 6–10 mice per group. Lungs were inflation fixed in situ in 10% phosphate-buffered formalin (Confix, Autralian Biostain Pty Ltd, VIC, Australia) at a distending pressure of 10 hPa for 1–2 hours before ligation and removal from the chest cavity. Lungs were immersion fixed in formalin overnight before being transferred to 70% ethanol and stored at 4°C until processing. Paraffin embedded lungs were sectioned at 5 μm thickness and stained with haematoxylin and eosin. Measurement of cytokines and mediators in BALF In order to characterise the primary inflammatory and cytokine response to RSV infection, we chose the appropriate kit to measure innate immune responses. This included tumour necrosis factor alpha (TNFα), interferon gamma (IFNγ), macrophage chemotactic protein 1 (MCP-1) and interleukins (IL) 6, 10 and 12 (p70 protein) and these were measured in BALF supernatants by cytometric bead assay (BD Biosciences, CA, USA) according to the manufacturer's instructions. Prostaglandin E2 (PGE2), IL-13, vascular endothelial growth factor (VEGF) and cysteinyl leukotrienes (cysLT) were measured as potential mediators of airway hyperresponsiveness using enzyme immunoassay kits (PGE2, cysLT: Cayman Chemicals, MI, USA; IL-13, VEGF: Quantikine, R&D Systems, MN, USA) according the manufacturer's instructions. Cytometric bead assay and cysLT ELISA were performed at 5, 7 and 21 d post RSV inoculation and at 5 and 21 d post diluent control inoculation. IL-13, VEGF and PGE2 were measured at 5 and 7 d post RSV inoculation and at 5 d post control inoculation. Measurement of epithelial permeability using Evans Blue dye Evans Blue dye (EBD) is a useful indicator of microvascular permeability [35]. EBD (Sigma-Aldrich, MO, USA) was administered intravenously to mice via the jugular vein following iv MCh challenge as described by Tulic et al. [36]. A slow bolus of 50 mg/kg EBD was delivered in a volume of 0.1 ml/10 g bodyweight through the existing iv cannula. Mice were ventilated for a further 30 minutes before post-EBD BAL was performed. The amount of EBD in BALF was quantified by reading the absorbance of the samples at 620 nm using a microplate reader (Bio-Tek Instruments, VT, USA). The amount of dye was calculated by interpolation on a standard curve in the range of 1 – 10 μg/ml [37]. Measurement of epithelial permeability was performed at 7 d post infection in adult mice only (n = 8 control, 7 RSV). Statistical analysis RSV groups were compared vs. combined control groups where no differences were observed between controls at 5 and 21 d. Differences in bodyweight, viral titre and EBD concentrations between groups were compared using unpaired t-test. Differences in total and differential cell counts, baseline physiology, cytokine and mediator assays were tested by 1-way analysis of variance (ANOVA) followed by Dunnett's post-hoc test for normally distributed data, and by Kruskal-Wallis ANOVA on ranks followed by Dunn's test for non normal data. Differences in MCh responsiveness in vivo between RSV infected and control animals were tested by 1-way ANOVA on PC200/150 data for aerosol MCh challenge, and by 2-way repeated measures ANOVA for iv MCh challenge. In vitro responsiveness of TSM segments was tested using 1-way ANOVA on EC50 data. Data are expressed as mean (SE). Graphs were prepared using SigmaPlot software (SigmaPlot 2000, SPSS Science, IL, USA). Statistical analysis was performed using SigmaStat software (version 2.03, SPSS Science, IL, USA). Significance was accepted at p < 0.05. Results Clinical illness Mice infected with RSV did not exhibit clinical signs of illness during the acute phase of infection. Adult mice infected with RSV did not decrease in bodyweight compared to controls (p = 0.41). RSV infected weanling mice gained weight at the same rate as control animals, both groups reaching 125–130% of their original bodyweight by 5 d post inoculation (p = 0.66; Figure 1). No mice were culled for excessive weight loss or clinical score ≥ 3. Figure 1 Total cells in BALF from adult and weanling mice inoculated with RSV or diluent control. Adult mice had significantly elevated total cell numbers in BALF at 7 and 10 d post inoculation that returned to control levels by 21 d. Weanling mice did not have increased cell numbers in BALF at any timepoint. Viral titre Adult and weanling mice had similar levels of RSV replication in lung homogenates at 5 d post inoculation (4.96 and 4.92 × 104 TCID50/g, respectively). Inflammation Adult mice Adult mice had significantly increased inflammatory cell numbers in BALF at 7 and 10 d post inoculation (p < 0.001). Cell numbers had returned to control levels by 21 d (Figure 1). Despite increased cell numbers, differential cell counts did not reveal a difference in the type of infiltrating cells at any timepoint and were dominated by macrophages (Figure 2). Mild peribronchiolar and perivascular inflammation was evident in histological sections at 5 d post RSV infection (Figure 3B), and had increased in severity at 7 d post infection (Figure 3C). Inflammatory cells were also visible in the lung parenchyma at 7 d (Figure 3C). Control mice did not show any evidence of inflammation at 5 d post inoculation (Figure 3A). Figure 2 Differential cell counts in adult and weanling mice after RSV and control inoculation. Macrophages were the predominant cell type in both age groups. Total macrophage and neutrophil numbers were increased in adult mice at 7 and 10 d post infection; however this did not reach statistical significance. Figure 3 Representative sections from adult (A-C) and weanling (D-F) mice inoculated with diluent control (A, D) or RSV (B, C, E, F), each showing an airway () and blood vessel (bv). Perivascular and peribronchiolar inflammation were evident to a small degree at 5 d post RSV (B); and to a much greater extent at 7 d post RSV (C) in adult mice. Some parenchymal inflammation was also present at 7 d. Little to no evidence of inflammation existed in weanling mice at 5 d post infection (E); however a small degree of perivascular and peribronchiolar inflammation was present at 7 d post infection (F). Based on morphology these cells were classified as lymphocytes. Control mice did not show any evidence of inflammation at either age (A, D). Bar = 50 μm. Weanling mice Inflammatory cell numbers in BALF did not change in weanling mice inoculated with RSV or diluent control (p = 0.191; Figure 1). Similarly, there was no difference in cell profile in BALF (Figure 2). Histological sections from weanling mice inoculated with diluent control and at 5 d post RSV infection showed little or no inflammatory infiltrate around airways, blood vessels or in the lung parenchyma (Figure 3D, E, respectively). Peribronchiolar and perivascular inflammation were evident to a small extent at 7 d post infection (Figure 3F), with infiltration of lymphocytes seen. Airway and parenchymal mechanics Baseline lung function In keeping with the mild inflammatory changes observed in histological sections, there was no evidence of airway obstruction or increased tissue stiffness at baseline in RSV-infected mice. RSV infection did not alter baseline Raw, G or H in adult mice (Table 1). Weanling mice had higher values of Raw, G and H than adult animals, consistent with age-related alterations in respiratory mechanics [38], although H decreased to approach adult values by 21 d (Table 1). Raw and G were not altered in RSV-infected weanling mice at baseline. H was decreased in weanling mice at 21 d post infection, but only when compared to 5 d controls (p = 0.003; p < 0.05 vs. 5 d control). Table 1 Baseline airway and tissue mechanics in adult and weanling mice. Values: mean (SE). Age Treatment Weight (g) Raw hPa.s.ml-1 G hPa.ml-1 H hPa.ml-1 Adult Control 5 d 19.3 (0.4) 0.33 (0.02) 5.1 (0.2) 37.3 (1.3) Control 21 d 17.9 (0.6) 0.33 (0.02) 5.4 (0.5) 36.5 (2.3) RSV 5 d 17.1 (0.2) 0.38 (0.03) 5.2 (0.2) 40.9 (1.8) RSV 7 d 18.2 (0.3) 0.35 (0.03) 5.9 (0.3) 44.3 (2.4) RSV 10 d 16.7 (0.3) 0.39 (0.03) 5.2 (0.3) 41.5 (2.1) RSV 21 d 18.7 (0.4) 0.43 (0.02) 4.8 (0.3) 40.3 (2.5) Weanling Control 5 d 13.9 (0.6) 0.51 (0.04) 7.0 (0.4) 61.6 (2.6) Control 21 d 16.7 (0.6) 0.48 (0.03) 6.5 (0.8) 57.7 (2.9) RSV 5 d 13.9 (0.5) 0.53 (0.08) 7.6 (0.4) 69.6 (5.7) RSV 7 d 15.1 (0.3) 0.52 (0.03) 7.8 (0.5) 64.0 (3.1) RSV 10 d 15.5 (0.5) 0.52 (0.05) 8.5 (0.7) 65.6 (6.7) RSV 21 d 16.3 (0.5) 0.39 (0.02) 6.3 (0.4) 45.1 (3.5) * p < 0.05 vs. d5 control, not significant vs. d21 control Responsiveness to MCh i) Aerosol MCh challenge Adult mice exhibited extreme hyperresponsiveness to aerosolised MCh (Figure 4) in both airway and tissue compartments at 5 and 7 d post RSV inoculation (Raw, G, H: p = 0.003, 0.007, <0.001, respectively), requiring an approximately 100-fold lower concentration of MCh than control animals to elicit a doubling of the response (Figure 5). The response to MCh at 10 d was more variable, with approximately half of the mice studied having returned to control levels of responsiveness by this timepoint. Responsiveness had returned to control levels in all animals studied by 21 d. Figure 4 Dose-response curves to aerosolised MCh challenge in adult mice showing airway resistance (A, B), tissue damping (C, D) and tissue elastance (E, F). Hyperresponsiveness was clearly evident in airways and tissues at 5 and 7 d post RSV infection (A, C, E). A mixed response was seen at 10 d post infection (B, D, and F). Figure 5 Concentrations of aerosolised MCh required to induce a doubling of the response to saline in the airways (A), or a 50% increase in the response of the lung parenchyma (B, C) of adult mice. Significantly lower concentrations of MCh were required to induce responses at 5 and 7 d post infection in Raw and H (A, C), and at 7 d in G (B). A mixed response was evident at 10 d post infection in both airway and tissue compartments. Weanling mice had more variable responses to MCh but were still hyperresponsive to aerosolised MCh at 5 and 7 d (p = 0.001, < 0.001, <0.001 for Raw, G, H respectively) (Figure 6). A mixed response was again seen at 10 d. Weanling mice required an approximately 10-fold lower concentration of MCh to elicit a response. Responsiveness had returned to control levels by 21 d. Figure 6 Concentrations of aerosolised MCh required to induce a doubling of the saline response in the airways (A), or a 50% increase in the response of the lung parenchyma (B, C) of weanling mice. Significantly lower concentrations of MCh were required to induce responses at 7 and 10 d post infection in Raw (A), 7 d in G (B) and 5 and 7 d in H (C). The response at 10 d post infection was more consistent than in adult mice, however responsiveness in general was much more variable in weanlings. ii) Intravenous MCh challenge Neither adult nor weanling mice exhibited increased airway or tissue responsiveness to iv MCh compared to controls at 7 d post inoculation. Weanling mice infected with RSV were slightly smaller than controls (control 13.7 (0.35) g; RSV 11.8 (0.35) g, causing a small upward shift in the curve that was not related to altered responsiveness (Figure 7). This weight difference was maintained from the time of inoculation and due to variation in litter size rather than weight loss from RSV-induced illness. Figure 7 Airway resistance in response to iv MCh challenge in adult and weanling mice at 7 d post infection. RSV infected mice (closed symbols) did not demonstrate increased responsiveness to any concentration of iv MCh compared to controls (open symbols) at either age. Raw returned to baseline levels following atropine administration. RSV-infected weanling mice had slightly elevated Raw throughout the iv challenge, although this was due to their smaller size rather than altered responsiveness. In vitro responsiveness TSM segments from adult and weanling mice infected with RSV did not exhibit increased responsiveness to EFS post inoculation (adult EC50 (Hz): RSV 2.59 (1.32) vs control 1.68 (0.56); weanling EC50 (Hz) RSV 2.23 (0.74) vs control 1.77(0.84). Similarly, there was no change in responsiveness to MCh at 7 d (Figure 8). Figure 8 Responsiveness of isolated TSM segments from adult and weanling mice to MCh challenge at 7 d post infection. RSV infected mice (closed symbols) did not demonstrate altered bronchoconstrictor responses in vitro compared to controls (open symbols) when infected as either adults or weanlings. Cytokines and mediators in BALF Cytometric bead assay IL-12 p70 was not detectable in BALF from adult mice at any timepoint, irrespective of treatment (data not shown). TNFα, IFNγ, MCP-1 and IL-6 were undetectable in control samples and at 5 and 21 d post RSV inoculation but were significantly increased at 7 d post RSV (p < 0.001; Figure 9A–C). IL-6 was increased at 7 d but did not reach significance in post-hoc analysis (p = 0.011; Figure 9D). IL-10 levels were not altered by RSV infection (p = 0.125, data not shown). Figure 9 Cytokine levels in BALF from adult mice. TNFα (A), IFNγ (B) and MCP-1 (C) concentrations were all increased at 7 d post infection and undetectable in controls and at other timepoints. IL-6 was elevated to a small extent at 7 d (D) but did not reach significance in post-hoc analysis. IL-12 p70, TNFα, IFNγ, MCP-1 and IL-6 were all below detectable levels in BALF from weanling mice at all timepoints (data not shown). Although detectable, IL-10 levels were not altered by RSV infection (data not shown). Prostaglandin E2 PGE2 was elevated in BALF from both adult and weanling mice, peaking at 7 d post infection (p < 0.001) (Figure 10). Figure 10 PGE2 levels in BALF from adult and weanling mice. Significantly elevated levels were detected at 7 d post RSV infection in both age groups. Cysteinyl leukotrienes Increased levels of cysLT were detected in BALF from adult and weanling mice (p = 0.029, 0.009 respectively), peaking at 5 d post RSV inoculation (Figure 11). Despite a great deal of variability, the increase was significant in adult mice at 5 d (p < 0.05), but did not reach significance in weanling mice. Figure 11 cysLT concentrations in BALF from adult and weanling mice. Elevated levels were detected at 5 d post RSV infection in adults and weanlings, although this did not reach significance in post-hoc analysis in weanlings. Upper and lower limits of detection for this assay were 8 and 500 pg/ml, respectively. IL-13 IL-13 was undetectable in all samples (data not shown). VEGF VEGF was elevated at 7 d post infection in both adult and weanling mice (both p < 0.001). Neither age group had elevated VEGF levels at 5 d post infection (Figure 12). Figure 12 VEGF levels in BALF from adult and weanling mice. Elevated levels were detected at 7 d post RSV infection. Microvascular permeability Microvascular permeability measured by Evans blue dye extravasation into BALF was not increased in adult mice at 7 d (p = 0.25; data not shown). Microvascular permeability was not measured in weanling mice. Discussion Our low dose model of RSV infection was successful in achieving viral replication and physiological alterations to airway (Raw) and parenchymal (G, H) function in the lungs of both adult and weanling mice. The level of responsiveness was somewhat dissociated from the observed inflammatory changes, particularly in the younger mice. As expected with the dose of RSV administered [15], mice did not lose weight or show clinical signs of illness during the acute phase of infection. Weanling mice infected with RSV gained weight at the same rate as controls. Inflammatory changes Adult mice showed modestly elevated inflammatory cell numbers in BALF that peaked at 7 d post infection and returned to control levels by 21 d (Figure 1). Inflammatory cell numbers were not elevated in adult mice at 5 d post infection, indicative of a delay between viral replication in the lung and initiation of the cell-mediated immune response. Weanling mice did not show any increase in infiltrating inflammatory cell numbers above control mice at any timepoint, suggesting that this level of viral replication was insufficient to induce a detectable cell-mediated immune response. The cell profile was not altered in either age group, and consisted predominantly of macrophages at all timepoints (Figure 2). Histological sections revealed a mild peribronchiolar and perivascular inflammation 5 d post infection in adult mice, becoming more extensive at 7 d. Weanling mice at 5 d did not appear different to controls, however a small degree of inflammation was seen in weanling mice at 7 d post infection (Figure 3). The cytokine profile measured in BALF from adult and weanling mice mirrored the inflammatory profile, with adult mice showing increased levels of TNFα, IFNγ, MCP-1, IL-6 and VEGF at 7 d post infection but not at 5 d (not measured at 10 d) (Figures 9 and 12). In keeping with inflammatory cell numbers, weanling mice did not have elevated TNFα, IFNγ, MCP-1 or IL-6 in BALF at any timepoint, although VEGF was elevated at 7 d post infection (Figure 12). IL-13 has previously been identified as an important mediator of AHR in RSV infected DBA/J and BALB/c mice [25], however it was undetectable in the present study. In the absence of significant numbers of infiltrating lymphocytes, the absence of IL-13 in these samples is unsurprising. These results suggest that the physiological changes observed in RSV infected mice in the present study were not due to cell-mediated inflammatory processes. Physiological changes Baseline physiology RSV infection did not alter baseline airway or parenchymal mechanics in mice infected as adults or weanlings (Table 1). Age-related differences in respiratory mechanics were apparent between age groups and in weanling mice between the 5 and 21 d timepoints, although this pattern was not altered by RSV infection. These data suggest that the low levels of inflammation observed in tissue sections were insufficient to cause airway obstruction (increased Raw) or stiffening of parenchymal tissues (increased G and/or H). Aerosol MCh Both adult and weanling mice demonstrated extreme airway and parenchymal hyperresponsiveness to aerosolised MCh, although the response in weanlings was more variable. Adult mice required an approximately 100-fold lower dose of MCh than controls for a doubling response (Figures 4, 5); weanlings required on average 10-fold less MCh than control animals (Figure 6). The concentration of MCh required for response seen in these mice is substantially lower than has been demonstrated by other studies (generally requiring ~10 mg/ml) using similar infective doses of RSV [22,23,39,40]. Control mice responded at a somewhat lower MCh dose than naïve BALB/c routinely measured in our laboratory (data not shown), indicating that the sucrose buffer solution may have induced some degree of hyperresponsiveness. Despite the level of responsiveness of the control animals, RSV infection still induced a clear leftward shift of both the airway and parenchymal dose-response curves representing increased sensitivity to bronchoconstrictor challenge. AHR to MCh in RSV-infected mice has primarily been detected using unrestrained plethysmography [22,23,39], an inherently non-specific means of measuring airway function [24]. We were not surprised at differences between our results and those obtained with Penh, given that FOT contains direct physiological information on airway and parenchymal behaviour. Penh data obtained during MCh challenge is also likely to be contaminated by increased nasal resistance due to respiratory secretions induced by cholinergic stimulation. A potential advantage of unrestrained plethysmography over FOT may lie in the ability to test unsedated animals, but we would expect sedation to reduce responsiveness rather than increase it [41,42]. More reliable physiological data comes from Dakhama et al. [40], who demonstrated AHR to MCh in intubated mice using total lung resistance. The magnitude of the response detected in the present study coupled with the similar pattern of responsiveness in the airway and parenchymal compartments suggests that degree of sensitivity to MCh detected in the present study is not simply a function of a more sensitive measurement technique relative to other studies. While partitioning of Zrs into airway and tissue mechanics allows detection of more subtle changes than total lung or respiratory system impedance, it seems unlikely that responses of this magnitude would not be detected using other measurement systems. Intravenous MCh and MCh in vitro In contrast to challenge with aerosolised MCh, responsiveness to iv MCh challenge was not altered in mice infected with RSV as adults or weanlings (Figure 7). Similarly, we did not observe increased responsiveness to MCh in tracheal segments from adult or weanling mice at 7 d (Figure 8). The contrasting effects of different routes of agonist delivery suggest that delivery of MCh directly to the epithelial surface is crucial in inducing hyperresponsiveness. Conflicting data exists in the literature on the response to iv MCh in RSV-infected mice; similar doses of iv MCh have been shown to induce hyperresponsiveness in mice infected with 3 × 105 pfu RSV [25] and 107 pfu RSV [15], and to be unable to induce hyperresponsiveness in mice infected with 107 pfu RSV [18]. Responsiveness to EFS in vitro In the present study, responsiveness of tracheal segments to EFS in vitro was unaltered in RSV-infected mice. Using a slightly higher dose of RSV (106 pfu), Dakhama et al. [40] demonstrated increased responsiveness of murine tracheal smooth muscle segments to EFS at 6 d post infection, without any increase in maximal tension. Similarly, Colasurdo and colleagues have demonstrated increased responsiveness to EFS but no change in maximal tension at 4 d post infection in cotton rats [43]. The reason for the discrepancy in EFS responsiveness between published studies and the results we report here is not clear. While the dose used by Dakhama et al. [40] was somewhat higher, inflammatory cell numbers in BALF were identical (in adult mice) to our study, suggesting similar severity of infection. Differences in responsiveness between mice and cotton rats does not seem surprising given the greater permissiveness of cotton rats to RSV infection, however the similar pattern of responses seen in the aforementioned studies suggests that species differences do not play a major role. The lack of an increased responsiveness to EFS in vitro does, however, argue against any alterations in neural control of airway smooth muscle (ASM) and the lack of increased responsiveness to MCh in vitro argues against alterations in ASM contractile properties following RSV infection. Mechanisms of route-specific hyperresponsiveness The discrepancy in responsiveness between in vivo and in vitro conditions may reflect differences in the site of responsiveness within the airway tree. The use of extrathoracic tracheal segments in vitro may ignore alterations in airway function occurring further down the airway tree. This may be particularly relevant when the site of RSV replication within the mouse lung is considered; the virus replicates mostly in small airways and alveolar epithelial cells rather than epithelial cells in large conducting airways. While technically more difficult, assessment of the responsiveness of bronchi or intrathoracic airways may yield a greater response in vitro [44]. Airway resistance as measured in vivo represents a greater proportion of the airway tree and thus may be more sensitive to changes occurring in regions of the airway other than the trachea. Differences in levels of responsiveness between different modes of agonist delivery in the present study bear strong resemblance to studies investigating the effects of cationic proteins on airway function published in the 1990's [45-47]. After the discovery that treatment with cationic proteins (major basic protein, poly-L-lysine and poly-L-arginine) increased airway responsiveness to inhaled MCh in rats [47], further investigation in isolated airways revealed that hyperresponsiveness was induced only if MCh was delivered to the luminal surface of the airway in vitro. Delivery of MCh to the external surface of the airway wall did not alter responsiveness irrespective of pre-treatment with cationic proteins [45,46]. Thus, hyperresponsiveness was only manifested when the challenging agonist had to cross the airway epithelium to reach the underlying ASM, and implicates the integrity of the epithelial layer as important in the response to inhaled agonists. Epithelial mechanisms are the most likely candidates to explain the results of the present study, in which hyperresponsiveness was only induced by aerosol delivery of MCh in vivo. We did not examine luminal vs. external administration of MCh in vitro, but can instead compare aerosol and iv delivery in vivo. Disruption of epithelial barrier function or alterations in signalling are potential candidates to explain the site-specificity of agonist action of airway responsiveness. VEGF is upregulated in nasal washings from RSV-infected children [48] and in human epithelial cells infected with RSV in vitro [48,49]. VEGF was found to be responsible for increased permeability of RSV-infected epithelial monolayers in culture [49]. In the present study, VEGF was detected at elevated levels in BALF of both adult and weanling mice at 7 d post infection but was not elevated at 5 d (Figure 12), suggesting that it may play a role in increasing airway responsiveness but cannot account for increased responsiveness at all timepoints. Despite elevated levels of VEGF, and contrary to the results of Kilani et al. in vitro [49], epithelial permeability as measured by Evans Blue dye extravasation into BALF was not increased at the time of peak responsiveness (7 d post infection) in RSV infected mice in vivo. These results suggest that epithelial barrier function remained intact in these animals; however the level of permeability of the epithelium in these mice simply may not have been great enough to allow extravasation of the large Evans Blue dye-albumin complex whilst being sufficient to allow greater access of MCh (~300 × smaller than albumin) to the ASM. Alternatively, the ability of the technique to detect small increases in epithelial permeability may have been limited by binding of Evans Blue dye to airway tissues [50]. Surfactant acts as part of the airway mucosal barrier and may be involved in the epithelial-specific response to MCh challenge observed in the present study. Multiple layers of phospholipid bind directly to the surface of the bronchial epithelium (reviewed in [51] and contribute to epithelial barrier function by masking bronchial irritant receptors that respond to MCh challenge [52,53]. Disruption of surfactant function by RSV infection (as demonstrated in mice by [54] and unmasking of irritant receptors may be sufficient to enhance the level of responsiveness to aerosolised MCh, although we have not performed any studies to directly test this hypothesis. Alterations in responsiveness induced by surfactant dysfunction in RSV infection would only be detected by MCh delivery directly to the epithelial surface and would not require increased epithelial permeability. Bypassing the epithelial layer by iv administration of MCh and the loss of the surfactant layer in vitro would not reveal disruption of surfactant function, and are supported by the results of the present study. Role of cytokines and mediators in AHR Airway and parenchymal hyperresponsiveness were dissociated from inflammatory changes in the low dose model of RSV infection used in the present study. The level of hyperresponsiveness seen in these mice far exceeded that which would be expected for the measured levels of inflammation. The dissociation between inflammatory and physiological changes is further emphasised by the similarity of physiological responses between adult and weanling mice despite their differing inflammatory profiles. In the absence of a significant population of inflammatory cells, the innate immune response and products of resident cells become potential candidates for induction of hyperresponsiveness. The potential roles of leukotrienes and prostaglandins in the hyperresponsiveness seen in RSV-infected mice in the present study were investigated due to their ability to influence ASM contraction. Cysteinyl leukotrienes have profound effects on airway function; they are potent activators of ASM contraction [55], they act on the vasculature to produce vasodilation and increase vascular permeability [56] and stimulate mucus secretion and interfere with mucociliary clearance [57]. The role of PGE2 in regulating airway function is more complex, due in part to the existence of four separate cell surface receptors with unique signal transduction mechanisms [58]. PGE2 is potently bronchoprotective in vitro (reviewed in [59], but has been shown to induce bronchoconstriction as well as bronchoprotection in humans [60,61] and animal models [62,63]. Although the distribution of PGE2 receptors within the lung has not been fully defined, mRNA expression of all four types has been detected in the mouse lung [64,65]. Airway epithelial cells, mast cells and alveolar macrophages are local sources of cysLT and PGE2, both of which have been shown to be elevated in the airways of children with bronchiolitis [66-68]. Cysteinyl leukotrienes were significantly elevated in both adult and weanling mice infected with RSV prior to influx of inflammatory cells at 7 d post infection (Figure 11), suggesting that epithelial cells were the main source. cysLT expression was detected earlier than PGE2, peaking at 5 d post infection in adult mice, and only elevated at 5 d in weanling mice. Although consistently detectable at 5 d post RSV in both age groups, cysLT levels varied markedly between animals. PGE2 was detected at significantly elevated levels in BALF from both adult and weanling mice at 7 d post infection (Figure 10). The similarity between adult and weanling PGE2 levels despite the dramatically different infiltrating cell populations at these two ages again suggests that epithelial cells were the predominant source. The extreme sensitivity of both adult and weanling mice to MCh challenge at the peak of PGE2 production indicates that the bronchoprotective effect of PGE2 at these concentrations was not sufficient to inhibit airway responsiveness. Coyle et al (JCI 1995) demonstrated that the cationic proteins major basic protein and poly-L-lysine increased immunoreactive kinins and kallikrein-like activity in vivo and that this mechanism explained the epithelial-dependant increase in MCh responsiveness. We did not have the opportunity of studying kinins and so can not comment on whether similar mechanisms may underlie the epithelial-dependent MCh responsiveness we report following RSV infection. Age-dependent effects of RSV infection Despite equal levels of viral replication in adult and weanling mice, significant differences were observed in the inflammatory response to RSV. The lack of a significant cell-mediated immune response in weanling mice suggests that differences in the level of the host innate immune response to RSV may have been responsible for the disparity in responsiveness between the two age groups. The lack of MCP-1 expression detected in BALF from weanling mice may be indicative of a general paucity of chemoattractant chemokine production in this age group. The similarity of physiological responses despite marked differences in cell mediated immunity between adult and weanling mice highlights the issues associated with characterising infection models solely in adult animals with mature immune systems. These data also argue for the need for a systematic study of the effect of age on the effects of viral infections in mouse models. Summary Infection of adult and weanling mice with 1 × 105 pfu RSV induced significant alterations in airway and parenchymal responsiveness to bronchoconstrictor challenge. Increased responsiveness occurred in the absence of baseline changes in airway or parenchymal physiology, and in conjunction with mild inflammatory changes. Route-specificity of MCh responsiveness and elevated levels of epithelial-derived mediators indicated that epithelial mechanisms were the main determinants of altered respiratory function. The model described in the present study may provide a useful basis for assessment of the specific physiological effects of mild RSV lower respiratory tract infection on airway function. Although great caution should always be maintained when translating data from mouse models to humans, VEGF-mediated increases in epithelial permeability may be a mechanism by which RSV mediates airways hyperresponsiveness in the human disease. Acknowledgements This work was supported by the National Health and Medical Research Council of Australia grant #139024. Rachel Collins was supported by scholarships from the National Health and Medical Research Council of Australia and the CRC for Asthma. ==== Refs Glezen P Denny FW Epidemiology of acute lower respiratory disease in children N Engl J Med 1973 288 498 505 4346164 Glezen WP Taber LH Frank AL Kasel JA Risk of primary infection and reinfection with respiratory syncytial virus Am J Dis Child 1986 140 543 546 3706232 Openshaw PJ Immunity and immunopathology to respiratory syncytial virus. The mouse model Am J Respir Crit Care Med 1995 152 S59 62 7551415 Harsten G Prellner K Lofgren B Heldrup J Kalm O Kornfalt R Serum antibodies against respiratory tract viruses: a prospective three-year follow-up from birth J Laryngol Otol 1989 103 904 908 2584849 Schmidt AC Johnson TR Openshaw PJM Braciale TJ Falsey AR Anderson LJ Wertz GW Groothuis JR Prince GA Melero JA Graham BS Respiratory syncytial virus and other pneumoviruses: a review of the international symposium-RSV 2003 Virus Res 2004 106 1 13 15522442 10.1016/j.virusres.2004.06.008 Glezen WP Paredes A Allison JE Taber LH Frank AL Risk of respiratory syncytial virus infection for infants from low-income families in relationship to age, sex, ethnic group, and maternal antibody level J Pediatr 1981 98 708 715 7229749 Green M Brayer AF Schenkman KA Wald ER Duration of hospitalization in previously well infants with respiratory syncytial virus infection Pediatr Infect Dis J 1989 8 601 605 2797955 Hall CB Hall WJ Speers DM Clinical and physiological manifestations of bronchiolitis and pneumonia. Outcome of respiratory syncytial virus Am J Dis Child 1979 133 798 802 463835 Parrott RH Kim HW Arrobio JO Hodes DS Murphy BR Brandt CD Camargo E Chanock RM Epidemiology of respiratory syncytial virus infection in Washington, D.C. II. Infection and disease with respect to age, immunologic status, race and sex Am J Epidemiol 1973 98 289 300 4355170 Moon L Rahman N Bhatia K Australia's children: their health and wellbeing 1998 1998 Canberra, AIHW AIHW Catalogue no. PHE 7 Lemanske RFJ The childhood origins of asthma (COAST) study Pediatr Allergy Immunol 2002 13 38 43 12688623 10.1034/j.1399-3038.13.s.15.8.x Prince GA Horswood RL Berndt J Suffin SC Chanock RM Respiratory syncytial virus infection in inbred mice Infect Immun 1979 26 764 766 546793 Taylor G Stott EJ Hughes M Collins AP Respiratory syncytial virus infection in mice Infect Immun 1984 43 649 655 6693171 Anderson JJ Norden J Saunders D Toms GL Scott R Analysis of the local and systemic immune responses induced in BALB/c mice by experimental respiratory syncytial virus infection J Gen Virol 1990 71 1561 1570 2197371 van Schaik SM Enhorning G Vargas I Welliver RC Respiratory syncytial virus affects pulmonary function in BALB/c mice J Infect Dis 1998 177 269 276 9466511 Stack AM Malley R Saladino RA Montana JB MacDonald KL Molrine DC Primary respiratory syncytial virus infection: pathology, immune response, and evaluation of vaccine challenge strains in a new mouse model Vaccine 2000 18 1412 1418 10618539 10.1016/S0264-410X(99)00399-0 Hayes PJ Scott R Wheeler J In vivo production of tumour necrosis factor-alpha and interleukin-6 in BALB/c mice inoculated intranasally with a high dose of respiratory syncytial virus J Med Virol 1994 42 323 329 8046422 Peebles RSJ Sheller JR Johnson JE Mitchell DB Graham BS Respiratory syncytial virus infection prolongs methacholine-induced airway hyperresponsiveness in ovalbumin-sensitized mice J Med Virol 1999 57 186 192 9892406 10.1002/(SICI)1096-9071(199902)57:2<186::AID-JMV17>3.0.CO;2-Q Stark JM McDowell SA Koenigsknecht V Prows DR Leikauf JE Le Vine AM Leikauf GD Genetic susceptibility to respiratory syncytial virus infection in inbred mice J Med Virol 2002 67 92 100 11920822 10.1002/jmv.2196 Mejias A Chavez-Bueno S Rios AM Saavedra-Lozano J Fonseca Aten M Hatfield J Kapur P Gomez AM Jafri HS Ramilo O Anti-respiratory syncytial virus (RSV) neutralizing antibody decreases lung inflammation, airway obstruction, and airway hyperresponsiveness in a murine RSV model Antimicrob Agents Chemother 2004 48 1811 1822 15105140 10.1128/AAC.48.5.1811-1822.2004 Jafri HS Chavez-Bueno S Mejias A Gomez AM Rios AM Nassi SS Yusuf M Kapur P Hardy RD Hatfield et Respiratory syncytial virus induces pneumonia, cytokine response, airway obstruction, and chronic inflammatory infiltrates associated with long-term airway hyperresponsiveness in mice J Infect Dis 2004 189 1856 1865 15122522 10.1086/386372 Schwarze J Hamelmann E Bradley KL Takeda K Gelfand EW Respiratory syncytial virus infection results in airway hyperresponsiveness and enhanced airway sensitization to allergen J Clin Invest 1997 100 226 233 9202075 Cho JY Miller M Baek KJ Castaneda D Nayar J Roman M Raz E Broide DH Immunostimulatory DNA sequences inhibit respiratory syncytial viral load, airway inflammation, and mucus secretion J Allergy Clin Immunol 2001 108 697 702 11692091 10.1067/mai.2001.119918 Bates J Irvin C Brusasco V Drazen J Fredberg J Loring S Eidelman D Ludwig M Macklem P Martin J Milic-Emili J Hantos Z Hyatt R Lai-Fook S Leff A Solway J Lutchen K Suki B Mitzner W Pare P Pride N Sly P The use and misuse of Penh in animal models of lung disease Am J Respir Cell Mol Biol 2004 31 373 374 15317683 Tekkanat KK Maassab HF Cho DS Lai JJ John A Berlin A Kaplan MH Lukacs NW IL-13-induced airway hyperreactivity during respiratory syncytial virus infection is STAT6 dependent J Immunol 2001 166 3542 3548 11207314 Graham BS Bunton LA Wright PF Karzon DT Reinfection of mice with respiratory syncytial virus J Med Virol 1991 34 7 13 1885946 Hierholzer JC Killington RA Mahy BWJ and Kangro HO Virus isolation and quantitation Virology Methods Manual 1996 London, Academic Press 25 46 Sly PD Hayden MJ Petak F Hantos Z Measurement of low-frequency respiratory impedance in infants Am J Resp Crit Care Med 1996 154 161 166 8680673 Pillow JJ Korfhagen TR Ikegami M Sly PD Overexpression of TGF-alpha increases lung tissue hysteresivity in transgenic mice J Appl Physiol 2001 91 2730 2734 11717240 Hantos Z Daroczy B Suki B Nagy S Fredberg JJ Input impedance and peripheral inhomogeneity of dog lungs J Appl Physiol 1992 72 168 178 1537711 10.1063/1.352153 Hirai T McKeown KA Gomes RFM Bates JHT Effects of lung volume on lung and chest wall mechanics in rats J Appl Physiol 1999 86 16 21 9887108 Sakai H Ingenito EP Mora R Abbay S Cavalcante FSA Lutchen KR Suki B Hysteresivity of the lung and tissue strip in the normal rat: effects of heterogeneities J Appl Physiol 2001 91 737 747 11457789 Sly PD Collins RA Thamrin C Turner DJ Hantos Z Volume dependence of airway and tissue impedances in mice J Appl Physiol 2003 94 1460 1466 12391040 Brown RH Mitzner W Airway response to deep inspiration: role of inflation pressure J Appl Physiol 2001 91 2574 2578 11717221 Patterson CE Rhoades RA Garcia JG Evans blue dye as a marker of albumin clearance in cultured endothelial monolayer and isolated lung J Appl Physiol 1992 72 865 873 1568982 10.1063/1.351760 Tulic MK Wale JL Holt PG Sly PD Modification of the inflammatory response to allergen challenge after exposure to bacterial lipopolysaccharide Am J Resp Cell Mol Biol 2000 22 604 612 Bell SC Rynell AC Matheson MJ Finnimore AJ Berend N Inhaled FMLP increases microvascular permeability in the rabbit trachea J Appl Physiol 1993 74 1337 1341 8482675 Bozanich EM Collins RA Thamrin C Hantos Z Sly PD Turner DJ Developmental changes in airway and tissue mechanics in mice J Appl Physiol 2005 99 108 113 15817717 10.1152/japplphysiol.01111.2004 Stark JM Khan AM Chiappetta CL Xue H Alcorn JL Colasurdo GN Immune and functional role of nitric oxide in a mouse model of respiratory syncytial virus infection J Infect Dis 2005 191 387 395 15633098 10.1086/427241 Dakhama A Park JW Taube C El Gazzar M Kodama T Miyahara N Takeda K Kanehiro A Balhorn A Joetham A Loader JE Larsen GL Gelfand EW Alteration of airway neuropeptide expression and development of airway hyperresponsiveness following respiratory syncytial virus infection Am J Physiol Lung Cell Mol Physiol 2005 288 L761 770 15608150 10.1152/ajplung.00143.2004 Holtzman MJ Hahn HL Sasaki K Skoogh BE Graf PD Fabbri LM Nadel JA Selective effect of general anesthetics on reflex bronchoconstrictor responses in dogs J Appl Physiol 1982 53 126 133 7118625 Skoogh BE Holtzman MJ Sheller JR Nadel JA Barbiturates depress vagal motor pathway to ferret trachea at ganglia J Appl Physiol 1982 53 253 257 7118638 Colasurdo GN Hemming VG Prince GA Loader JE Graves JP Larsen GL Human respiratory syncytial virus affects nonadrenergic noncholinergic inhibition in cotton rat airways Am J Physiol Lung Cell Mol Physiol 1995 268 L1006 11 Chiba Y Ueno A Sakai H Misawa M Hyperresponsiveness of bronchial but not tracheal smooth muscle in a murine model of allergic bronchial asthma Inflamm Res 2004 53 636 642 15693613 10.1007/s00011-004-1305-x Coyle AJ Mitzner W Irvin CG Cationic proteins alter smooth muscle function by an epithelium-dependent mechanism J Appl Physiol 1993 74 1761 1768 8514694 Coyle AJ Ackerman SJ Irvin CG Cationic proteins induce airway hyperresponsiveness dependent on charge interactions Am Rev Respir Dis 1993 147 896 900 8385431 Uchida DA Ackerman SJ Coyle AJ Larsen GL Weller PF Freed J Irvin CG The effect of human eosinophil granule major basic protein on airway responsiveness in the rat in vivo. A comparison with polycations Am Rev Respir Dis 1993 147 982 988 8466137 Lee CG Yoon HJ Zhu Z Link H Wang Z Gwaltney JMJ Landry M Elias JA Respiratory syncytial virus stimulation of vascular endothelial cell growth factor/vascular permeability factor Am J Respir Cell Mol Biol 2000 23 662 669 11062145 Kilani MM Mohammed KA Nasreen N Hardwick JA Kaplan MH Tepper RS Antony VB Respiratory syncytial virus causes increased bronchial epithelial permeability Chest 2004 126 186 191 15249461 10.1378/chest.126.1.186 Saria A Lundberg JM Evans blue fluorescence: quantitative and morphological evaluation of vascular permeability in animal tissues J Neurosci Methods 1983 8 41 49 6876872 10.1016/0165-0270(83)90050-X Hills BA An alternative view of the role(s) of surfactant and the alveolar model J Appl Physiol 1999 87 1567 1583 10562593 Hills BA Asthma: is there an airway receptor barrier? Thorax 1996 51 773 776 8795662 Hills BA Chen Y Suppression of neural activity of bronchial irritant receptors by surface-active phospholipid in comparison with topical drugs commonly prescribed for asthma Clin Exp Allergy 2000 30 1266 1274 10971473 10.1046/j.1365-2222.2000.00850.x van Schaik SM Vargas I Welliver RC Enhorning G Surfactant dysfunction develops in BALB/c mice infected with respiratory syncytial virus Pediatr Res 1997 42 169 173 9262218 Barnes NC Piper PJ Costello JF Comparative effects of inhaled leukotriene C4, leukotriene D4, and histamine in normal human subjects Thorax 1984 39 500 504 6463929 Drazen JM Austen KF Lewis RA Clark DA Goto G Marfat A Corey EJ Comparative airway and vascular activities of leukotrienes C-1 and D in vivo and in vitro PNAS 1980 77 4354 4358 6933488 Marom Z Shelhamer JH Bach MK Morton DR Kaliner M Slow-reacting substances, leukotrienes C4 and D4, increase the release of mucus from human airways in vitro Am Rev Respir Dis 1982 126 449 451 7125334 Narumiya S Sugimoto Y Ushikubi F Prostanoid receptors: structures, properties, and functions Physiol Rev 1999 79 1193 1226 10508233 Morrison KJ Gao Y Vanhoutte PM Epithelial modulation of airway smooth muscle American Journal of Physiology 1990 258 L254 262 2193539 Mathe AA Hedqvist P Effect of prostaglandins F2 alpha and E2 on airway conductance in healthy subjects and asthmatic patients Am Rev Respir Dis 1975 111 313 320 1091185 Cuthbert MF Effect on airways resistance of prostaglandin E1 given by aerosol to healthy and asthmatic volunteers Br Med J 1969 4 723 726 5354877 Martin JG Suzuki M Maghni K Pantano R Ramos-Barbon D Ihaku D Nantel F Denis D Hamid Q Powell WS The immunomodulatory actions of prostaglandin E2 on allergic airway responses in the rat J Immunol 2002 169 3963 3969 12244197 Tilley SL Hartney JM Erikson CJ Jania C Nguyen M Stock J McNeisch J Valancius C Panettieri RAJ Penn RB Koller BH Receptors and pathways mediating the effects of prostaglandin E2 on airway tone Am J Physiol Lung Cell Mol Physiol 2003 284 L599 606 12618422 Ushikubi F Hirata M Narumiya S Molecular biology of prostanoid receptors; an overview J Lipid Mediat Cell Signal 1995 12 343 359 8777578 10.1016/0929-7855(95)00022-I Katsuyama M Nishigaki N Sugimoto Y Morimoto K Negishi M Narumiya S Ichikawa A The mouse prostaglandin E receptor EP2 subtype: cloning, expression, and northern blot analysis FEBS Lett 1995 372 151 156 7556658 10.1016/0014-5793(95)00966-D Volovitz B Welliver RC De Castro G Krystofik DA Ogra PL The release of leukotrienes in the respiratory tract during infection with respiratory syncytial virus: role in obstructive airway disease Pediatr Res 1988 24 504 507 3174295 van Schaik SM Tristram DA Nagpal IS Hintz KM Welliver RC Welliver RC Increased production of IFN-gamma and cysteinyl leukotrienes in virus-induced wheezing J Allergy Clin Immunol 1999 103 630 636 10200012 10.1016/S0091-6749(99)70235-6 Sznajer Y Westcott JY Wenzel SE Mazer B Tucci M Toledano BJ Airway eicosanoids in acute severe respiratory syncytial virus bronchiolitis J Pediatr 2004 145 115 118 15238918 10.1016/j.jpeds.2004.03.049	16324223	PMC1325022	CC BY	2021-01-04 16:23:24	no	Respir Res. 2005 Dec 5; 6(1):142	utf-8	Respir Res	2,005	10.1186/1465-9921-6-142	oa_comm
==== Front BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1731633225910.1186/1471-2164-6-173Research ArticleThe odorant receptor repertoire of teleost fish Alioto Tyler S [email protected] John [email protected] Department of Molecular and Cell Biology, Functional Genomics Laboratory, Helen Wills Neuroscience Institute, University of California, Berkeley, California 94720, USA2 Grup de Recerca en Informàtica Biomèdica, Institut Municipal d'Investigació Mèdica, Universitat Pompeu Fabra, Centre de Regulació Genòmica, Psg. Marítim 37-49, 08003 Barcelona, Spain2005 6 12 2005 6 173 173 30 8 2005 6 12 2005 Copyright © 2005 Alioto and Ngai; licensee BioMed Central Ltd.2005Alioto and Ngai; 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 Vertebrate odorant receptors comprise three types of G protein-coupled receptors: the OR, V1R and V2R receptors. The OR superfamily contains over 1,000 genes in some mammalian species, representing the largest gene superfamily in the mammalian genome. Results To facilitate an informed analysis of OR gene phylogeny, we identified the complete set of 143 OR genes in the zebrafish genome, as well as the OR repertoires in two pufferfish species, fugu (44 genes) and tetraodon (42 genes). Although the genomes analyzed here contain fewer genes than in mammalian species, the teleost OR genes can be grouped into a larger number of major clades, representing greater overall OR diversity in the fish. Conclusion Based on the phylogeny of fish and mammalian repertoires, we propose a model for OR gene evolution in which different ancestral OR genes or gene families were selectively lost or expanded in different vertebrate lineages. In addition, our calculations of the ratios of non-synonymous to synonymous codon substitutions among more recently expanding OR subgroups in zebrafish implicate residues that may be involved in odorant binding. ==== Body Background The perception and discrimination of thousands of different odorants by the vertebrate olfactory system results from the activation of specific odorant receptors expressed by olfactory neurons in the nose. The first odorant receptors were identified in the rat [1] and belong to what is now referred to as the "OR" superfamily of odorant receptors [2]. ORs exhibit a predicted seven transmembrane topology and sequence motifs characteristic of the A family (rhodopsin-like or Class I) of G protein-coupled receptors. Subsequent to the initial discovery of the OR superfamily of odorant receptors, two unrelated types of G protein-coupled receptors (GPCRs) were identified in the mammalian vomeronasal organ, the V1R receptors [3] and the V2R receptors [4-7]. The vomeronasal V1R and V2R receptors are thought to subserve signaling to pheromonal compounds [2]. The OR gene superfamily is the largest multigene superfamily described in mammalian genomes. The completion of both the Celera and public consortium versions of the mouse genome confirmed the existence of about 1068 potential intact OR genes (comprising at least 228 subfamilies) and 334 pseudogenes [8,9]. In humans, there are ~340 intact OR genes and ~300 pseudogenes [10-12]. By way of contrast, molecular cloning and genomic DNA blot hybridizations in fish species suggest an OR repertoire size approximately five- to ten-fold smaller than that of mammalian species [13,14]. An understanding of how vertebrate olfactory receptor repertoires evolved can be gained from comparing the properties and organization of genes from divergent vertebrate species. In this regard, the zebrafish, Danio rerio, provides a useful model for comparative genomics studies. Recent studies have demonstrated that the zebrafish genome encodes only 1 V1R-like receptor [15] (T.S.A and J.N., unpublished results) and ~60 olfactory C-family (Class III) GPCRs related to the mammalian V2R family (T.S.A., P. Luu, E. VanName, and J.N., manuscript in preparation); one fish olfactory C-family receptor has been shown to be activated by amino acids [16,17], which are potent odorants for fish. In the OR superfamily, approximately 28 genes and 5 pseudogenes were identified previously in zebrafish using PCR and homology-based techniques [14,18-22]. Although a number of phylogenetic reconstructions have been made [8,9,18,23-28], a more accurate view of the OR superfamily's evolutionary history would be facilitated by comparisons between genomic datasets that include a more complete representation of member genes from each species (see also [29]). For the present study, we carried out genome data mining on the zebrafish genome sequence provided by the Sanger Institute Danio rerio Sequencing Project and found 143 potentially intact genes belonging to the zebrafish OR superfamily. We find that despite the limited size of the zebrafish OR repertoire, it comprises eight diverse OR families, with family members sharing on average ~40% amino acid identity. In addition, OR genes from two pufferfish species – fugu and tetraodon – can be grouped into six families which overlap with the zebrafish gene families. Analysis of the ratio of possible non-synonymous to synonymous codon substitutions suggests that OR genes in general are under negative or purifying selection; only a small number of residues within the transmembrane domains – the likely sites of odorant binding – appear to have undergone positive selection. Based on these findings we propose a model for the evolution of the vertebrate OR repertoire. Results and discussion Prediction of zebrafish OR genes The third (Zv3) and fourth (Zv4) draft zebrafish genome assemblies of whole genome shotgun sequence (5.7 × coverage) were searched for OR gene sequences using a modification of the method described for identifying OR sequences from the mouse genome [8]. The protein coding sequences of the vast majority of known OR genes characterized to date are uninterrupted by introns, which obviates the need for splice site prediction in the identification of most OR genes. Our gene prediction strategy was to combine a low-threshold BLAST search with profile Hidden Markov Model- (HMM) based gene prediction with the program Genewise . The Genewise results were post-processed using custom Perl scripts to generate complete ORFs. This process was repeated in an iterative fashion, as follows. The zebrafish genome assembly was subjected to TBLASTN search with a representative set of known zebrafish ORs (<50% percent identity among members of this set). The gene prediction program Genewise was then run on the genomic sequences surrounding each unique BLAST hit using a profile hidden Markov model (HMM) of the OR superfamily (see Methods for details). In each round, the newly predicted OR genes were added as queries for the next BLAST search. They were also aligned to previous members, and a new profile HMM was constructed for use in the next round of gene prediction. Initial query sequences included members of each of the OR9, OR2, OR5, OR4, and OR13 subfamilies described previously [18]. Predicted genes were subjected to a set of criteria for inclusion in the final set of OR genes. First, all coding sequences were required to be longer than 700 base pairs in length and show highest sequence similarity to previously characterized OR sequences. To be considered full-length, the deduced amino acid sequence was required to be greater than 275 amino acids, contain seven predicted transmembrane domains and exhibit the presence of a conserved N-linked glycosylation site with the pattern N-X- [TS]-X (where X is any amino acid residue except for proline) at the N-terminus. Sequences failing to meet these criteria and those lacking start or stop codons due to assembly gaps were considered to be partial genes. FASTY comparison to known OR peptide sequences was used to generate conceptual translations of potential pseudogenes and to determine the number of disruptions (frame shifts, early stop codons and in-frame deletions). To compensate for disruptions in OR coding sequences due to possible sequencing and/or assembly errors, we adopted a classification system previously proposed for mining of the mouse genome [8]. In the present study, a gene was considered intact if it had a complete coding sequence with up to one disruption, or a pseudogene if it was either a partial sequence with one or more disruptions or a full-length sequence with two or more disruptions. Based on these criteria, our search identified 143 intact OR genes (136 with no disruptions), 7 partial genes, 10 pseudogenes greater than 700 bp in length, and 15 gene fragments shorter than 700 bp (these shorter fragments were excluded from further analysis) [see Additional file 2 and Additional file 10]. Thus, we believe our total OR gene count is a conservative estimate of the true size of the OR repertoire, with between 78% (136/175) and 86% ([143+7]/175) of identifiable OR sequences consisting of potentially functional OR genes. How complete is the predicted OR gene repertoire? To address this question, we extracted 65 zebrafish OR sequences (28 non-redundant published OR genes or cDNAs, 4 unpublished full-length cDNAs, and 33 ESTs) from Genbank and determined whether or not they were represented in the set of predicted genes by aligning them to OR genomic sequence. The 33 ESTs were first consolidated into 23 clusters based on overlapping sequences. Eight of these corresponded to known genes, while 15 clusters were novel, bringing the non-redundant set of Genbank genes to 47. Forty-five out of 47 of these sequences were identified in our search [see Additional file 10]. The two that we were unable to identify, OR115-15 and OR128-14 (an unpublished cDNA that was identified in the more recent Zv5 genome assembly and an EST, respectively) belong to two of the largest subfamilies in the repertoire. Thus, we are confident that the repertoire of odorant receptors described here includes nearly all of the OR genes encoded in the zebrafish genome. However, because some gaps remain in the assembly near several small receptor gene clusters, we expect that as the genome sequence is refined, a few additional receptor genes will be found. OR nomenclature and classification OR families are typically defined as monophyletic groups with members that share greater than 40% amino acid identity, whereas subfamily members share greater than 60% amino acid identity [30]. Using these operational definitions, we classified the zebrafish OR genes into families and subfamilies by reconstructing their phylogeny by neighbor-joining with 1000 bootstrap replicates. Clades of OR genes with less than ~40% and less than ~60% inter-branch amino acid identity were used to group genes into distinct families and subfamilies, respectively. The average percent identity between families is approximately 25% while the maximum observed percent identity between any two ORs of different families is 39%. To unify the naming for zebrafish OR genes, we propose a revised nomenclature based on the following rationale. Both newly predicted and previously described OR genes were named (or re-named) according to subfamily membership. Subfamilies were numbered sequentially starting at the number 101 (to avoid confusion with previous zebrafish OR nomenclature) in a depth-first traversal of the phylogenetic tree. Within subfamilies, ORs were numbered sequentially according to genomic position, if known. The new nomenclature showing subfamily membership and correspondence to previously identified zebrafish OR genes are shown in Table S1 [see Additional file 10]. Genomic distribution of zebrafish OR genes Previous studies have demonstrated that OR genes are clustered in vertebrate genomes [8,9,18,31,32]. In mammalian genomes, OR genes are distributed widely, residing on 18 chromosomes in the mouse [8] and 21 chromosomes in humans [10,11]. From the zebrafish Zv3 and Zv4 assemblies, we found that 119 of the identified zebrafish OR genes are distributed in five major clusters containing between 14 and 31 genes each. There are two clusters on chromosome 15, two on chromosome 21, one on chromosome 10, several small clusters on chromosomes 8, 14 and 17, and in a few cases, genes exist as singletons (Figure 1) [see Additional file 10]. Subfamilies are largely contiguous (see below) and subfamily members usually share the same transcriptional orientation, suggesting tandem duplication as a mechanism of expansion within a subfamily [18,33]. We were able to assign genomic locations for ~80% of the OR genes we identified (29 remain unassigned). Figure 1 Chromosomal distribution of zebrafish OR genes. The majority of OR genes are organized in large clusters at only a few loci in the zebrafish genome. OR genes are depicted as boxes above (plus strand) or below (minus strand) a line representing each chromosome that encodes ORs. Genes are color-coded according to subfamily. Phylogeny of zebrafish OR genes Using a neighbor joining algorithm (see Methods), we constructed a phylogenetic tree of the 143 intact OR genes and 4 full-length pseudogenes identified in the two zebrafish genome assemblies, using the zebrafish melanocortin receptors as an outgroup (Figure 2a). Based on this analysis and the criteria set forth above, zebrafish ORs could be classified into 8 families (≥40% intra-family sequence identity) and 40 subfamilies (≥60% intra-subfamily sequence identity). An additional thirteen sequences comprising 7 partial genes and 6 pseudogene fragments were subsequently assigned to subfamilies based on their sequence similarities and additional phylogenetic analyses (data not shown). Most of the gene families contain between 12 and 40 genes each; the two smallest families, Family A and Family B, contain 6 and 1 genes each, respectively. The intra-subfamily identity threshold was lowered for three subfamilies, OR102, OR115 and OR125, to generate monophyletic clades [see Additional file 11]. High bootstrap support (Figure 2a) justify these classifications, with all subfamilies exhibiting bootstrap scores of 100%. Figure 2 Phylogeny of zebrafish and other vertebrate OR families. (a) Phylogeny of zebrafish receptors. A neighbor joining tree was constructed based on an alignment of the predicted amino acid sequences of 143 intact genes and 4 full-length pseudogenes identified from the zebrafish genome [see Additional file 2]. OR genes are named by subfamily and colored by family. The eight gene families are labeled A-H. The zebrafish melanocortin receptor branch (dotted line labeled "mcr") indicates the root of the tree. Bootstrap scores for each family are indicated in parentheses. (b) Phylogenetic relationship among zebrafish and mouse odorant receptors. The following sets of genes were aligned and used to construct a tree by neighbor joining: the mouse odorant receptors, mORs [8,9]; the subset of 136 intact zebrafish ORs with no disruptions identified in this study (highlighted in red); and mouse melanocortin receptors (mcr). Note the presence of OR gene subfamilies OR112, OR113, and OR114 (Family A) within the Class I clade and zebrafish OR101-1 (Family B) within the Class II clade. Bootstrap scores corresponding to selected nodes are indicated.(c) Phylogeny of the complete OR repertoires of zebrafish, fugu and tetraodon identified in this study. One-hundred-thirty-six non-disrupted genes from zebrafish, 42 non-disrupted genes from fugu, and 42 non-disrupted genes from tetraodon were used in this neighbor joining analysis. Families are labeled A-H and correspond to the zebrafish families shown in panel A. Bootstrap scores corresponding to selected nodes are indicated. The topology of the phylogenetic tree shown in Figure 2a is supported by three additional lines of evidence. First, we calculated all possible pairwise identities both within and between different groups of OR sequences. With the three exceptions noted above for subfamilies OR102, OR115, and OR125, the minimum percent identity within each subfamily is ≥ 62% [see Additional file 11]. Importantly, the maximum inter-subfamily identity is 44% [see Additional file 12] and the maximum interfamily identity is 38% [see Additional file 13]; both of these values are well below the ≥ 62% identity typically observed between members within a given subfamily. Thus, the sorting of ORs by neighbor joining analysis into distinct families and subfamilies is supported by an analysis based on all possible pairwise identities. Second, highly related OR genes are tightly clustered in the zebrafish genome, with the members of a given subfamily residing adjacent to one another, uninterrupted by more distantly related genes [18,33]. In the present analysis, we found that the assignment of OR subfamilies by neighbor joining analysis indeed is consistent with this genomic organization; out of 23 multigene subfamilies, members from only five (OR111, OR113, OR126, OR128, and OR133) are found in genomic clusters interrupted by genes from other subfamilies (Figure 1) [see Additional file 10]. Third, we constructed a phylogenetic tree using a maximum likelihood algorithm; at both the family and subfamily levels, maximum likelihood analysis yields tree topologies comparable to those derived by neighbor joining [see Additional file 5]. Comparison of fish and mammalian OR repertoires To gain additional insight into how the OR gene superfamily evolved in vertebrates, the zebrafish ORs were aligned to additional sets of vertebrate OR sequences. OR genes were predicted from the genome sequences of two pufferfish species, fugu (Takifugu rubripes) [27] and tetraodon (Tetraodon nigroviridis) [34], using methods identical to those used for finding zebrafish ORs. Forty-four (3 with one disruption) and 42 (6 with one disruption) intact genes were found in fugu [27,29] and tetraodon, respectively [see Additional file 3 and Additional file 4]. As the genome sequence data for these two species represent ~95% and ~92% coverage, respectively, we expect that the OR genes identified here comprise the majority of each species' OR repertoire. Thus, the OR repertoires of both pufferfish species appear to be only ~one-third the size of the zebrafish repertoire. A summary of the OR genes identified in zebrafish, fugu, and tetraodon genomes is provided in Table 2. Table 1 Average pairwise identities between odorant receptor families. Pairwise comparisons were performed between each member of a family with each member of the family to be compared. The average percent identity was then calculated for all comparisons between each pair of families. Similar results were obtained for consensus sequences of each family generated from Hidden Markov Models (HMMs) of each family. Zebrafish families are labeled A-H according to Figure 2a. The two major groups of mouse genes are denoted as "Class I" and "Class II." The maximum percent identity between families used to guide family classification of phylogenetic clades was 40% identity. A 100 B 31 100 C 28 34 100 D 29 31 29 100 E 28 28 26 26 100 F 29 29 27 28 32 100 G 25 30 28 26 25 26 100 H 18 18 17 18 19 19 18 100 Class I 32 32 29 28 27 28 24 17 100 Class II 28 38 28 29 28 29 25 18 30 100 A B C D E F G H Class I Class II Table 2 Summary of identified teleost OR genes. OR sequences identified in the present study are listed in this table. Zebrafish Fugu Tetraodon Intact genes a 143 (7) 44 (3) 42 (6) Partial genes b 7 9 4 Pseudogenes c 10 4 11 Total 160 57 57 a A gene is operationally defined as "intact" if it possesses a full-length OR protein coding sequence with no more than one disruption. For each species, the number of genes with a single disruption is listed in parentheses; the remainder contain no disruptions. b A sequence encoding ≤ 275 contiguous amino acids, missing specific features characteristic of ORs (see the text for details), or missing start or stop codons is classified as a partial gene. c A sequence is defined as a pseudogene if it is a partial gene with one or more disruption or a full-length gene with 2 or more disruptions. Nine-hundred-thirty-five mouse OR sequences [8,9] were either downloaded from Genbank (864 genes) or extracted from MGSCv3 using published coordinates (71 genes) [9]. Phylogenetic trees were computed for ORs from zebrafish and mouse (Figure 2b) [see Additional file 6] and zebrafish, fugu and tetraodon (Figure 2c) [see Additional file 7]. The location of the melanocortin receptor branch represents the root of each tree. Mouse ORs can be classified into two groups, Class I and Class II, each showing on average greater than 40% intra-group sequence identity [8]. Based on their greater similarity to the limited number of fish OR genes identified prior to the present study, Class I genes from amphibians and mammals have been referred to as "fish-like" [8,10,24]. However, our analysis of the complete set of zebrafish OR genes indicates that this view cannot be generalized to the entire fish OR repertoire. Mammalian Class I and Class II genes can in fact be grouped more closely with only two out of eight ~equidistantly-related zebrafish families; Class I genes show close similarity to only a small subset of zebrafish OR genes (OR112-1, OR113-1, OR113-2 and OR114-1, which together comprise Family A), and one zebrafish gene (OR101-1, comprising the single member Family B) clusters together with mammalian Class II genes (Figure 2b). We base these conclusions on phylogenetic reconstructions as determined by neighbor joining (Figure 2b) and maximum likelihood [see Additional file 6], as well as on a separate calculation of average pairwise identities of genes between families (Table 1) [see Additional file 13]. In all cases, the alignment of mouse and zebrafish genes was gap-minimized and trimmed to remove N- and C-terminal tails [see Additional file 2]. Overall, mouse Class I exhibits similar average pairwise identity to the zebrafish families (27.3 ± 4.8% identity [mean ± standard deviation]; range: 17 – 32%) as mouse Class II (27.7 ± 5.5%; range: 18 – 38%); the difference in mean values is not significant in a two-tailed t-test (p = 0.89). Calculations comparing consensus sequences representing each family yielded similar results (data not shown). A comparison of teleost OR genes further reveals that six of the eight zebrafish OR families overlap with pufferfish families; families B and G do not appear to be present in pufferfish (Figure 2c) [see Additional file 7]. In our phylogeny comparing zebrafish and pufferfish genes by neighbor joining, we find low bootstrap support for Family E (score = 47), likely reflecting this family's closer proximity to Family F in the multi-species tree as compared to the tree generated with zebrafish OR genes alone (however, Family E has high bootstrap support by maximum likelihood analysis [see Additional file 7]). Interestingly, in zebrafish the most divergent family (Family H) shows only 17–19% identity to other families (versus 25–34% interfamily identity amongst the other families; Table 1). The location of the outgroup melanocortin receptor between Family H and the other families supports the conclusion that this family is the result of a very ancient gene duplication event. Based on the degree of divergence from other OR gene families, it is possible that the genes comprising Family H may not in fact encode bona fide odorant receptors. However, the predicted zebrafish Family H receptors retain one of the highly conserved OR signature motifs (see below), and one member of this family (OR137-7) was previously identified as an EST from a zebrafish olfactory epithelium cDNA library [see Additional file 10]. In addition, when zebrafish Family H sequences were used in BLAST searches of both the non-redundant protein sequence database and the mouse genome sequence, previously identified OR sequences were identified as the closest hits (data not shown). Family H also forms a cluster distinct from non-OR GPCRs in a phylogenetic tree comprising mouse and zebrafish ORs together with a set of 199 non-OR Type I (rhodopsin class) mouse GPCRs [see Additional file 8]. Thus, for the present purposes we consider the Family H sequences operationally as OR genes. More generally, this phylogenetic reconstruction based on OR and non-OR GPCRs reveals that the ORs as a group are distinct from the other Type I GPCRS. A similar phylogeny for vertebrate OR genes was recently described [29]. This study placed zebrafish, fugu, Xenopus and chicken OR genes into groups roughly comparable to those described here in Figure 2, with Family A corresponding to these authors' Group β (which clusters closely with human Class I genes); the single zebrafish gene comprising Family B falling within Group γ /human Class II; Family C corresponding to Group ε; Families D and G corresponding to Group ζ (which is not a monophyletic clade); Families E and F corresponding to Group δ; and Family H corresponding to Group η. Two highly divergent groups (not identified or retained in our search) – termed κ and θ – were also described, although their identities as OR genes are unclear [29]. Conserved motifs in predicted OR protein sequences Previous studies of vertebrate ORs have identified a number of conserved sequence motifs characteristic of these receptors [8,12,35]. These include the following: an N-linked glycosylation site NX [TS]X in the N-terminal domain; the motif MA [FY] [DE]RYVAIC located at the third transmembrane domain (TM3)/second intracellular loop (IC2) junction which is thought to interact with G-proteins (specifically Golf); three conserved cysteine residues in the second extracellular loop (EC2) thought to partake in disulfide bonding; and the motif KAFSTCXSH in IC3 containing an intracellular cysteine conserved in GPCRs and potential phosphorylation sites. We found that these motifs are conserved in all the zebrafish OR families, with the exception of Family H, in which only the MAYDRYVAIC motif is conserved. This sequence conservation is illustrated by a sequence logo generated from the alignment of predicted full-length zebrafish OR coding sequences (Figure 3a). In this representation, the relative frequency with which an amino acid appears at a given position is reflected by the height of its one-letter amino acid code in the logo, with the total height at a given position proportional to the level of sequence conservation. Interestingly, when compared to the sequence logo representing the alignment of mouse Class I and Class II ORs (Figure 3b), the zebrafish OR logo shows lower conservation amongst the predicted zebrafish receptor sequences (reflected by the more numerous and shorter letters at individual positions in the logo), revealing the greater diversity within the zebrafish vs. mouse OR superfamily (Table 1). Figure 3 Sequence logos of zebrafish and mouse OR families. Conservation of predicted amino acid sequence for the zebrafish (a) and mouse (b) OR repertoires is shown graphically (see the text). Y axis, information content. X axis, residue position. For this analysis, positions with gaps in more than 95% of sequences, as well as poorly aligned N- and C-terminal sequences, were removed. Positions in the species-specific logos are identical according to this alignment. The logo was generated from this alignment using the program WebLogo (G.E. Crooks, G. Hon, J.-M. Chandonia and S.E. Brenner, personal communication), available at . Adaptive evolution of OR genes What evolutionary processes might explain the diversity of OR gene sequences? This diversity could be the result of genetic drift, with polymorphisms in the population being fixed at a rate consistent with the absence of selective pressure. Alternatively, the sequence diversity could reflect true functional divergence, perhaps in the ligand specificities of the encoded receptors. Considering the diversity of OR proteins encoded in vertebrate genomes and the even greater diversity of compounds detected by these receptors, ligand binding sites within ORs may be expected to be under positive selective pressure as organisms evolve new receptor proteins to recognize odorant compounds. As for other Type 1 (rhodopsin class) GPCR ligands, odorants are thought to bind to OR proteins in the plane of the membrane, in contact with residues in the transmembrane domains [2]. Consistent with this notion, previous studies have demonstrated that odorant receptor genes have been subject to positive selective pressure, especially in the transmembrane domains thought to coordinate odorant binding [13,36] (however, see [37]). We therefore attempted to pinpoint the precise codon sites – and thus the amino acid residues – that may have been subjected to positive selection during the evolution of the zebrafish OR superfamily. To this end, we used the relative frequency of non-synonymous vs. synonymous codon substitutions to assess the selective processes acting on these receptor genes [38]. Where there is no positive or negative selection on a sequence, the number of non-synonymous changes relative to the number of possible non-synonymous changes (dN) would be equal to the number of synonymous changes relative to the number of possible synonymous changes (dS) – i.e., dN/dS = 1. Significant deviations of dN/dS from unity reflect selection on the sequence; a dN/dS ratio > 1 indicates that a region has undergone positive selection, whereas a dN/dS ratio < 1 indicates negative or "purifying" selection [38]. For our analysis, we aligned 136 full-length zebrafish OR coding sequences containing no disruptions and calculated dN/dS ratios based on a gap-minimized alignment (Table 3 and Figure 4a). We found that when the OR coding sequence is partitioned broadly into transmembrane domains (TMs) 1–7 and non-transmembrane domains (excluding the N- and C-terminal tails), none of these regions exhibits positive selection. Rather, with average dN/dS ratios <1, these protein regions all appear to be under negative or purifying selection. Interestingly, TMs 1, 3, 4, 5, and 6 display significantly higher average dN/dS ratios than the combined intracellular and extracellular loops (p < 1 × 103; see Figure 4a). The observation that these transmembrane domains were in general under less purifying selection than other regions of the protein is consistent with the possibility that they may have adapted to bind different odorants. In contrast, TM2 and TM7 display significantly lower average dN/dS ratios compared to the complete coding sequence (p < 0.05). The apparently stronger negative selection on TM2 and TM7 (as compared to the other transmembrane regions) suggests that these transmembrane domains subserve a common – perhaps structural – role in these receptors. Table 3 Comparison of selective pressure by transmembrane domain. TM1 TM2 TM3 TM4 TM5 TM6 TM7 non-TM CDS Length (amino acids) 26 21 19 24 25 19 24 110 268 Mean pairwise dN/dSa 0.330 0.276 0.457 0.574 0.437 0.513 0.282 0.304 0.325 TM vs loops p-value (2 -tailed t test)b 9.9E-04 0.81 2.1E-12 7.6E-21 1.2E-18 2.5E-10 1.00 TM vs CDS p-value (2 -tailed t test)b 0.45 0.025 2.1E-08 5.9E-16 7.9E-12 6.0E-07 0.0024 Number of positively selected sites (p < 0.1)c 0 0 1 1 0 0 0 2 4 Number of negatively selected sites (p < 0.1)c 21 16 12 15 15 14 19 80 192 a The mean ratios of non-synonymous substitutions per non-synonymous site (dN) to synonymous substitutions per synonymous site (dS) were calculated for each transmembrane region, the aggregate of the non-transmembrane regions, as well as the entire OR coding sequence. dN/dS ratios were calculated only for pairwise comparisons in which mutational saturation had not been reached. A constant set of 262 informative pairs was used for all calculations. b Probabilities that the average dN/dS ratio for the particular TM domain is no different than the average for the entire CDS or the non-TM domains, using a two-tailed t test. c Number of sites with probabilities greater than 0.9 (p < 0.1) of being positively or negatively selected for each domain. Ancestral sequences were reconstructed, and at every non-constant site, dN and dS were calculated. If dN < (or >) dS, a p-value derived from a two-tailed binomial distribution was used to assess significance. Figure 4 Sites under positive and negative selection in OR coding sequences. Nucleotide alignments were generated from the corresponding amino acid alignment [see Additional file 2] after removal of N- and C-terminal sequences and gap removal with respect to OR124-3, and subjected to two analyses of selective pressure. (a) Analysis of dN/dS ratios by sub-regions of OR coding sequences indicates that OR genes in general are under negative or purifying selection (dN/dS < 1). However, analysis of informative pairwise comparisons reveals that transmembrane domains (TMs) 3–6, and to a lesser degree TM1, have significantly higher average pairwise dN/dS ratios when compared with the average for non-transmembrane coding sequence (asterisks; p < 1 × 10-3). Error bars show standard errors of the means. (b) A schematic representation of OR124-3 (an example OR) with transmembrane domains one through seven shown from left to right. SLAC analysis reveals sites under positive selection (dN/dS > 1) with p < 0.1 (red), p < 0.2 (orange), p < 0.5 (yellow), and sites under negative selection (dN/dS < 1) with p < 0.05 (dark blue), p < 0.1 (light blue), p < 0.2 (turquoise), p < 0.5 (green). The null hypothesis is that a site is neutrally evolving with dN/dS = 1. Yellow and green residues may be considered neutrally evolving. Sites on the representative OR124-3 sequence with dN/dS > 1 and p < 0.1 are: V108 (dN/dS = 1.39), F145 (dN/dS = 1.52), E261 (dN/dS = 1.48) and T262 (dN/dS = 1.42). Snake plot generated using the RbDe web service [48]. We hypothesized that specific codon sites corresponding to odorant-binding residues might have been positively selected as coding sequences diverged after gene duplication events. Accordingly, identification of these sites would suggest possible ligand binding sites. We therefore performed a site-by-site analysis of dN/dS ratios based on the alignment of the set of 136 full-length intact coding sequences used above. To avoid a potentially high rate of false positives common with pooled site methods (see [39]), we used the more conservative likelihood individual site (IS) method [40] based on the original proposed IS approach [41]. The phylogenetic relationships between sequences were determined and a substitution model was estimated from the data. The ancestral codon sequences at each node were then reconstructed and the dN and dS values were calculated for each codon site. In Figure 4b, the probability of being under positive or negative selection (dN/dS values different than dN/dS = 1.0) for each codon site is indicated on a snake plot of a representative OR amino acid sequence, OR124-3 (see also Table 3). By these criteria, only two sites within the transmembrane domains (one in TM3 and one in TM4) appear to have been subjected to positive selection, consistent with the notion that they may play a role in contacting ligands. Interestingly, two adjacent sites in the short third extracellular loop (very close to the top of TM6) also exhibit dN/dS ratios > 1. Overall, our characterization of dN/dS ratios reveals a striking paucity of sites exhibiting signs of positive selection, possibly reflecting the dominating influence of negative selection throughout the receptor coding region. Alternatively, since non-synonymous substitutions may occur only sporadically over evolutionary time, the signatures of less recent substitutions may no longer be detected by this analysis of the entire zebrafish OR family. Evolution of the vertebrate OR gene repertoire The characterization of the complete OR repertoires from both fish and mammalian species allows an informed analysis of OR gene evolution in the vertebrate lineage. One noteworthy feature of our phylogenetic reconstruction is the presence of a group of zebrafish and pufferfish subfamilies which together form a putative OR family (Family H) more divergent than the other families are to each other (Figure 2c and Table 1); based on a BLAST search of the mouse genome using representative zebrafish Family H sequences, this family is absent from the mouse. We hypothesize that the node between this branch of the tree and the others is the root representing the most ancient gene duplication event observable in the teleost lineage. This is supported by the placement at this node of the melanocortin receptor (outgroup) branch. In addition, when we aligned five OR sequences from lamprey [25,42] to the teleost ORs, they formed two additional families on either side of the melanocortin receptor branch, one which is clearly an OR family (more similar to teleost OR families A-G than to H) and one which appears as an outgroup (equidistant from all teleost families A-H and more dubious as an OR family) [see Additional file 9]. Since the lamprey diverged before the teleost/tetrapodon split, these observations provide further support for this node as the root of the tree. It should be noted, however, that until the lamprey genome has been fully sequenced, we will only have a partial picture of the ancestral OR repertoire. We expect that characterization of the entire lamprey OR repertoire will shed light on more ancient evolutionary events. From our analysis of mammalian, teleost and lamprey OR sequences, we propose the following model for OR gene evolution in vertebrates. OR genes in present-day vertebrates likely descended from eight ancestral OR genes (or gene families) that existed at the time of the split between ray-finned and lobe-finned fish (the ancestors of teleosts and tetrapods, respectively) approximately 450 million years ago (mya) [43]. A phylogenetic reconstruction based on mouse and zebrafish ORs and 199 mouse non-OR GPCRs [see Additional file 8] indicates that the ORs form a group distinct from all other Type I GPCRs, possibly reflecting a very ancient duplication event(s) and/or rapid divergence of the ORs in the evolution of Type I GPCRs. Our estimate of ancestral OR gene number is based on the identification of 8 OR gene families in teleosts, two of which show somewhat higher similarity to the 2 OR gene families in mammals. The grouping of zebrafish and pufferfish OR genes into common families indicates that the gene duplication events that gave rise to the major OR families probably occurred prior to the speciation of teleosts. In addition, the greater similarities between zebrafish Family A and mouse Class I, and between zebrafish Family B and mouse Class II infer that the ancestral genes for these families existed before the tetrapodon/teleost split. Our model therefore suggests a history during which ancestral genes or gene families were selectively lost during the evolution of the different vertebrate lineages. Of the ancestral families, zebrafish retained 8 families, fugu and tetraodon retained 6 families, and mammals retained 2 families. It should be noted that the low bootstrap score (47) for Family E in the comparison of zebrafish and pufferfish OR genes (Figure 2c) raises the possibility that Families E and F (which are adjacent to each other in the teleost phylogenetic tree) may have arisen from a more recent duplication in the teleost lineage. Alternatively, the genes in these groups may have been subjected to gene conversion events, with the effect of homogenizing the sequences between these two families. It is also possible that the 4–6 gene families unique to teleosts descended from Family A/Class I and/or Family B/Class II ancestral genes, after the tetrapodon/teleost split. Such a scenario seems unlikely, however, considering the roughly equivalent degree of divergence exhibited between 7 out of the 8 teleost gene families (including Families A and B). Moreover, amphibian and avian OR genes can be grouped into 6 out of the 8 identified OR families (Families A, B/Class II, C, E, F and H), further implicating the presence of common ancestral genes for these families prior to the tetrapodon/teleost split [29]. Mechanisms of gene or family loss in a particular vertebrate lineage may have involved a number of processes, for example, gene conversion, pseudogenization of all genes in a family, unequal crossover recombination events during meiosis, or larger chromosomal rearrangements. From the available data we cannot infer the precise order and rate of OR gene family expansion and contraction, or speciation events. Nonetheless, six of the retained OR gene families were subject to a substantial net expansion and diversification in zebrafish (and to a lesser degree in the pufferfish), while the other two ancestors gave rise to the present-day mammalian Class I and Class II ORs as well as a small number of zebrafish genes. We hypothesize that relaxed selective pressure on a subset of the ancestral tetrapodon OR repertoire led to the loss of major OR gene families in the mammalian lineage. The expansion within the two remaining gene families was likely driven by the adaptation to the terrestrial odorous environment. Thus, different selective pressures found in the aquatic and terrestrial environments led to different sizes and shapes of the OR repertoires of fish and mammals. It is generally thought that the diversity of OR sequences – as represented in the number of receptor families – underlies the diversity of chemical structures or "odor space" that can be detected by an organism's olfactory system. Thus, with ~6–8 OR gene families retained over evolutionary time (vs. 2 in mammals), fish may be capable of detecting a larger diversity of chemical structures than mammals. However, the larger total number of OR sequences in mammals (~1,000 vs. ~100 in fish) presumably allows a finer discrimination amongst the compounds that are detected by the mammalian olfactory system. Methods Iterative data mining Genome-wide searches of the third (Zv3) and fourth (Zv4) draft zebrafish genome assemblies made available by the Sanger Center on Nov 27, 2003, and July 12, 2004, respectively, were performed several times using the predicted ORs from each previous round to increase our querying power. This iterative data-mining approach has been published for finding OR genes in the mouse genome [8]. A detailed description of our protocol is provided in the Supplement [see Additional file 1]. Alignment and tree construction For multiple alignments of OR genes, ClustalX 1.81 [44] was used with default parameters and gaps were inspected manually and edited in xced to ensure integrity of transmembrane domains and proper alignment of anchoring OR motifs. N- and C-terminal tails were trimmed for all alignments. The neighbor-joining algorithm as implemented by PFAAT was used to generate unrooted phylogenetic trees from these alignments using the BLOSUM 50 similarity matrix; positions with greater than 40% gaps were excluded. One thousand bootstraps were performed to assess the support at each tree node. Trees were visualized with unrooted [45]. Maximum likelihood analysis was carried out using PHYML [46] on the same processed amino acid alignments described above. Bootstrap analysis with 100 replicates was carried out using the JTT model of amino acid substitution. The consensus tree including bootstrap support for each node was plotted for each dataset using either ATV [47] or unrooted. The sequences used for comparison to the zebrafish OR genes were obtained from Genbank and included the set of intact MORs [Genbank: AY072961] – [Genbank: AY074256] [8] plus 71 newly identified OR genes extracted from MGSCv3 using coordinates from the online supplement to [9], 5 full-length lamprey OR receptors [Genbank: AAC82383, Genbank: AAC82384, Genbank: AAC82385, Genbank: CAA10135, Genbank: CAA10136]. [25], the zebrafish melanocortin 1, 2, 3, 4, 5a and 5b receptors [Genbank: NP_851301.1, Genbank: NP_851302.1, Genbank: NP_851303.1, Genbank: NP_775385.1, Genbank: NP_775386.1, Genbank: NP_775387.1], and 199 mouse non-OR Class A GPCRs extracted from the GPCRDB . Fugu and tetraodon OR sequences were predicted from the current genome assemblies [27,34] using the methods described above for zebrafish. For the calculation of percent identities, mouse and zebrafish amino acid sequences were multiply aligned and trimmed of their N- and C-terminal tails as described above. Calculations of average, minimum and maximum intra-family, inter-family, intra-subfamily and inter-subfamily percent identities were based on percent identities calculated for all pairs of amino acid sequences in this multiple alignment. dN/dS analysis The dN/dS ratios for multi-codon regions (i.e. individual transmembrane domains or loop regions) of the odorant receptor coding sequence were determined using previously published methods [38]. To make inferences about selective pressure (positive and negative selection) on individual codons (sites) within the coding sequence of the zebrafish OR genes, the Single Likelihood Ancestor Counting (SLAC) package , which implements the Suzuki-Gojobori method [41], was used. Details regarding both of these methods are provided in the Supplement [see Additional file 1]. Genbank accession numbers All sequences described in this study have been deposited in Genbank under accession numbers [Genbank: DQ305986] – [Genbank: DQ306145] (zebrafish), [Genbank:DQ306146] – [Genbank: DQ306202] (tetraodon), and [Genbank:DQ306203] – [Genbank: DQ306259] (fugu). Authors' contributions TSA carried out the analysis. Both authors participated in the design of the study and writing of the manuscript. Supplementary Material Additional File 1 Supplement: Methods and legends for Figures S1-S8 and Tables S1-S4 Click here for file Additional File 2 Figure S1. Multiple sequence alignment of zebrafish OR amino acid translations. Click here for file Additional File 3 Figure S2. Multiple sequence alignment of fugu OR amino acid translations. Click here for file Additional File 4 Figure S3. Multiple sequence alignment of tetraodon OR amino acid translations. Click here for file Additional File 5 Figure S4. Phylogeny of zebrafish ORs using maximum likelihood analysis. Click here for file Additional File 6 Figure S5. Phylogeny of zebrafish and mouse ORs using maximum likelihood analysis. Click here for file Additional File 7 Figure S6. Phylogeny of zebrafish, fugu and tetraodon ORs using maximum likelihood analysis. Click here for file Additional File 8 Figure S7. Phylogeny of zebrafish and mouse ORs rooted by mouse non-OR GPCRs. Click here for file Additional File 9 Figure S8. Phylogeny of zebrafish, fugu, tetraodon and lamprey ORs. Click here for file Additional File 10 Table S1. The zebrafish OR repertoire. Click here for file Additional File 11 Table S2. Pairwise intra-subfamily percent identities for zebrafish OR subfamilies. Click here for file Additional File 12 Table S3. Pairwise inter-subfamily percent identities for zebrafish OR subfamilies. Click here for file Additional File 13 Table S4. Pairwise inter-group percent identities for zebrafish OR families and mouse Class I and Class II ORs. Click here for file Acknowledgements This work was supported by a grant from the National Institute on Deafness and Other Communications Disorders, National Institutes of Health (J.N.) and a genomics training grant from the National Institutes of Health (T.S.A.). We thank member of our lab for helpful discussions and K. Scott for comments on the manuscript. ==== Refs Buck L Axel R A novel multigene family may encode odorant receptors: a molecular basis for odor recognition Cell 1991 65 175 187 1840504 10.1016/0092-8674(91)90418-X Mombaerts P Genes and ligands for odorant, vomeronasal and taste receptors Nat Rev Neurosci 2004 5 263 278 15034552 10.1038/nrn1365 Dulac C Axel R A novel family of genes encoding putative pheromone receptors in mammals Cell 1995 83 195 206 7585937 10.1016/0092-8674(95)90161-2 Matsunami H Buck LB A multigene family encoding a diverse array of putative pheromone receptors in mammals Cell 1997 90 775 784 9288756 10.1016/S0092-8674(00)80537-1 Ryba NJ Tirindelli R A new multigene family of putative pheromone receptors Neuron 1997 19 371 379 9292726 10.1016/S0896-6273(00)80946-0 Herrada G Dulac C A novel family of putative pheromone receptors in mammals with a topographically organized and sexually dimorphic distribution Cell 1997 90 763 773 9288755 10.1016/S0092-8674(00)80536-X Yang H Shi P Zhang YP Zhang J Composition and evolution of the V2r vomeronasal receptor gene repertoire in mice and rats Genomics 2005 86 306 315 16024217 10.1016/j.ygeno.2005.05.012 Zhang X Firestein S The olfactory receptor gene superfamily of the mouse Nat Neurosci 2002 5 124 133 11802173 Zhang X Rodriguez I Mombaerts P Firestein S Odorant and vomeronasal receptor genes in two mouse genome assemblies Genomics 2004 83 802 811 15081110 10.1016/j.ygeno.2003.10.009 Glusman G Yanai I Rubin I Lancet D The complete human olfactory subgenome Genome Res 2001 11 685 702 11337468 10.1101/gr.171001 Malnic B Godfrey PA Buck LB The human olfactory receptor gene family Proc Natl Acad Sci U S A 2004 101 2584 2589 14983052 10.1073/pnas.0307882100 Zozulya S Echeverri F Nguyen T The human olfactory receptor repertoire Genome Biol 2001 2 RESEARCH0018 11423007 10.1186/gb-2001-2-6-research0018 Ngai J Dowling MM Buck L Axel R Chess A The family of genes encoding odorant receptors in the channel catfish Cell 1993 72 657 666 7916654 10.1016/0092-8674(93)90395-7 Barth AL Dugas JC Ngai J Noncoordinate expression of odorant receptor genes tightly linked in the zebrafish genome Neuron 1997 19 359 369 9292725 10.1016/S0896-6273(00)80945-9 Pfister P Rodriguez I Olfactory expression of a single and highly variable V1r pheromone receptor-like gene in fish species Proc Natl Acad Sci U S A 2005 102 5489 5494 15809442 10.1073/pnas.0402581102 Luu P Acher F Bertrand HO Fan J Ngai J Molecular determinants of ligand selectivity in a vertebrate odorant receptor J Neurosci 2004 24 10128 10137 15537883 10.1523/JNEUROSCI.3117-04.2004 Speca DJ Lin DM Sorensen PW Isacoff EY Ngai J Dittman AH Functional identification of a goldfish odorant receptor Neuron 1999 23 487 498 10433261 10.1016/S0896-6273(00)80802-8 Dugas JC Ngai J Analysis and characterization of an odorant receptor gene cluster in the zebrafish genome Genomics 2001 71 53 65 11161797 10.1006/geno.2000.6415 Weth F Nadler W Korsching S Nested expression domains for odorant receptors in zebrafish olfactory epithelium Proc Natl Acad Sci USA 1996 93 13,321 13,326 8552589 10.1073/pnas.93.23.13321 Vogt RG Lindsay SM Byrd CA Sun M Spatial patterns of olfactory neurons expressing specific odor receptor genes in 48-hour-old embryos of zebrafish Danio rerio J Exp Biol 1997 200 ( Pt 3) 433 443 9057306 Barth AL Justice NJ Ngai J Asynchronous onset of odorant receptor expression in the developing zebrafish olfactory system Neuron 1996 16 23 34 8562087 10.1016/S0896-6273(00)80020-3 Byrd CA Jones JT Quattro JM Rogers ME Brunjes PC Vogt RG Ontogeny of odorant receptor gene expression in zebrafish, Danio rerio J Neurobiol 1996 29 445 458 8656210 10.1002/(SICI)1097-4695(199604)29:4<445::AID-NEU3>3.0.CO;2-8 Irie-Kushiyama S Asano-Miyoshi M Suda T Abe K Emori Y Identification of 24 genes and two pseudogenes coding for olfactory receptors in Japanese loach, classified into four subfamilies: a putative evolutionary process for fish olfactory receptor genes by comprehensive phylogenetic analysis Gene 2004 325 123 135 14697517 10.1016/j.gene.2003.10.011 Freitag J Krieger J Strotmann J Breer H Two classes of olfactory receptors in Xenopus laevis Neuron 1995 15 1383 1392 8845161 10.1016/0896-6273(95)90016-0 Freitag J Beck A Ludwig G von Buchholtz L Breer H On the origin of the olfactory receptor family: receptor genes of the jawless fish (Lampetra fluviatilis) Gene 1999 226 165 174 10084886 10.1016/S0378-1119(98)00575-7 Freitag J Ludwig G Andreini I Rossler P Breer H Olfactory receptors in aquatic and terrestrial vertebrates J Comp Physiol [A] 1998 183 635 650 9839455 10.1007/s003590050287 Aparicio S Chapman J Stupka E Putnam N Chia JM Dehal P Christoffels A Rash S Hoon S Smit A Gelpke MD Roach J Oh T Ho IY Wong M Detter C Verhoef F Predki P Tay A Lucas S Richardson P Smith SF Clark MS Edwards YJ Doggett N Zharkikh A Tavtigian SV Pruss D Barnstead M Evans C Baden H Powell J Glusman G Rowen L Hood L Tan YH Elgar G Hawkins T Venkatesh B Rokhsar D Brenner S Whole-genome shotgun assembly and analysis of the genome of Fugu rubripes Science 2002 297 1301 1310 12142439 10.1126/science.1072104 Glusman G Sosinsky A Ben-Asher E Avidan N Sonkin D Bahar A Rosenthal A Clifton S Roe B Ferraz C Demaille J Lancet D Sequence, structure, and evolution of a complete human olfactory receptor gene cluster Genomics 2000 63 227 245 10673334 10.1006/geno.1999.6030 Niimura Y Nei M Evolutionary dynamics of olfactory receptor genes in fishes and tetrapods Proc Natl Acad Sci U S A 2005 102 6039 6044 15824306 10.1073/pnas.0501922102 Lancet D Ben-Arie N Olfactory Receptors Current Biology 1993 3 668 674 15335857 10.1016/0960-9822(93)90064-U Ben-Arie N Lancet D Taylor C Khen M Walker N Ledbetter DH Carrozzo R Patel K Sheer D Lehrach H North MA Olfactory receptor gene cluster on human chromosome 17: possible duplication of an ancestral receptor repertoire Hum Mol Gen 1994 3 229 235 8004088 Young JM Friedman C Williams EM Ross JA Tonnes-Priddy L Trask BJ Different evolutionary processes shaped the mouse and human olfactory receptor gene families Hum Mol Genet 2002 11 535 546 11875048 10.1093/hmg/11.5.535 Kratz E Dugas JC Ngai J Odorant receptor gene regulation: implications from genomic organization Trends Genet 2002 18 29 34 11750698 10.1016/S0168-9525(01)02579-3 Jaillon O Aury JM Brunet F Petit JL Stange-Thomann N Mauceli E Bouneau L Fischer C Ozouf-Costaz C Bernot A Nicaud S Jaffe D Fisher S Lutfalla G Dossat C Segurens B Dasilva C Salanoubat M Levy M Boudet N Castellano S Anthouard V Jubin C Castelli V Katinka M Vacherie B Biemont C Skalli Z Cattolico L Poulain J De Berardinis V Cruaud C Duprat S Brottier P Coutanceau JP Gouzy J Parra G Lardier G Chapple C McKernan KJ McEwan P Bosak S Kellis M Volff JN Guigo R Zody MC Mesirov J Lindblad-Toh K Birren B Nusbaum C Kahn D Robinson-Rechavi M Laudet V Schachter V Quetier F Saurin W Scarpelli C Wincker P Lander ES Weissenbach J Roest Crollius H Genome duplication in the teleost fish Tetraodon nigroviridis reveals the early vertebrate proto-karyotype Nature 2004 431 946 957 15496914 10.1038/nature03025 Pilpel Y Lancet D The variable and conserved interfaces of modeled olfactory receptor proteins Protein Sci 1999 8 969 977 10338007 Gilad Y Bustamante CD Lancet D Paabo S Natural selection on the olfactory receptor gene family in humans and chimpanzees Am J Hum Genet 2003 73 489 501 12908129 10.1086/378132 Gimelbrant AA Skaletsky H Chess A Selective pressures on the olfactory receptor repertoire since the human-chimpanzee divergence Proc Natl Acad Sci U S A 2004 101 9019 9022 15178761 10.1073/pnas.0401566101 Nei M Gojobori T Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions Mol Biol Evol 1986 3 418 426 3444411 Nei M Selectionism and Neutralism in Molecular Evolution Mol Biol Evol 2005 Aug 24; [Epub ahead of print] Pond SL Frost SD A simple hierarchical approach to modeling distributions of substitution rates Mol Biol Evol 2005 22 223 234 15483327 Suzuki Y Gojobori T A method for detecting positive selection at single amino acid sites Mol Biol Evol 1999 16 1315 1328 10563013 Berghard A Dryer L A novel family of ancient vertebrate odorant receptors J Neurobiol 1998 37 383 392 9828044 10.1002/(SICI)1097-4695(19981115)37:3<383::AID-NEU4>3.0.CO;2-D Hedges SB The origin and evolution of model organisms Nat Rev Genet 2002 3 838 849 12415314 10.1038/nrg929 Thompson JD Gibson TJ Plewniak F Jeanmougin F Higgins DG The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools Nucleic Acids Res 1997 25 4876 4882 9396791 10.1093/nar/25.24.4876 Perriere G Gouy M WWW-query: an on-line retrieval system for biological sequence banks Biochimie 1996 78 364 369 8905155 10.1016/0300-9084(96)84768-7 Guindon S Lethiec F Duroux P Gascuel O PHYML Online--a web server for fast maximum likelihood-based phylogenetic inference Nucleic Acids Res 2005 33 W557 W559 15980534 10.1093/nar/gki352 Zmasek CM Eddy SR ATV: display and manipulation of annotated phylogenetic trees Bioinformatics 2001 17 383 384 11301314 10.1093/bioinformatics/17.4.383 Konvicka K Campagne F Weinstein H Interactive construction of residue-based diagrams of proteins: the RbDe web service Protein Eng 2000 13 395 396 10877849 10.1093/protein/13.6.395	16332259	PMC1325023	CC BY	2021-01-04 16:32:47	no	BMC Genomics. 2005 Dec 6; 6:173	utf-8	BMC Genomics	2,005	10.1186/1471-2164-6-173	oa_comm
==== Front Clin Pract Epidemiol Ment HealthClinical Practice and Epidemiology in Mental Health : CP & EMH1745-0179BioMed Central 1745-0179-1-271635172110.1186/1745-0179-1-27ResearchPsychiatric patients turnaround times in the emergency department Kropp Stefan [email protected] Christoph [email protected] Wildt Bert [email protected] Udo [email protected] Martin [email protected] Irina [email protected] Marc [email protected] Department of Psychiatry, Psychotherapy and Psychosomatics, Landesklinik Teupitz, 15755 Teupitz, Germany2 Department of Clinical Psychiatry and Psychotherapy, Hannover Medical School, 30623 Hannover, Germany3 Departement of Psychiatry and Psychotherapy, University Erlangen-Nuremberg, Germany4 Department of Social Psychiatry and Psychotherapy, Hannover Medical School, 30623 Hannover, Germany2005 13 12 2005 1 27 27 2 5 2005 13 12 2005 Copyright ©2005 Kropp et al; licensee BioMed Central Ltd.2005Kropp et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Background To analyze the turnaround times of psychiatric patients within the Emergency Department (ED) from registration to discharge or hospitalization in a University Hospital in 2002. Methods Data from a one-year period of psychiatric admissions to the emergency service at a University Hospital were monitored and analyzed focused on turnaround times within the ED. Information on patients variables such as age, sex, diagnosis, consultations and diagnostic procedures were extracted from the patients' charts. Results From 34.058 patients seen in the ED in 2002, 2632 patients were examined by psychiatrists on duty. Mean turnaround time in the ED was 123 (SD 97) minutes (median 95). Patients to be hospitalized on a psychiatric ward stayed shorter within the ED, patients who later were admitted to another faculty, were treated longer in the ED. Patients with cognitive or substance related disorders stayed longer in the ED than patients with other psychiatric diagnoses. The number of diagnostic procedures and consultations increased the treatment time significantly. Conclusion As the number of patients within the examined ED increases every year, the relevant variables responsible for longer or complicated treatments were assessed in order to appropriately change routine procedures without loss of medical standards. Using this basic data, comparisons with the following years and other hospitals will help to define where the benchmark of turnaround times for psychiatric emergency services might be. ==== Body Background Turnaround time is an important parameter that strongly influences patients and staff satisfaction in the emergency department and there are early reports considering this important issue [1]. In the department examined it is defined as the time from patient arrival to either discharge or hospitalization. The measurement of turnaround time is an helpful variable of efficacy which is feasible in most emergency departments. Measuring turnaround times may have two major goals: improving the medical care delivered within reasonable time and taking care of for patients' satisfaction [2-5]. Adhering to a comprehensive philosophy of quality management today is vital for hospitals to compete with other hospitals and healthcare providers. The quality of health care has many different components and is difficult to define. One fundamental principle of quality management is that quality can best be improved if it is measured and compared [6]. Within the emergency department, the measurement of psychiatric service quality with different quality indicators might be useful. As waiting and turnaround times might be crucial to the outcome of a medical disease and the reduction of waiting times will positively influence patients perceptions of a hospital and its services, the purpose of this study is to examine the turnaround times of psychiatric patients according to general and diagnose-related variables as a basis of one possible quality indicator. Methods For the period of one year, all psychiatric emergency consultations (n = 2632) within the ED of the Hanover Medical School in Germany were retrospectively monitored, including the time of registration and discharge from ED. We assessed general patients variables and diagnostic procedures in the ED, which might have an impact on the turnaround times. Each procedure such as physical examination, laboratory tests, ECG, X-ray or cranial CT-scan counted as one item. When patients refused physical examination and no other procedures were done, the number of items was zero. All statistical analyses were performed with SPSS™ 12. Besides descriptive statistics, nonparametric methods such as Pearson's Chi Quare test, Mann-Whitney-U-test, Kruskal-Wallis-test and correlation analysis by Spearman were performed. All tests were two-sided. The significance level was set at α 00.05 or less. Results 2632 patients were assessed, 48.4% were female. The mean age was 43.5 (SD 16.0) years. Female patients were significantly older than male patients (Mann-Whitney-U, Z=-3.4, p = 0.001). 567 patients were secondly referred to a psychiatrist by faculties for consultation (e.g. internal medicine, neurology, surgery). 945 patients were admitted to psychiatric wards and 104 patients to non psychiatric wards. 107 patients refused hospitalization. Substance-related problems (ICD-10 F1X, 672 patients) and psychotic disorders including schizophrenia (ICD-10 F2X, 391 patients) were the most common diagnoses, followed by somatoform, anxiety and neurotic disorders (ICD-10 F4X, 332 patients). 104 patients were involuntary admitted acording to PsychKG (German law concerning psychiatric practice). In 2222 patients, the time from door to release or to hospitalisation was determined. The mean time of stay in ED was 123 (SD 97) minutes, median 95 minutes. 77.9% of all psychiatric patients were released or hospitalized within 3 hours, 88.3% within 4 hours (Fig. 1). Figure 1 Turnaround times. Turnaround times of the general population of 2222 psychiatric patients within the emergency department in the year 2002. To acknowledge possible influence factors, the first analysis was done taking into account working shifts within the staff of the ED. During the day shift, 26.8% of the patients stayed for more than 3 hours in the ED, during the night shift only 17.9% had to stay that long (Chi-Square = 19.8, df = 2, p < 0.001). By dividing the day into six four hour periods, significant differences between the categorized length of stay could be detected (Chi-Square = 55.6, df = 10, p < 0.001). Between midnight and 4 h, the percentage of patients staying longer than three hours dropped to 7,5%. We found no difference between working days or weekends (Chi-Quare = 0.03, df = 2, p = 0.986). Patients who were hospitalized stayed longer than the discharged patients (Mann-Whitney-U, Z=-3.2, p = 0.002). Patients admitted via the ED to a psychiatric ward stayed shorter in the ED than patients who were admitted to another faculty, for example 14.2% of the later admitted psychiatric patients stayed 3 or more hours in the ED. But 25.9% of the patients admitted to other faculties (Chi-Square = 60.4, df = 2, p = <0.001). Patients who belonged to the urban catchment area were discharged significantly more often within an hour than patients from outside the catchment area (28.0% vs. 21.0%; Chi-Square = 13.3, p = 0.001). The established diagnosis had a significant impact on turnaround times (Chi-Square = 154.3, df = 14, p < 0.001): patients with dementia or cognitive disorders often stayed longer than three hours in the ED (51.3%) and they were less likely to stay only one hour (11.5%) or even 1 to 3 hours in the ED (37.2%). Patients with substance-related disorders were treated less likely only up to one hour within the emergency department (16.8%). The group of patients with schizophrenia or personality disorders often stayed only up to one hour in the ED (schizophrenia 42.8%; personality disorders 32.8%). The groups of patients with anxiety, mood or adjustment disorders, or other psychiatric disorders did not reveal relevant differences from the mean time of all patients within the ED. As shown in Fig. 2, with more diagnostic procedures performed, the time patients had to stay within the ED increased significantly (Spearman's rho = 0.132, p = <0.001). When patients were referred for consultations to other faculties (i.e. neurology, internal medicine, surgery) the time within the ED also increased according to the number of consultations as shown in Fig. 3 (Kruskal-Wallis-H, Chi-Square = 489.9, df = 2, p = <0.001). Figure 2 Diagnostic procedures and time. Turnaround times in percent of time intervals according to the number of diagnostic procedures provided to the population of psychiatric patients (0 – 4 procedures). Figure 3 Consultations. Turnaround times in percent of time intervals according to the number of consultations given to psychiatric patients by other facultys (none up to 2 consultations by different faculties). Discussion There is little information of turnaround times within psychiatric services in general EDs. In one sample from Bern, Switzerland [7], the mean length of stay was 77 minutes, but treatment seemed to be organized very differently and the number of psychiatric patients per year was also very different, so a real comparison can not be drawn to this data. But data on turnaround times within general EDs is also hard to obtain. It is of general knowledge, that shorter turnaround times are occurring in rural hospitals and the longest times occur in major academic centers with more than 400 beds. The average of general turnaround time for major teaching hospitals was 204 minutes (median 210 minutes) [8]. In another major academic hospital, the average time for later admitted patients was 330, for discharged patients 123 minutes [9]. For all hospitals in the US, the average waiting time is reported to be 5.8 hours in hospitals with overcrowded EDs [10]. In contrast to only examining the waiting time in the ED until the exact moment medical treatment starts, we calculated the turnaround times of the psychiatric service in the general emergency department of a major academic center providing medical emergency services for nearly every medical subdicipline. From our medical understanding of a psychiatric emergency as well as from the patients' perspective the total time spent in ED – including sufficient treatment and discharge or hospitalization – is vital. Therefore, we retrospectively evaluated the complete process concerning different psychiatric diagnoses and other variables. Women were significantly older than men. This circumstance could be explained by the fact that women are more frequently longer socially integrated, which may be related to differential use of mental health services by men and women. The mean turnaround time of 123 minutes seems quite long compared to the data from Switzerland [7]. Compared to the waiting time without treatment and the phenomenon of overcrowded EDs in the US [10,11], it seems rather short. More interestingly, the data concerning different diagnoses and procedures: patients with dementia stay longer than any other group within the ED, which might be related to other medical conditions which also need treatment and assessment. This also seems to be the case with substance related disorders, where patients often need treatment by other faculties. Patients with personality or schizophrenic disorders seem to be treated faster than any other group. Many of these patients were already known and belong to the urban catchment area. Usually they come for crisis intervention and dont need treatment by other faculties. These points could explain the short turnaround times. Often consultations by other faculties and diagnostic procedures are needed to help and treat psychiatric patients adequately, in some cases these additional examinations might be done for legal reasons. In all cases, every consultation and additional examination expands the turnaround time. As have been recently shown, further research is needed concerning the underlying variables of turnaround times, clinician-based decisions and the quality of care of psychiatric emergency services [12]. The results should be discussed carefully and controversial and are only part of a mosaic of quality dimensions. Patient and staff satisfaction is only one dimension of quality and remain less important than the objective adequacy and accuracy of clinical management according to standard medical criteria [13]. Practioners are at least obligated to provide the most effective care most efficiently. Only in terms of cost-effectiveness we must discuss how to maintain a good standard of care containing costs. Or in the contrary a long turnaround time may be the result of problems in the management of resources and this may influence the health results. For our perspective the health care professional must take into account patient preferences as well as social preferences in assesing and assuring quality [14]. Conclusion Although competition between hospitals and health care providers in Europe now starts to be recognized, few data is available on turnaround times of psychiatric services within general emergency departments. As patients and staff satisfaction or dissatisfaction is strongly correlated not only with the medical treatment but also with turnaround times we have chosen this simple measurement in combination with the psychiatric diagnosis and easily obtainable variables to start evaluating the quality of our service to patients. With this first data, comparisons with following years and other hospitals will help to define where the benchmark of turnaround times for psychiatric patients might be. Competing interests The author(s) declare that they have no competing interests. Authors' contributions SK conceived and designed the evaluation and helped to draft the manuscript. CA participated in designing the evaluation and performed parts of the statistical analysis. BT re-evaluated the clinical data and revised the manuscript. UR evaluated and performed the statistical analysis and revised the manuscript. MO and IA collected the clinical data, interpreted them and revised the manuscript. MZ re-analyzed the clinical and statistical data and revised the manuscript. All authors read and approved the final manuscript. Table 1 Diagnosis and time. Turnaround times and diagnoses according to DSM-IV, time intervals and mean with SD and median are also given. Diagnosis < 1 h >1 h <3 h >3 h Mean(SD); median Delirium, Dementia, Cognitive Disorders 11,5% 37,2% 51,3% 197 (SD 129); 182 Substance-Related Disorders 16,8% 55,6% 27,6% 139 (SD 102); 115 Schizophrenia and Other Psychotic Disorders 42,8% 47,6% 9,6% 89 (SD 78); 65 Mood Disorders 25,6% 57,8% 16,6% 118 (SD 94); 90 Anxiety Disorders, Adjustment Disorders 26,4% 52,2% 21,4% 120 (SD 90); 95 Personality Disorders 21,8% 53,6% 13,6% 98 (SD 73); 80 Acknowledgements We are indebted to Bernhard Brüggen and the whole team of the Emergency Department of the Hanover Medical School for their valuable help with this work and to Patricia Hogreve for re-reading the paper and correcting it. ==== Refs Baker B Rochon J Lenth of stay, short stay units and psychiatric emergency admissions Can J Psychiatry 1989 34 39 42 2924247 Naumann S Miles JA Managing waiting patient's perceptions. The role of process control J Manag Med 2001 15 376 386 10.1108/EUM0000000006184 11765320 Summers M Happell B Patient satisfaction with psychiatric services provided by a Melbourne tertiary hospital emergency department J Psychiatr Ment Health Nurs 2003 10 351 357 10.1046/j.1365-2850.2003.00600.x 12755921 Kihlgren AL Nilsson M Skovdahl K Palmblad B Wimo A Older patients awaiting emergency department treatment Scand J Caring Sci 2004 18 169 176 10.1111/j.1471-6712.2004.00266.x 15147480 Walrath JM Tomallo-Bowman R Maguire JM Emergency Department: Improving Patient Satisfaction Nurs Econ 2004 22 71 74 15108475 Kissling W Seemann U Piwernetz K Quality management in psychiatry Int Clin Psychopharmacol 2001 16 suppl 3 S15 S24 10.1097/00004850-200104003-00003 11459328 Schnyder U Klaghofer R Leuthold A Buddeberg C Characteristics of psychiatric emergencies and the choice of intervention strategies Acta Psychiatr Scand 1999 99 179 187 10100912 Southall AC Harris VV Patient ED turnaround times: a comparative review Am J Emerg Med 1999 17 151 153 10.1016/S0735-6757(99)90049-9 10102315 Chan L Reilly KM Salluzzo RF Variables That Affect Patient Turnaround Times in an Academic Emergency Department Am J Med Qual 1997 12 183 186 9385728 Lewin Group (for) the American Hospital Association) Emergency department overload: a growing crisis: The results of the American Hospital Association Survey of Emergency Department (ED) and Hospital Capacity 2002 Falls Church, VA: American Hospital Association Trzeciak S Rivers EP Emergency department overcrowding in the United States: an emerging treat to patient safety and public health Emerg Med J 2003 20 402 405 10.1136/emj.20.5.402 12954674 Hepp U Moergeli H Trier SN Milos G Schnyder U Attempted Suicide: factor Leading to Hospitalization Can J Psychiatry 2004 49 736 742 15633851 Donabedian A Criteria, norms and standards of quality: what do they mean? Am J Public Health 1981 71 409 412 7468883 Donabedian A The seven pillars of quality? Arch Pathol Lab Med 1990 114 1115 1118 2241519	16351721	PMC1325024	CC BY	2021-01-04 18:01:02	no	Clin Pract Epidemiol Ment Health. 2005 Dec 13; 1:27	utf-8	Clin Pract Epidemiol Ment Health	2,005	10.1186/1745-0179-1-27	oa_comm
==== Front BMC PediatrBMC Pediatrics1471-2431BioMed Central London 1471-2431-5-461633666710.1186/1471-2431-5-46Research ArticleAccuracy of clinical pallor in the diagnosis of anaemia in children: a meta-analysis Chalco Juan P [email protected] Luis [email protected] Carlos [email protected] Nilton Y [email protected] Carlos A [email protected] Instituto de Salud del Niño, Lima, LI 05, Perú2 Universidad Peruana Cayetano Heredia, Lima, LI 05, Perú3 Universidad Nacional Mayor de San Marcos, Lima, LI 05, Perú4 Universidad San Martin de Porres, Lima, LI 05, Perú5 Hospital de Emergencias Pediátricas, Lima, LI 05 Perú2005 8 12 2005 5 46 46 28 7 2005 8 12 2005 Copyright © 2005 Chalco et al; licensee BioMed Central Ltd.2005Chalco 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 Anaemia is highly prevalent in children of developing countries. It is associated with impaired physical growth and mental development. Palmar pallor is recommended at primary level for diagnosing it, on the basis of few studies. The objective of the study was to systematically assess the accuracy of clinical signs in the diagnosis of anaemia in children. Methods A systematic review on the accuracy of clinical signs of anaemia in children. We performed an Internet search in various databases and an additional reference tracking. Studies had to be on performance of clinical signs in the diagnosis of anaemia, using haemoglobin as the gold standard. We calculated pooled diagnostic likelihood ratios (LR's) and odds ratios (DOR's) for each clinical sign at different haemoglobin thresholds. Results Eleven articles met the inclusion criteria. Most studies were performed in Africa, in children underfive. Chi-square test for proportions and Cochran Q for DOR's and for LR's showed heterogeneity. Type of observer and haemoglobin technique influenced the results. Pooling was done using the random effects model. Pooled DOR at haemoglobin <11 g/dL was 4.3 (95% CI 2.6–7.2) for palmar pallor, 3.7 (2.3–5.9) for conjunctival pallor, and 3.4 (1.8–6.3) for nailbed pallor. DOR's and LR's were slightly better for nailbed pallor at all other haemoglobin thresholds. The accuracy did not vary substantially after excluding outliers. Conclusion This meta-analysis did not document a highly accurate clinical sign of anaemia. In view of poor performance of clinical signs, universal iron supplementation may be an adequate control strategy in high prevalence areas. Further well-designed studies are needed in settings other than Africa. They should assess inter-observer variation, performance of combined clinical signs, phenotypic differences, and different degrees of anaemia. ==== Body Background The global prevalence of anaemia is estimated in 2 billion people, that is, in about 30% of the worldwide population[1]. An even larger number of people present iron deficiency [1]. Every 9 of 10 persons affected of anaemia live in developing countries [2]. Anaemia prevalence in Latin America is 46% in children [3], with differences within countries. In Peru and Chile it is 50% and 8%, respectively [4,5]. Anaemia is related to impaired physical growth and mental development [6]. It is also associated to a higher risk of infant and child mortality, particularly when it co-exists with malnutrition and other risk factors [7]. It is therefore important to make a timely and accurate diagnosis and initiate an early intervention to reduce the negative impact of anaemia. The laboratory diagnosis of anaemia through any of several techniques is not widely available and its cost is often unaffordable in poor areas of the world. This stimulated several studies to assess the accuracy of clinical signs for screening of anaemia. The Integrated Management of Childhood Illness (IMCI) strategy developed by the World Health Organization recommends the use of palmar pallor as the initial screening tool [8]. This recommendation is based mainly on the interpretation of results of studies performed in the Gambia [9], Kenya [10], and Malawi [11]. None of these studies showed in fact a clear superiority of palmar pallor. Only the Kenya study showed that palmar pallor performed better than conjunctival pallor when used by health workers but not by study physicians [10]. One of them used packed red cells volume as the gold standard [9]. Packed red cells volume is a controversial gold standard for anaemia, as it varies with different physiologic and pathologic conditions such as hydration status, and its correlation with haemoglobin is not optimal [12]. Thus we were prompted to perform a systematic review to assess the accuracy of clinical pallor in the diagnosis of anaemia. The specific objective of the study was to answer the question of whether there is a clinical sign that best predicts the presence or absence of anaemia in children. The signs most frequently assessed in primary studies are conjunctival, palmar and nailbed pallor. The review did not include respiratory and cardiovascular signs as they are unspecific for anaemia and are furthermore related to severe anaemia with haemodynamic repercussion. Methods The review was aimed to include all studies performed in children aged 0 through 18 years old fulfilling pre-established inclusion criteria. Inclusion criteria 1. Studies on individual or combined accuracy of conjunctival, palmar or conjunctival pallor in the clinical diagnosis of anaemia. 2. Studies performed in children 0 through 18 years old. 3. Original articles. Review articles and letters to editors were not considered, except when they had enough information to assess the diagnostic performance of clinical signs of anaemia. 4. Prospective or retrospective studies performed in outpatient or inpatient children. 5. Articles with enough information to assess the diagnostic performance of clinical signs of anaemia, namely sensitivity, specificity, likelihood ratios and predictive values. 6. Studies in which haemoglobin was used as the gold standard. Exclusion criteria 1. Studies not related to assessment of clinical signs in the diagnosis of anaemia. 2. Studies with insufficient information for deriving the diagnostic performance of clinical signs. 3. Studies in which it was not used a gold standard or those in which haemoglobin was not the gold standard Search strategies Two independent reviewers (JPC, CA) made an Internet search of the literature. The databases searched were the National Library of Medicine database from 1966 through January, 2002 and EMBASE from 1986 through January, 2002. In addition we searched the American and Caribbean Health Sciences Literature (Literatura Americana y del Caribe en Ciencias de la Salud, LILACS) database from 1986 through February, 2002 and the African Health Anthology database from 1924 through July, 2002. This search was combined with a manual tracking of articles deemed relevant and found in the references section of primary and qualitative review articles. Details of the key words used are presented as an appendix [See Additional File 1]. The abstracts of the primarily identified articles were read by the same two independent reviewers to assess whether they were related to the clinical diagnosis of anaemia. Those deemed to be relevant were then retrieved and read as full papers. Any discrepancy between the reviewers was solved by consensus. Methodological quality of primary studies We assessed the methodological quality of primary studies according to modified published recommendations [13]. The quality score was derived by ascribing 2 points for each of the major criteria related to systematic and blind application of clinical signs and gold standard to all patients, and 1 point for each of the remaining criteria. The maximum possible score was 16 and the minimum was 0. The final validity rating was reached by consensus. The quality criteria details are presented as an appendix [see Additional File 2]. Methods for calculating the diagnostic performance of index tests Table 2 × 2 were reconstructed from the original data. Sensitivity, specificity, predictive values, and likelihood ratios with their corresponding 95% CIs were calculated for each primary study. Calculations were performed separately for each clinical sign and by different haemoglobin thresholds used in the primary studies. Whenever the 2 × 2 tables contained a 0 cell, 0.5 was added to all cells to avoid undefined results. The diagnostic odds ratio (DOR) of each primary study was calculated according to the following formula [14]: DOR = [Sensitivity/(1-sensitivity]/[(1-specificty)/specificity] The DOR represents the ratio of the odds of a positive test result in subjects with the disease to the odds of a positive test result in subjects without the disease. A DOR of 1 means that the test has no discriminative power. When the DOR is more than one, the odds of a positive test result are higher in the diseased group. Methods of homogeneity assessment Studies were analyzed separately for homogeneity of results by clinical sign and by haemoglobin threshold through chi-square test for proportions (sensitivity and specificity), through Cochran Q for LR's and DOR's [15] and through DOR graphic plotting of individual studies, along with their 95% CI graphs [16]. Mathematical pooling Pooled proportions (sensitivity and specificity) were calculated through the weighted averages taking into account the sample size of each study. Likewise, DOR's and LR's were pooled. The Mantel-Haenszel fixed effects model was planned to use if the studies were homogeneous for the diagnostic performance indexes and the DerSimonian Laird random effect model if they showed heterogeneity [17]. The 95% CIs were also calculated for all the pooled diagnostic indexes [16]. LR's and DOR's were recalculated after outlier's exclusion. Diagnostic performance and 95% CIs of individual studies, homogeneity assessment, mathematical pooling and weighing were performed through the use of Metadisc software version Beta 1.1.1 [18]. Exploration of heterogeneity Potential sources of heterogeneity on diagnostic performance were assessed through Metaregression [18]. Pre-specified potential influential covariates included clinical setting (outpatients or inpatients), continent of study (Africa, Asia, Latin America), age group (children up to 5 years old, children older than 5 years old), technique of haemoglobin measurement (Hemocue®, spectrophotometry, Coulter®), whether or not the study setting was endemic for malaria or for intestinal worms, type of observer (physician, nurse, technician, parents), and methodological quality score (continuous variable). For each haemoglobin threshold category and for each test, multivariate metaregression was run including the above signaled covariates to assess whether any of them showed a significant influence on lnDOR. The metaregression was weighted by study size and the threshold effect was not considered, as there were not additional cutoff points within each pre-specified haemoglobin threshold. Post-test probabilities of anaemia To graphically illustrate the relative usefulness of each particular clinical sign of anaemia at each haemoglobin threshold, different pre-test probability values were plotted against post-test probability values for both positive (LR+) and negative (LR-) results, before and after outlier's exclusion. The post-test probability for a disorder is another way to assess the value of a diagnostic test. It represents the chances that your patient has a disease. It incorporates information about the disease prevalence, the patient pool, and specific patient risk factors (pre-test probabilities) and information about the diagnostic test itself (the LR). The LR is used to assess how good a diagnostic test is and to help in selecting an appropriate diagnostic test(s) or sequence of tests. The LRs have advantages over sensitivity and specificity because they are less likely to change with the prevalence of the disorder, they can be calculated for several levels of the symptom/sign or test, they can be used to combine the results of multiple diagnostic test and they can be used to calculate the post-test probability for a target disorder. Post-test probabilities can be calculated for different clinical scenarios or settings with various possible pre-test probabilities (disease prevalence), using positive (LR+) and negative (LR-) results for the interest tests. Results Adapted QUORUM statement checklist and flow diagram of the study are included as an appendix [see Additional File 3]. Literature search The number of primarily found articles was 225. Two hundred and two papers were excluded after abstract reading because they were nor relevant to the study objective. Twelve studies were excluded after reading them as full papers, because they were not performed in children (8 studies) [19-26], did not use haemoglobin as reference test (1) [27], did not assess individual signs of anaemia (1) [28], did not present separately results for children (1) [29], or did not perform clinical assessment of pallor (1) [30]. Finally, eleven articles were included in the meta-analysis [10,11,31-39]. All the studies we found had been performed in developing countries, mostly in children underfive. Eight were performed in Africa[10,11,31,35-39], one in Pakistan[32], one in Bangladesh and Uganda [34] and one in Brazil [33]. The Uganda component of one study was excluded as it used packed red cells volume as gold standard [34]. Most studies reported their results using pre-specified thresholds. All used one or more of the following haemoglobin categories: <11 g/dL, <8 g/dL, 7 g/dL, and < 5 g/dL. Only one study reported the results for 7 thresholds [36]. In this case, we re-constructed the results in the above noted 4 categories to allow the comparison of results with the other primary studies. Table 1 summarizes main characteristics of primary studies, including the scores of methodological quality. There were studies that evaluated more than one sub-group of subjects and such results are shown separately. Table 1 Summary of primary studies characteristics Author Year Country Location Ages Number Setting Pallor Haemoglobin cut-off assessed (g/dL) Wurapa FK31 1986 Zambia Rural < 4 y 12 Outpatient C <11 Thaver IH32 1994 Pakistan Urban 6 m-5 y 947 Outpatient C,N,P,T <11 Luby SP11 1995 Malawi Rural < 6 y 1104 Outpatient C,N,P,T <11, <8, <5 Sdepanian VL33 1996 Brazil Urban 6 m-5 y 143 Outpatient C,G,N,P,T <11 Kalter HD34 1997 Bangladesh Urban 2 m-5 y 482 Emergency C,P <11, <8, <5 Zucker JR10 1997 Kenya NS 2 m-5 y 1666 Outpatient C,N,P,T <8, <5 NS 2 m-5 y 1048 Inpatient C,N,P,T <8, <5 Stoltzfus RJ35 1999 Tanzania Urban < 5 ya 613 Outpatient C,N,P <7, Urban < 5 yb 537 Outpatient C,N,P <7 Urban > 5 y 3302 Outpatient C,N,P <7 Getaneh T36 2000 Ethiopia Urban 2 m-5 y 574 Outpatient C,N,P,T <11, <8, <7, <5 Muhe L37 2000 Ethiopia Rural 2 m-5 y 2540 Outpatient C,N,P,T <8, <5 Wamae CN38 2000 Kenya Rural 2–4 y 574 Outpatient P <11 Desai MR 29 2002 Kenya Rural < 5 y 3782 Outpatient C,G,N,P,T <7, <5 Observers Kappa Anaemia Prev. Malaria area Worm area Haemoglobin Technique Quality Score Physicians No 16% NS NS Coulter 12 Physicians No 78% NS NS Hemocue 12 Health workers No 82% NS NS Spectrophotometer 13 Paediatricians/Res. No 41% NS NS Coulter 11 Paediatricians No 81% No NS Hemocue 11 Physicians No 59% NS NS Hemocue 12 Physicians No 91% NS NS Hemocue Health workers No 81% NS Yes Hemocue 14 Health workers No 52% NS Yes Hemocue Health workers No 32% NS Yes Hemocue Nurses Yes 46% Yes Yes Hemocue 12 Physicians Yes 61% Yes Ns Hemocue 13 Health workers No 61% Yes Yes Spectrophotometer 12 Parents No 66% NS NS Hemocue 14 NS: Not stated, C: Conjunctiva, N: Nailbed, P: Palm, T: Tongue, G: General, Res.: residents, Prev.: prevalence. a and b are referred to cohorts studied in 1996 and 1997, respectively. Homogeneity assessment Chi-square test for proportions and Cochran Q for LR's and DOR's showed heterogeneity for results of primary studies within each threshold. For graphical display of the heterogeneity, 95% CIs for DOR's of individual studies are shown in Figure 1, 2, 3, 4. Figure 1 Individual and pooled DOR's at Hb <11 g/dL. Figure 2 Individual and pooled DOR's at Hb <8 g/dL. Figure 3 Individual and pooled DOR's at Hb <7 g/dL. Figure 4 Individual and pooled DOR's at Hb <5 g/dL. Outliers at each haemoglobin category were identified through the DOR's graphs. Point estimates with confidence limits were plotted for each individual study. Those studies or results whose DOR's graphs were outside the 95% bounds of the pooled DOR were considered outliers. At haemoglobin <11 g/dL there was one outlier for palmar pallor [38]. At haemoglobin <8 g/dL there were 2 outliers for conjunctival pallor [10,37], one for nailbed [37], and one for palmar pallor [37]. At haemoglobin <5 g/dL there was 1 outlier for conjunctival pallor [39], 2 for palmar pallor [10,39], and one for nailbed pallor [39]. Mathematical pooling As diagnostic performance of primary tests showed heterogeneity, mathematical pooling for them was calculated using the DerSimonian Laird random effects model, to incorporate variation among studies. With this method the weighted average of LR's and DOR's logs are calculated. Tables 2, 3, 4, 5 show the pooled sensitivities, specificities, LR's, DOR's with their 95% CIs. Table 6 shows the pooled LR's and DOR's after exclusion of outliers. Table 2 Pooled diagnostic performance markers at Hb <11 g/dL Clinical Pallor Total N Diseased Sensitivity (95% IC) Specificity (95% IC) Likelihood Ratio+(CI, 95%) Likelihood Ratio-(CI, 95%) DOR (CI, 95%) AUC* SROC** Conjunctiva 3195 2367 43.6(41.7–45.6) 81.4(78.6–83.9) 2.3(1.7–3.1) 0.7(0.5–0.9) 3.7(2.3–5.9) 0.7058 Palm 3885 2731 39.2(37.4–41.1) 86.7(84.6–88.5) 3.0(2.0–4.6) 0.7(0.6–0.8) 4.3(2.6–7.2) 0.7270 Nailbed 2534 1867 29.2(27.2–31.3) 88.5(85.8–90.7) 2.7(1.6–4.5) 0.8(0.7–0.9) 3.4(1.8–6.3) 0.6942 AUC: area under the curve, SROC: Summary receiver operating characteristics curve. Table 3 Pooled diagnostic performance markers at Hb <8 g/dL Clinical Pallor Total N Diseased Sensitivity (95% IC) Specificity (95% IC) Likelihood Ratio+(CI, 95%) Likelihood Ratio-(CI, 95%) DOR (CI, 95%) AUC SROC** Conjunctiva 8867 2711 70.8(69–72.5) 69(67.8–70.1) 2.5(1.5–4.1) 0.4(0.3–0.7) 5.9(2.4–14.8) 0.7687 Palm 8998 2713 80.9(79.4–82.3) 67.7(66.5–68.8) 2.7(2.3–3.0) 0.3(0.2–0.4) 10.1(6.2–16.4) 0.8274 Nailbed 6841 2043 79.7(77.9–81.4) 71.2(69.9–72.4) 3.0(2.5–3.6) 0.3(0.1–0.5) 12.4(5.9–25.9) 0.8477 AUC: area under the curve, SROC: summary receiver operating characteristics curve. Table 4 Pooled diagnostic performance markers at Hb <7 g/dL Clinical Pallor Total N Diseased Sensitivity (95% IC) Specificity (95% IC) Likelihood Ratio+(CI, 95%) Likelihood Ratio-(CI, 95%) DOR (CI, 95%) AUC SROC** Conjunctiva 8693 428 36.9(32.5–41.6) 88.7(88–89.4) 4.5(1.9–1.1) 0.8(0.6–0.9) 6.6(2.6–17.2) 0.7821 Palm 8726 429 35.7(31.3–40.3) 89.2(88.5–89.8) 4.2(2.3–7.7) 0.7(0.6–0.9) 6.8(3.2–14.4) 0.7851 Nailbed 8716 428 32.2(28–36.8) 90.8(90.1–91.4) 5.8(2.2–15.2) 0.6(0.5–0.9) 10.3(2.7–39.1) 0.8297 AUC: area under the curve, SROC: summary receiver operating characteristics curve Table 5 Pooled diagnostic performance markers at Hb <5 g/dL Clinical Pallor Total N Diseased Sensitivity (95% IC) Specificity (95% IC) Likelihood Ratio+(CI, 95%) Likelihood Ratio-(CI, 95%) DOR (CI, 95%) AUC SROC** Conjunctiva 12649 603 47.6(43.6–51.7) 88.1(87.5–88.7) 8.4(3.9–18.5) 0.6(0.5–0.7) 20.6(10.1–42.2) 0.8889 Palm 12780 603 56.6(52.5–60.5) 87.9(87.3–88.5) 7.7(3.1–19.0) 0.5(0.4–0.5) 21.6(10.5–44.4) 0.8922 Nailbed 10623 494 61.1(56.7–65.4) 87.7(87.1–88.4) 7.9(2.7–22.7) 0.4(0.4–0.5) 22.9(9.8–54.0) 0.8964 AUC: area under the curve, *SROC: summary receiver operating characteristics curve. Table 6 Pooled likelihood ratios and diagnostic odds ratios for index tests after excluding outliers Hb <11 g/dL Hb <8 g/dL Hb <5 g/dL Clinical pallor LR+ LR- DOR LR+ LR- DOR LR+ LR- DOR Conjunctiva --- --- --- 2.6 0.4 6.4 7.1 0.5 19.7 --- --- --- (2.2 – 3.0) (0.2 – 0.7) (3.7 – 11.0) (2.8 – 17.6) (0.4 – 0.7) (11.5 – 33.5) Palmar 2.6 0.8 3.5 2.7 0.3 8.3 6.9 0.5 23.8 (1.8 – 3.6) (0.7 – 0.8) (2.3 – 5.1) (2.2 – 3.2) (0.2 – 0.5) (5.5 – 12.7) (2.3 – 20.4) (0.4 – 0.6) (13.4 – 42.3) Nailbed --- --- --- 3.1 0.3 9.3 5.6 0.4 16.6 --- --- --- (2.3 – 4.2) (0.2 – 0.6) (4.9 – 17.7) (2.1 – 15.3) (0.4 – 0.6) (7.7 – 35.9) Numbers in parenthesis denote 95% CIs. Dotted lines (---) denote absence of outliers. Hb <7 g/dL did not have outliers. Pooled DOR at haemoglobin <11 g/dL was 4.3 (95% CI 2.6–7.2) for palmar pallor, 3.7 (2.3–5.9) for conjunctival pallor, and 3.4 (1.8–6.3) for nailbed pallor. For the same haemoglobin threshold, pooled LR+ was 3.0 (95% CI 2.0–4.6) for palmar pallor, 2.7 (1.6–4.5) for nailbed pallor, 2.3 (1.7–3.1) for conjunctival pallor. Also for haemoglobin <11 g/dL, pooled LR- was 0.7 (CI 95% 0.6–0.8) for palmar pallor, 0.7 (0.5–0.9) for conjunctival pallor, and 0.8 (0.7–0.9) for nailbed pallor. DOR's and LR's were slightly better for nailbed pallor at all other haemoglobin thresholds. The pooled diagnostic parameters did not vary substantially after excluding outliers, except that the modest DOR superiority for palmar pallor at haemoglobin <11 g/dL disappeared and improved over the other signs at haemoglobin <5 g/dL (Table 6). Exploration of heterogeneity Method of haemoglobin measurement, type of examiner, continent, clinical setting (outpatients or inpatients) and quality score entered as covariates for most studies. Multivariate metaregression revealed that the only influential covariates on LnDOR for palmar pallor were type of observer (β = 3.16, p = 0.04; RDOR = 23.6, 95% CI = 1.05–531) and haemoglobin technique (β = -5.02, p = 0.03; RDOR = 0.01, 95% CI = 0.00–0.4) at haemoglobin <8 g/dL. There was not any other covariate significantly influencing on LnDOR for the other index clinical signs at any other haemoglobin threshold. Post-test probabilities of anaemia Figures 5, 6, 7, 8 show the post-test probabilities for each particular sign at different haemoglobin thresholds before and after exclusion of outliers. Mild anaemia accounts for the greatest burden of disease in the world. Thus, considering different possible settings according to anaemia prevalence (pre-test probability), we illustrate here some post-test probabilities for positive and negative results of clinical signs of anaemia at haemoglobin <11 g/dL (Figure 5). For a clinical sign present and at 8% of anaemia prevalence, the post-test probability of disease increased to 21% for palmar pallor, to 19% for nailbed pallor, and to 17% for conjunctival pallor. For the same threshold and at 50% of anemia prevalence, the post-test probability increased to 75% for palmar pallor, to 73% for nailbed pallor, and to 70% for conjunctival pallor. And at 80% of anemia prevalence, the post-test probability of disease increased to 92% for palmar and nailbed pallor, and to 90% for conjunctival pallor. Like for DOR's and LR's, the discrete superiority of palmar pallor disappeared when outliers were excluded (Figure 5). At the same haemoglobin threshold (<11 g/dL), when a sign was absent, the post-test probability decrease was modest for any of the clinical signs of anaemia (Figure 5). The same trend was observed at other haemoglobin thresholds (Figures 6, 7, 8). Again, the exclusion of outliers did not change substantially the post-test probabilities. Figure 5 Post-test probabilities of positive and negative results of clinical signs of anaemia for different pre-test probabilities, at Hb<11 g/dL, with and without outliers. Figure 6 Post-test probabilities for positive and negative results of clinical signs of anaemia for different pre-test probabilities, at Hb<8 g/dL, with and without outliers. Figure 7 Post-test probabilities of positive and negative results of clinical signs of anaemia for different pre-test probabilities, at Hb<7 g/dL (no outliers were found). Figure 8 Post-test probabilities of positive and negative results of clinical signs of anaemia for different pre-test probabilities, at Hb<5 g/dL, with and without outliers. Discussion We did not find a highly accurate clinical sign for diagnosing anaemia. Palmar pallor was modestly superior at haemoglobin less than 11 g/dL and nailbed was slightly superior at all other haemoglobin thresholds. After exclusion of outliers, nailbed pallor performed slightly better than the other signs, except at haemoglobin less than 5 g/dL, where palmar pallor improved somewhat over conjunctival and nailbed pallor. Sensitivity of clinical signs ranged widely from 29.2% through 80.9% at different haemoglobin thresholds. Only palmar pallor showed 80.9% of sensitivity at haemoglobin less than 8 g/dL. And only nailbed pallor reached a 90.8% of specificity at haemoglobin less than 7 g/dL. This means that the rates of false positive and false negative results are unacceptably high for the clinical diagnosis of anaemia. The prevalence of asymptomatic anaemia in children may be as high as 87% in some areas of the world such south-eastern Tanzania [40]. Iron supplementation for all children in such a setting with a silent burden of anaemia has been suggested as a control strategy [40]. This can be associated with periodical deworming in tropical and subtropical countries [41]. A technical document supporting the use of palmar pallor as part of IMCI guidelines states that for detection of severe anaemia clinical signs should be as sensitive and specific as possible, to avoid missing referral for a potentially life-saving blood transfusion, and to avoid unnecessary referrals which would burden the families and the health facilities [8]. We did not find high values of sensitivity and specificity for any of the clinical signs of anaemia. Palmar pallor, the recommended sign, did not perform particularly better. The pooled sensitivities are higher and the pooled specificities lower than those for more severe anaemia. This is due to the fact that different studies did not assess necessarily the same haemoglobin cut-offs, which is shown in Table 1. By contrast, the diagnostic odds ratio, which constitutes a single test performance indicator, increases at lower haemoglobin cut-off values. This is one of the reasons we chose to include DOR's as another summary statistics for pooling accuracy of clinical signs in our study. In addition, DOR offers the advantage of overcoming the under-estimation of diagnostic accuracy that often occurs if one pools the results of primary studies just in terms of sensitivity and specificity. We further explored whether the post-test probabilities of anaemia changed substantially when clinical signs were present or absent for different prevalences of the condition (pre-test probabilities). Pre-test probabilities of 50% and 8% represent the anaemia prevalence of Peru and Chile, respectively [4,5], and 80% is close to prevalence recently reported in southern Tanzania [40]. At haemoglobin less than 11 g/dL, post-test probabilities of anaemia did not show a substantial change in presence or absence of clinical signs, except in a scenario with 8% of prevalence. On an individual basis, a good clinical sign would lead correctly, when present, to prescribe iron and conversely, when absent, to withhold it. On the other side, an accurate diagnosis of severe anaemia should lead to a prompt referral for blood transfusion and additional interventions depending on the underlying causes of anaemia. However, at haemoglobin less than 5 g/dL, post-test probabilities of disease when a clinical sign was present increased up to 5 times only in a scenario with 8% of anaemia prevalence, but less than the double in scenarios with 50% and 80% of prevalence. The post-test probabilities decreased only slightly when a clinic sign was absent, at both haemoglobin less than 11 and 5 g/dL, irrespective of anaemia prevalence. Thus, this meta-analysis does not support the recommendation of taking a management decision on the basis of presence or absence of any of the clinical signs of anaemia assessed. There are some limitations of primary studies included that may have influenced on the results of the meta-analysis. First, most studies did not assess inter-observer variation. Due to the subjective component in the appreciation of clinical pallor, it is important to quantify this factor. Second, most studies were performed in Africa, limiting their generalizability to other regions of the world due to phenotypic differences, varying anaemia prevalence and different causes such as malaria and intestinal parasitosis. For instance, it has been documented that a greater palmar pigmentation in Bangladesh was associated with a decreased sensitivity of palmar pallor [34]. In addition, high rates of blepharoconjunctivitis may obscure conjunctival pallor and also decrease its sensitivity [42]. Third, the diagnostic accuracy of combined signs was rarely performed [32,34]. Combining signs may increase the performance of clinical signs, even if such an evaluation may also increase in complexity. Trade-offs between combination of clinical signs and complexity of evaluation should be considered if combined signs display better diagnostic accuracy. Conclusion We did not find a highly accurate clinical sign of anaemia. In view of poor performance of clinical signs, universal iron supplementation of children may be an adequate control strategy at public health level, particularly in high prevalence areas, as was recently suggested [40]. Further well-designed studies are needed for settings other than Africa. They should assess inter-observer variation, performance of combined clinical signs, phenotypic differences and different degrees of anaemia. Competing interests Luis Huicho is one of the principal investigators of the Multi-Country Evaluation of Integrated Management of Childhood Illness (IMCI), coordinated by the Department of Child and Adolescent Health of WHO and supported by the Bill and Melinda Gates Foundation and the US Agency for International Development. This project is aimed at evaluating the impact, cost and effectiveness of IMCI. No other conflict of interest declared by any other author. Authors' contributions J.PC took part in conception, design, data collection, management and analysis; LH in conception, design, data management and analysis; CA in data collection and analysis; NYC and CAB in data analysis. All authors contributed to interpretation of the data and writing of the report. Pre-publication history The pre-publication history for this paper can be accessed here: Supplementary Material Additional File 1 Key words used in the search. details of key words used in the search Click here for file Additional File 2 Criteria for the assessment of the methodological quality of primary studies. Details of the quality criteria and scores ascribed for each of them Click here for file Additional File 3 Adapted QUORUM statement checklist and flow diagram of the systematic review. It includes the completed QUORUM statement checklist and the flowchart Click here for file Acknowledgements Thanks are due to Juan José Chalco and Bertha Huarez for assistance in edition of figures. ==== Refs International Nutritional Anemia Consultative Group INACG Symposium 12 March 1999, Durban, South Africa ILSI Research Foundation, Washington, D.C 1 60 WHO Turning the tide of malnutrition: responding to the challenge of the 21st century 2000 Geneva: WHO (WHO/NHD.007) (accessed Oct 17, 2005) International Nutritional Anemia Consultative Group Report of the 2003 International Nutritional Anemia Consultative Group Symposium Integrating programs to move iron deficiency and anemia control forward (accessed Oct 17, 2005) Instituto Nacional de Estadistica e Informatica Encuesta Demográfica y de Salud Familiar 2000 2001 Lima, Perú Olivares M Pizarro F Hertrampf E Walter T Arredondo M Letelier A Fortificación de alimentos con hierro en Chile Rev Chil Nutr 2000 27 340 44 Pollit E The developmental and probabilistic nature of the functional consequences of iron-deficiency anemia in children J Nutr 2001 131 S669 75 Brabin BJ Premji Z Verhoeff F An analysis on anemia and child mortality J Nutr 2001 131 S636 45 WHO Division of Child and Development Integrated management of childhood illness: conclusions Bull World Health Org 1997 75 S119 28 Weber MW Kellingray SD Palmer A Jaffar S Mulholland EK Greenwood BM Pallor as a clinical sign of severe anaemia in children: an investigation in the Gambia Bull World Health Org 1997 75 S113 18 Zucker JR Perkins BA Jafari H Otieno J Obonyo C Campbell CC Clinical signs for the recognition of severe anaemia in western Kenya Bull World Health Org 1997 75 S97 102 Luby SP Kasembe PN Redd CZ Ciba C Nwanyanwu OC Hightower AW Franco C Chitsulo L Wirima JJ Olivar MA Using clinical signs to diagnose anaemia in African children Bull World Health Organ 1995 73 477 82 7554019 Graitcer PL Goldsgy JB Nichaman MZ Hemoglobins and hematocrits: are they equally sensitive in detecting anemias? Am J Clin Nut 1981 34 61 4 Mulrow CD Linn WD Gaul MK Pugh JA Assessing quality of a diagnostic test evaluation J Gen Intern Med 1989 4 288 95 2760697 Moses LE Shapiro D Littenberg B Combining independent studies of a diagnostic test into a summary ROC curve: data-analytic approaches and some additional considerations Stat Med 1993 12 1293 1316 8210827 Devillé WL Buntinx F Bouter LM Montori VM de Cet HCW vander Windt AWM Bezemer PD Conducting systematic reviews of diagnostic studies: didactic guidelines BMC Med Res Methodol 2002 2 9 12097142 Leemis LM Trivedi KS A Comparison of approximate interval estimators for the Bernoulli parameter Am Stat 1996 50 63 8 DerSimonian R Laird N Meta-analysis in clinical trials Controlled Clin Trials 1986 7 177 88 3802833 Zamora J Muriel A Abraira V Meta-DiSc for Windows. A Software package for the Meta-analysis of Diagnostic Tests XI Cochrane Colloquium Barcelona 2003 (accessed Oct 17, 2005) Gjorup T Bugge PM Hendriksen C Jensen AM A critical evaluation of the clinical diagnosis of anemia Am J Epidemiol 1986 124 657 65 3752058 Strobach RS Anderson SK Doll DC Ringenberg QS The value of the physical examination in the diagnosis of anemia. Correlation of the physical findings and the hemoglobin concentration Arch Intern Med 1988 148 831 2 3355303 Sanchez-Carrillo CI Ramirez-Sanchez TJ Zambrana-Castaneda M Selwyn BJ Test of a noninvasive instrument for measuring hemoglobin concentration Int J Technol Assess Health Care 1989 5 659 67 2699472 Sanchez-Carrillo CI Bias due to conjunctiva hue and the clinical assessment of anemia J Clin Epidemiol 1989 42 751 4 2760666 Nardone DA Roth KM Mazur DJ McAfee JH Usefulness of physical examination in detecting the presence or absence of anemia Arch Intern Med 1990 150 201 4 2297289 Wallace DE McGreal GT O'Toole G Holloway P Wallace M McDermott EW Blake J The influence of experience and specialisation on the reliability of a common clinical sign Ann R Coll Surg Engl 2000 82 336 8 11041034 Hung OL Kwon NS Cole AE Dacpano GR Wu T Chiang WK Goldfrank LR Evaluation of the physician's ability to recognize the presence or absence of anemia, fever, and jaundice Acad Emerg Med 2000 7 146 56 10691073 Ingram CF Lewis SM Clinical use of WHO haemoglobin colour scale: validation and critique J Clin Pathol 2000 53 933 7 11265179 Weber MW Kellingray SD Palmer A Jaffar S Mulholland EK Greenwood BM Pallor as a clinical sign of severe anaemia in children: an investigation in the Gambia Bull World Health Org 1997 S113 8 Montresor A Albonico M Khalfan N Stoltzfus RJ Tielsch JM Chwaya HM Savioli L Field trial of a haemoglobin colour scale: an effective tool to detect anaemia in preschool children Trop Med Int Health 2000 5 129 33 10747273 Ekunwe EOA Predictive value of conjunctival pallor in the diagnosis of anaemia West Afr J Med 1997 16 246 50 9473962 Kent AR Elsing SH Hebert RL Conjunctival vasculature in the assessment of anemia Ophthalmology 2000 107 274 7 10690824 Wurapa FK Bulsara MK Boatin BA Evaluation of conjunctival pallor in the diagnosis of anaemia J Trop Med Hyg 1986 89 33 6 3746991 Thaver IH Baig L Anaemia in children: Part I. Can simple observations by primary care provider help in diagnosis? J Pak Med Assoc 1994 44 282 84 7861584 Sdepanian VL Silvestrini WS de Morais MB Diagnostic limitations of the physical examination in the identification of children with anemia Rev Assoc Med Bras 1996 42 169 74 9138361 Kalter HD Burnham G Kolstad PR Hossain M Schillinger JA Khan NZ Saha S Kednya-Mugisha N Schwartz B Black RE Evaluation of clinical signs to diagnose anaemia in Uganda and Bangladesh, in areas with and without malaria Bull World Health Organ 1997 75 S103 11 Stoltzfus RJ Edward-Raj A Dreyfuss ML Albonico M Montresor A Dhoj Thapa M West KP JrChwaya HM Savioli L Tielsch J Clinical pallor is useful to detect severe anemia in populations where anemia is prevalent and severe J Nutr 1999 129 1675 81 10460203 Getaneh T Girma T Belachew T Teklemariam S The utility of pallor detecting anemia in under five years old children Ethiop Med J 2000 38 77 84 11144886 Muhe L Oljira B Degefu H Jaffar S Weber MW Evaluation of clinical pallor in the identification and treatment of children with moderate and severe anaemia Trop Med Int Health 2000 5 805 10 11123829 Wamae CN Mwanza J Makama S Palmar Pallor as an Indicator for Anthelminthic Treatment Among Ill Children Aged 2–4 Years – Western Kenya, 1998 MMWR 2000 49 278 81 10778792 Desai MR Philips-Howard PA Terlouw DJ Wannemuehler KA Odhacha A Kariuki SK Nahlen BL ter Kuile FO Recognition of pallor associated with severe anaemia by primary caregivers in western Kenya Trop Med Int Health 2002 7 831 39 12358617 Schellenberg D Schellenberg JR Mushi A Savigny D Mgalula L Mbuya C Victora CG The silent burden of anaemia in Tanzanian children: a community-based study Bull World Health Organ 2003 8 581 90 14576890 INACG/WHO/UNICEF Stoltzfus RJ, Dreyfuss ML Guidelines for the Use of Iron Supplements to Prevent and Treat Iron Deficiency Anaemia 2000 International Life Sciences Institute Washington, DC) 2000 Simoes EAF Desta T Tessema T Gerbresellassie T Dagnew M Gove S Performance of health workers after training in integrated management of childhood illness in Gondar, Ethiopia Bull World Health Organ 1997 75 S43 53	16336667	PMC1325025	CC BY	2021-01-04 16:31:07	no	BMC Pediatr. 2005 Dec 8; 5:46	utf-8	BMC Pediatr	2,005	10.1186/1471-2431-5-46	oa_comm
==== Front Reprod Biol EndocrinolReproductive Biology and Endocrinology1477-7827BioMed Central London 1477-7827-3-711635617510.1186/1477-7827-3-71ResearchClinical benefit of metaphase I oocytes Vanhoutte Leen [email protected] Sutter Petra [email protected] der Elst Josiane [email protected] Marc [email protected] Infertility Centre, Ghent University Hospital, De Pintelaan 185, 9000 Ghent, Belgium2005 15 12 2005 3 71 71 20 10 2005 15 12 2005 Copyright © 2005 Vanhoutte et al; licensee BioMed Central Ltd.2005Vanhoutte 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 studied the benefit of using in vitro matured metaphase I (MI) oocytes for ICSI in patients with a maximum of 6 mature metaphase II (MII) oocytes at retrieval. Methods In 2004, 187 ICSI cycles were selected in which maximum 6 MII oocytes and at least one MI oocyte were retrieved. MI oocytes were put in culture to mature until the moment of ICSI, which was performed between 2 to 11 hours after oocyte retrieval (day 0). In exceptional cases, when the patient did not have any mature oocyte at the scheduled time of ICSI, MI oocytes were left to mature overnight and were injected between 19 to 26 hours after retrieval (day 1). Embryos from MI oocytes were chosen for transfer only when no other good quality embryos from MII oocytes were available. Outcome parameters were time period of in vitro maturation (IVM), IVM and fertilization rates, embryo development, clinical pregnancy rates, implantation rates and total MI oocyte utilization rate. Results The overall IVM rate was 43%. IVM oocytes had lower fertilization rates compared to in vivo matured sibling oocytes (52% versus 68%, P < 0.05). The proportion of poor quality embryos was significantly higher in IVM derived oocytes. One pregnancy and live birth was obtained out of 13 transfers of embryos exclusively derived from IVM oocytes. This baby originated from an oocyte that was injected after 22 hrs of IVM. Conclusion Fertilization of in vitro matured MI oocytes can result in normal embryos and pregnancy, making IVM worthwhile, particularly when few MII oocytes are obtained at retrieval. ==== Body Background A proportion of human oocytes collected during an IVF or ICSI procedure remains meiotically immature at the germinal vesicle (GV) or metaphase I (MI) stage. Several publications have shown that this proportion fluctuates around 15 to 20% [1,2]. It is not exactly known why some of the oocytes are unresponsive to the maturation trigger in vivo. Different explanations are possible. When ovarian hyperstimulation is started, follicles may be at different stages of development, producing oocytes of varying degrees of maturity. Since follicles are aspirated prior to rupture, the collected oocytes come from a heterogeneous pool of follicles, including luteinizing as well as degenerating follicles [3]. It is also possible that smaller antral follicles are aspirated during oocyte retrieval, which can result in the collection of immature oocytes [4,5]. Eventually, the proportion of immature oocytes can be dependent on patients' characteristics (such as cause of infertility, age, ovarian reserve) and the stimulation protocol used. Immature oocytes from superovulated cycles can undergo the final stages of meiotic maturation spontaneously in vitro. MI oocytes have already undergone the process of germinal vesicle breakdown (GVBD) and may progress to the metaphase II (MII) stage within a few hours of in vitro culture. This allows them to be injected by ICSI at the same time as their sibling mature MII oocytes. The clinical use of MI oocytes from stimulated cycles has been studied by several research groups. It has been reported that these in vitro matured oocytes yield lower fertilization rates [6-8], abnormal embryonic development [7-9] and lower implantation rates [6] than in vivo matured oocytes. Development to term is limited to rare cases [6,8,9]. As a consequence, immature oocytes from stimulated cycles are generally considered to be a side-product and only in vivo matured MII oocytes are used for ICSI. Nevertheless, for patients in whom a low number of MII oocytes are retrieved, the use of in vitro matured MI oocytes may be worthwhile in order to increase the number of injectable oocytes at the time of ICSI. We therefore designed the present study to determine if, indeed, in vitro matured MI oocytes could have a clinical application in our IVF-program in a selected population of patients with a low number of MII oocytes at retrieval. This was analyzed within the framework of a daily laboratory practice, without changing the routine of ovarian stimulation, oocyte retrieval, ICSI procedure, embryo culture or transfer. Methods Patient selection The study included all ICSI cycles over a one year period (2004) in which two inclusion criteria were fulfilled: 1) a maximum of 6 mature (MII) oocytes and 2) at least one MI oocyte present at retrieval. Our ICSI program has been approved as infertility treatment by the Ghent University Hospital Ethical Committee. Ovarian stimulation, IVM and oocyte handling for ICSI All patients underwent controlled ovarian stimulation after cycle synchronization with a standard contraceptive pill for 2–6 weeks. A short gonadotrophin-releasing hormone (GnRH) agonist protocol was used, consisting of 0.1 mg of triptorelin (Decapeptyl, Ipsen, France) from day 5 onwards after discontinuation of the oral contraceptive. This was followed by human menopausal gonadotrophin (hMG; Menopur, Ferring, Germany) or follicle stimulating hormone (FSH; either Gonal-F, Serono, Switzerland or Puregon, Organon, The Netherlands) from day 7 after discontinuation of the pill onwards. The starting dose was usually 150 IU, but this dose was adjusted after 7 days of hMG or FSH administration, according to the individual response of the patient. Known poor responders were started on 300 IU of hMG or FSH daily. The follicular phase was monitored by means of transvaginal ultrasound scanning of the ovaries and serum estradiol measurement if judged necessary. An injection of 5,000 or 10,000 IU human chorionic gonadotrophin (hCG; Pregnyl, Organon, The Netherlands) was administered when half of all mature follicles had reached a mean diameter of at least 20 mm, measured in two planes. Oocyte retrieval was scheduled 34 to 36 hrs after hCG administration. Oocytes were denuded enzymatically by a brief exposure of the cumulus-oocyte complexes to 80 IU/ml hyaluronidase (Type VIII; Sigma Chemical Co., Bornem, Belgium), followed by mechanical denudation approximately 1 to 2 hrs after oocyte collection. The nuclear status of denuded oocytes was subsequently recorded. GV oocytes were not considered for ICSI. MI oocytes were defined as those oocytes in which no GV and no first polar body were visible. These oocytes were put in culture to mature. The culture medium for in vitro maturation (IVM) was either Sydney IVF Fertilization Medium (Cook, Ltd., Limerick, Ireland) or Early Cleavage Medium (Irvine Scientific, Brussels, Belgium). MI oocytes were left to mature until the time when ICSI for the particular patient was carried out. This was in a time-frame of 2 to 11 hrs after oocyte retrieval (day 0). In exceptional cases, when the patient did not have any mature oocyte at the scheduled time of ICSI, MI oocytes were left to mature overnight and, when matured, were injected between 19 to 26 hrs after oocyte retrieval (day 1). In vivo and/or in vitro matured oocytes were injected with patient's sperm. Sperm preparation for ICSI and details for the microinjection procedure have been described elsewhere [10]. After injection, oocytes were cultured in either Sydney IVF Cleavage Medium (Cook) or Early Cleavage Medium. Embryo evaluation and transfer Assessment of fertilization took place between 16 to 20 hrs after ICSI. Embryos were evaluated based on the number of blastomeres and the degree of fragmentation. Embryos with less than 10% anucleated fragments were classified as 'excellent'. Embryos with either 10–20% or >20% anucleated fragments were classified as 'good' and 'poor' quality embryos, respectively. Embryos with at least one blastomere with more than one nucleus were classified as multinucleated embryos and were considered as 'poor' quality embryos as well, regardless of the degree of fragmentation. Transfer of embryos was carried out on day 2 or day 3. Embryos from in vitro matured MI oocytes were chosen for transfer only when no other good quality embryos from MII oocytes were available. The number of embryos transferred was limited by Belgian law based on female age, cycle number and embryo quality [11]. Pregnancy was diagnosed by the detection of a positive serum hCG at least 14 days after embryo transfer, followed by a rise in hCG levels. All patients received a transvaginal ultrasound scan between 6 and 7 weeks of pregnancy to differentiate between biochemical and clinical (presence of an intra-uterine gestational sac with fetal heart beat) pregnancies and to diagnose ectopic implantations. All pregnancies were monitored further by transvaginal ultrasound until 12 weeks of amenorrhoea. Statistical analysis For comparison among groups, results were analyzed with Chi-square and Fisher's exact test when appropriate. When the P value was <0.05, the difference was considered significant. Results Patient and cycle characteristics In the year 2004, 180 patients underwent 187 ICSI cycles in which maximum 6 MII oocytes and at least one MI oocyte were retrieved. This is 12.7% of the total number of ICSI cycles performed in our infertility center during the same year. The mean age of the patients was 35.9 ± 4.67 years (range 25–48). A total of 1208 oocytes were collected. Three hundred of these oocytes were at the MI stage at the moment of oocyte denudation (24.8%; mean of 1.6 MI oocytes/cycle; range 1–6). Hundred thirty-two oocytes were at the GV stage (10.9%) and 80 oocytes (6.6%) were degenerated or damaged at the moment of denudation. Comparison between in vitro matured MI oocytes and in vivo matured oocytes Overall, 43% (129/300) of MI oocytes matured to the MII stage. Maturation and injection of at least one MI oocyte was achieved in approximately half of the ICSI cycles (55%; n = 102). In vivo matured sibling MII oocytes, retrieved in the same treatment cycles, were injected in parallel and served as the control group. ICSI was performed with fresh ejaculate in 81.4% of the cycles and with frozen ejaculate in 9.8% of the cycles. Frozen epididymal or testicular spermatozoa were used in respectively 1.0% and 7.8% of the cycles. The results of fertilization and embryo development in the two groups of oocytes (IVM + control) are presented in Table 1. The fertilization rate of matured MI oocytes was significantly lower compared to the fertilization rate of in vivo matured oocytes (52% versus 68%, P < 0.05). Also the embryo quality was different between the two groups. In the in vitro matured group, significantly less embryos of excellent quality and more embryos of poor quality were obtained compared to the in vivo matured group. This was observed on day 2 (p < 0.05) as well as on day 3 (p < 0.05). Table 1 Comparison of fertilization and embryonic development between in vitro and in vivo matured sibling oocytes In vitro matured MI oocytes In vivo matured sibling oocytes (control) Total N° of oocytes injected 129 339 N° (%) of normal fertilized oocytesa 67 (52)* 229 (68) Embryo development on day 2 (%)b Excellent 24* 41 Good 27 32 Poor 49* 27 Embryo development on day 3 (%)b Excellent 22* 38 Good 28 32 Poor 50* 30 * Statistically different between columns (p < 0.05). a Normal fertilized oocytes were defined as oocytes containing 2 pronuclei en 2 polar bodies. b The proportions of excellent, good and poor embryos are expressed per total number of embryos obtained on day 2 or day 3. Comparison between different time periods of IVM A second evaluation of the results was done by splitting up the period of IVM, between collection of the oocytes and the time of ICSI, in different time intervals (Table 2). Table 2 Comparison of maturation rates, fertilization rates and embryo development between different time periods of IVM Time interval of IVM N° of MI oocytes N° (%) of in vitro matured oocytes N° (%) of fertilized oocytes N° (%) of excellent + good quality embryos on Day 2 (c) N° (%) of excellent + good quality embryos on Day 3 2 to 4 hrs 147 60 (41) 26 (43)a 15 (58) 14 (61) >4 to 7 hrs 136 57 (42) 31 (54)a 15 (48) 9 (38) >7 to 11 hrs 9 6 (67) 6 (100)b 3 (50) 2 (40) 19 to 26 hrs 8 6 (75) 4 (67)ab 1 (25) 2 (100) a,b Different letters indicate significant differences within columns (p < 0.05). c Thirteen embryos were transferred on day 2. A number of 147 oocytes, coming from 100 ICSI cycles, were evaluated for maturity within 2–4 hrs of IVM culture, 136 oocytes (77 ICSI cycles) within >4–7 hrs of culture, 9 oocytes (5 ICSI cycles) within >7–11 hrs of culture, and 8 oocytes (5 ICSI cycles) were left to mature overnight and were evaluated the day after retrieval, between 19–26 hrs of culture. A time-dependent increase in the progression to maturation was noted, ranging from 41% mature oocytes after 2–4 hrs of IVM to 75% after 19–26 hrs of IVM, but this trend was not statistically different. The fertilization rate in the group of >7–11 hrs of IVM (100%) was significantly higher compared to 2–4 hrs of IVM (43%) and >4–7 hrs of IVM (54%) (p < 0.05), but not to 19–26 hrs of IVM (67%). There was no difference between the number of excellent and good quality embryos in the different IVM time interval groups. Embryo transfer, clinical pregnancy and implantation rates The most important parameter to evaluate whether the use of in vitro matured MI oocytes has a clinical benefit is the pregnancy outcome. Table 3 represents the results of embryo transfer, clinical pregnancy and implantation rates. A distinction was made between cycles in which at least one MI oocyte was matured in vitro (n = 102) and ICSI cycles with no matured MI oocytes (n = 85). The group of cycles with matured MI oocytes was further split-up in 3 groups: 1) transfers involving exclusively embryos derived from in vitro matured MI oocytes, 2) mixed transfers and 3) transfers involving exclusively embryos derived from in vivo matured MII oocytes. A double embryo transfer involving embryos exclusively derived from MI oocytes resulted in a singleton pregnancy and the birth of a healthy baby girl. This baby originated from an oocyte that was injected after 22 hrs of IVM. There was no statistical difference between clinical pregnancy rates and embryo implantation rates in the three groups. However, when the numbers of exclusively MI and mixed transfers were pooled and compared to exclusively MII transfers, the clinical pregnancy rates (8.8% versus 27.0%) and embryo implantation rates (5.6% versus 15.1%) were significantly lower in the former group (p < 0.05). Table 3 Clinical pregnancy and implantation rates in cycles with maximum 6 MII oocytes cycles with at least one MI oocyte matured (n = 102) cycles with no MI oocyte matured (n = 85) Total (n = 187)a Exclusively embryos derived from MI oocytes Mixed transfers Exclusively embryos derived from sibling MII oocytes N° of transfers 13 21 63 81 178 N° of embryos transferred 17 54 119 141 331 N° (%) of clinical pregnancies per embryo transfer 1 (7.7) 2 (9.5) 17 (27.0) 15 (18.5) 35 (19.7) N° (%) of implanted embryos 1 (5.9) 3 (5.6)b 18 (15.1)b 15 (10.6) 37 (11.2) No statistical difference between columns. a In 9 ICSI cycles, no embryo transfer was done. b A twin pregnancy was obtained in these groups. Total MI oocyte utilization rate Forty-two embryos derived from normally fertilized in vitro matured oocytes were used for transfer and 9 embryos were cryopreserved. This means that 51 out of 67 embryos originating from MI oocytes were used. The total MI oocyte utilization rate (= percentage of embryos transferred and frozen per fertilized oocyte) was 76%. Discussion The present study aimed to analyze, for the first time, the clinical benefit of MI oocytes in a selected group of patients with a low number of mature oocytes at retrieval. To achieve this goal, we selected ICSI cycles in which a maximum of 6 mature oocytes and at least one MI oocyte were obtained at oocyte retrieval. The results show that fertilization rate and developmental capacity of the embryos was significantly reduced in IVM oocytes compared with control sibling oocytes. One live birth obtained after transfer of embryos exclusively derived from IVM oocytes illustrates that the use of MI oocytes is not of major issue in IVF programs, but may be an option for patients with low numbers of MII oocytes. A low number of oocytes at retrieval might be a result of low ovarian response to gonadotrophin stimulation. Low response to stimulation occurs in approximately 10% of the ART population [12]. There is no universally accepted definition for low response. One of the criteria is the number of oocytes retrieved. Faber et al. [13] used an oocyte retrieval rate of ≤ 4 mature oocytes as cut-off limit, while De Sutter et al. [14] and Moreno et al. [15] categorized < 5 and ≤ 6 oocytes retrieved, respectively, as low responding patients, without distinguishing mature and immature oocytes. Based on these different definitions, it can be concluded that the patients in our study can be categorized as 'relatively poor responders'. In each cycle of this study, at least one MI oocyte was present. The proportion of immature oocytes (GV +MI = 35.7%) in this group of relatively poor responding patients is high compared to the percentages described in the literature after ovarian stimulation in a non-selected group of patients (15–20%; [1,2]) and, as a consequence, the total MI oocyte utilization rate of 76% shows that embryos from in vitro matured MI oocytes were used at high frequency for transfer or cryopreservation. In the majority of the ICSI cycles, the maturation status of the MI oocytes was checked between 2 to 7 hours after oocyte retrieval. Within this time-frame, 41.3% of the collected MI oocytes extruded their polar body. These maturation rates are comparable with those achieved by others working with MI oocytes retrieved from stimulated cycles. Chian et al. [16] obtained a maturation rate of 46.1% and 52.0% after 6 hrs of in vitro culture and Strassburger et al. [8] obtained 45.1% matured oocytes after 4 hrs of culture. Other studies describe lower maturation rates, like the study of Devos et al. [6] (26.7% maturation after 4 hrs of culture) and the study of Chen et al. [9] (16.4% maturation after 9 hrs of culture). Variations in maturation rates between studies might be explained by different starting and ending points for in vitro maturation. Also the culture conditions, the use of more suitable types of media and/or the addition of serum, growth factors and hormones might influence maturation rates, subsequent fertilization and embryo development of in vitro matured oocytes [16-18]. The group of Balakier et al. [7] performed an exact time recording of polar body extrusion during IVM of MI oocytes. They found that the highest fertilization rate and the lowest incidence of multinucleation were obtained when injection of the oocytes was performed between 3 to 6 hrs after extrusion of the first polar body. This indicates that oocyte maturation is not completed upon reaching the MII stage. In the present study, we did not perform an accurate kinetic experiment, but proper timing of polar body extrusion as well as injection may enhance the outcome of in vitro matured oocytes. Whatever the conditions of in vitro maturation applied, our study and the majority of other studies show consistently lower fertilization rates of in vitro matured MI oocytes compared to sibling in vivo matured oocytes [6-8]. The proportion of poor quality embryos was higher after in vitro maturation. This is in accordance with other publications, which describe more cleavage arrest [7,8], a higher number of multinucleated blastomeres [7] and a reduced development to the blastocyst stage [9] after IVM. On the contrary, the study of Devos et al. [6], performed on a large group of 896 ICSI cycles, found the same proportions of excellent and fair quality embryos after IVM compared to in vivo matured oocytes. This might be explained by the IVM incubation time of maximum 4 hrs applied in this study. Longer IVM incubation times could result in oocyte ageing. However, we were not able to find a statistical difference in fertilization rate and embryo quality between the different IVM time intervals in the in vitro matured group, although a larger sample would be necessary for a more conclusive statement in the IVM time intervals of >7–11 hrs and 19–26 hrs. Nevertheless, of special interest was the fact that the live birth we obtained from exclusively IVM oocytes originated from an oocyte that was matured in vitro for a period of 22 hrs. Conclusion We may conclude that the use of in vitro matured MI oocytes can be of benefit to obtain pregnancy in patients with a low number of MII oocytes. IVM culture conditions and time schedule for ICSI must be refined to achieve optimum fertilization and development. Authors' contributions LV and JVDE designed the study. LV collected and analyzed the data of the study and wrote the manuscript. All the authors corrected the manuscript and approved the final version. PDS and MD treated the patients. Acknowledgements The authors wish to thank the clinical and laboratory staff of the Infertility Center. The study was funded by the hospital infertility program. ==== Refs Cha KY Chian RC Maturation in vitro of immature human oocytes for clinical use Hum Reprod Update 1998 66 103 120 9683349 10.1093/humupd/4.2.103 Smitz J Nogueira D Vanhoutte L de Matos DG Cortvrindt R Gardner DK, Weissman A, Howles CM, Shoham Z Oocyte in vitro maturation Textbook of Assisted Reproductive Technology 2004 Chapter 10 2 Londen: Martin Dunitz LTD 125 161 Stouffer RL Zelinski-Wooten MB Overriding follicle selection in controlled ovarian stimulation protocols: quality vs quantity Reprod Biol Endocrinol 2004 2 32 15200679 10.1186/1477-7827-2-32 Ectors FJ Vanderzwalmen P Van Hoeck J Nijs M Verhaegen G Delvigne A Schoysman R Leroy F Relationship of human follicular diameter with oocyte fertilization and development after in-vitro fertilization or intracytoplasmic sperm injection Hum Reprod 1997 12 2002 2005 9363720 10.1093/humrep/12.9.2002 Triwitayakorn A Suwajanakorn S Pruksananonda K Sereepapong W Ahnonkitpanit V Correlation between human follicular diameter and oocyte outcomes in an ICSI program J Assist Reprod Genet 2003 20 143 147 12762412 10.1023/A:1022977002954 De Vos A Van de Velde H Joris H Van Steirteghem A In-vitro matured metaphase-I oocytes have a lower fertilization rate but similar embryo quality as mature metaphase-II oocytes after intracytoplasmic sperm injection Hum Reprod 1999 14 1859 1863 10402405 10.1093/humrep/14.7.1859 Balakier H Sojecki A Motamedi G Librach C Time-dependent capability of human oocytes for activation and pronuclear formation during metaphase II arrest Hum Reprod 2004 19 982 987 15033953 10.1093/humrep/deh158 Strassburger D Friedler S Raziel A Kasterstein E Schachter M Ron-El R The outcome of ICSI of immature MI oocytes and rescued in vitro matured MII oocytes Hum Reprod 2004 19 1587 1590 15131077 10.1093/humrep/deh236 Chen SU Chen HF Lien YR Ho HN Chang HC Yang YS Schedule to inject in vitro matured oocytes may increase pregnancy after intracytoplasmic sperm injection Arch Androl 2000 44 197 205 10864367 10.1080/014850100262173 Dozortsev D Rybouchkin A De Sutter P Qian C Dhont M Human oocyte activation following intracytoplasmic sperm injection: the role of the sperm cell Hum Reprod 1995 10 403 407 7769071 Ombelet W De Sutter P Van der Elst J Martens G Multiple gestation and infertility treatment: registration, reflection and reaction – the Belgian project Hum Reprod Update 2005 11 3 14 15528214 10.1093/humupd/dmh048 Fasouliotis SJ Simon A Laufer N Evaluation and treatment of low responders in assisted reproductive technology: a challenge to meet J Assist Reprod Genet 2000 17 357 373 11077616 10.1023/A:1009465324197 Faber BM Mayer J Cox B Jones D Toner JP Oehninger S Muasher SJ Cessation of gonadotropin-releasing hormone agonist therapy combined with high-dose gonadotropin stimulation yields favorable pregnancy results in low responders Fertil Steril 1998 69 826 830 9591487 10.1016/S0015-0282(98)00040-5 De Sutter P Dhont M Poor response after hormonal stimulation for in vitro fertilization is not related to ovarian aging Fertil Steril 2003 79 1294 1298 12798873 10.1016/S0015-0282(03)00264-4 Moreno C Ruiz A Simon C Pellicer A Remohi J Intracytoplasmic sperm injection as a routine indication in low responder patients Hum Reprod 1998 13 2126 2129 9756282 10.1093/humrep/13.8.2126 Chian RC Chung JT Downey BR Tan SL. Maturational and developmental competence of immature oocytes retrieved from bovine ovaries at different phases of folliculogenesis Reprod Biomed Online 2002 4 127 132 12470574 Cekleniak NA Combelles CM Ganz DA Fung J Albertini DF Racowsky C A novel system for in vitro maturation of human oocytes Fertil Steril 2001 75 1185 1193 11384647 10.1016/S0015-0282(01)01789-7 Roberts R Franks S Hardy K Culture environment modulates maturation and metabolism of human oocytes Hum Reprod 2002 17 2950 2956 12407055 10.1093/humrep/17.11.2950	16356175	PMC1325026	CC BY	2021-01-04 16:26:08	no	Reprod Biol Endocrinol. 2005 Dec 15; 3:71	utf-8	Reprod Biol Endocrinol	2,005	10.1186/1477-7827-3-71	oa_comm
==== Front BMC GeriatrBMC Geriatrics1471-2318BioMed Central London 1471-2318-5-151632115910.1186/1471-2318-5-15Research ArticleGait disorders are associated with non-cardiovascular falls in elderly people: a preliminary study Montero-Odasso Manuel [email protected] Marcelo [email protected] Gustavo [email protected] Enrique R [email protected] Roberto [email protected] Luis A [email protected] Division of Geriatric Medicine, McGill University, Montreal, Canada2 Geriatric Medicine Section, Hospital Italiano de Buenos Aires, Buenos Aires, Argentina3 Internal Medicine Department, Hospital Italiano de Buenos Aires, Buenos Aires, Argentina4 Informatics and Biostatistics Department, Hospital Italiano de Buenos Aires, Buenos Aires, Argentina5 Geriatric Fellowship, Faculty of Medicine, University of Buenos Aires, Buenos Aires, Argentina6 Solidage: McGill University, Université de Montréal Research Group on Integrated Services for Older Persons, Montréal, Canada2005 1 12 2005 5 15 15 8 7 2005 1 12 2005 Copyright © 2005 Montero-Odasso et al; licensee BioMed Central Ltd.2005Montero-Odasso 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 association between unexplained falls and cardiovascular causes is increasingly recognized. Neurally mediated cardiovascular disorders and hypotensive syndromes are found in almost 20 percent of the patients with unexplained falls. However, the approach to these patients remains unclear. Gait assessment might be an interesting approach to these patients as clinical observations suggests that those with cardiovascular or hypotensive causes may not manifest obvious gait alterations. Our primary objective is to analyze the association between gait disorders and a non-cardiovascular cause of falls in patients with unexplained falls. A second objective is to test the sensitivity and specificity of a gait assessment approach for detecting non-cardiovascular causes when compared with intrinsic-extrinsic classification. Methods Cross-sectional study performed in a falls clinic at a university hospital in 41 ambulatory elderly participants with unexplained falls. Neurally mediated cardiovascular conditions, neurological diseases, gait and balance problems were assessed. Gait disorder was defined as a gait velocity < 0.8 m/s or Tinetti Gait Score <9. An attributable etiology of the fall was determined in each participant. Comparisons between the gait assessment approach and the attributable etiology regarding a neurally mediated cardiovascular cause were performed. Fisher exact test was used to test the association hypothesis. Sensitivity and specificity of gait assessment approach and intrinsic-extrinsic classification to detect a non-cardiovascular mediated fall was calculated with 95% confidence intervals (CI95%). Results A cardiovascular etiology (orthostatic and postprandial hypotension, vasovagal syndrome and carotid sinus hypersensitivity) was identified in 14% of participants (6/41). Of 35 patients with a gait disorder, 34 had a non-cardiovascular etiology of fall; whereas in 5 out of 6 patients without a gait disorder, a cardiovascular diagnosis was identified (p < 0.001). Sensitivity and specificity of the presence of gait disorder for identifying a non-cardiovascular mediated cause was 97.1% (CI95% = 85–99) and 83% (CI95% = 36–99), respectively. Conclusion In community dwelling older persons with unexplained falls, gait disorders were associated with non-cardiovascular diagnosis of falls. Gait assessment was a useful approach for the detection of a non-cardiovascular mediated cause of falls, providing additional value to this assessment. ==== Body Background Falls are one of the giants of geriatric medicine, constituting a worldwide prevalent problem with substantial clinical and public health implications. Multiple risk factors have been identified as contributors to the fall syndrome and, accordingly, the list is highly heterogeneous including age-associated changes, neuro-sensory impairments, muscular weakness, comorbidities, cardiovascular mediated problems, polypharmacy, and environmental hazards, among others [1,2]. The most accepted classification of falls is based on whether risk factors are related to an intrinsic disorder or due to an extrinsic hazard [3,4]. However, a co-existence between these intrinsic and extrinsic factors has been found in almost 80 percent of elderly fallers limiting the clinical applicability of this classification [5]. When a principal intrinsic or extrinsic contributor is not clearly identified, patients with recurrent falls are classified as unexplained elderly fallers [2,6]. There is growing evidence of the association between unexplained falls and cardiovascular mediated causes [6-8]. Specifically, neurally mediated cardiovascular disorders and hypotensive syndromes account for up to 20 percent of these cases [8-11]. However, the approach to patients with unexplained falls remains unclear [2,12]. As clinical observation suggests that those with cardiovascular or hypotensive causes may not manifest obvious gait impairment, the assessment of gait impairments might be an interesting approach to these patients [3,10]. Accordingly, the principal objective of this study is to analyze the relationship between gait disorders and cardiovascular mediated falls under the hypothesis that gait disorders are mostly associated with a non-cardiovascular mediated falls. A second objective is to test the sensitivity and specificity of a gait assessment approach for detecting non-cardiovascular cause when compared with the intrinsic-extrinsic falls classification. Methods Study population and data collection Consecutive patients referred to our Falls Clinic at "Hospital Italiano de Buenos Aires' between January 2001 and January 2002 were included in the study. All participants were community dwelling persons over 65 years of age with a history of recurrent and unexplained falls. A fall was defined as "an unintentional change in position resulting in coming to rest at a lower level or on the ground." [5]. Recurrent falls were defined as 2 or more falls in the previous six months [6]. Based on previous studies, unexplained falls were defined as falls without a principal cause attributed after a general practitioner evaluation [2,10]. A consent form was obtained for each participant. Health and cognitive status, medications, and activities of daily living (ADLs) were recorded. Comorbidity was registered by a list of chronic diseases. A fall questionnaire, which included a list of intrinsic and extrinsic risk factors, place and circumstances of the fall, a list of comorbidities, use of psychopharmacological medication, level of physical activity and a history of fear of falling, was administered to each participant. A complete physical examination with supine and erect blood pressure measurement after 3 minutes of active standing, was also performed. All participants were assessed for neurally mediated cardiovascular conditions, neurological diseases, and gait and balance problems. In addition to a standard biochemical and hematological profile, a baseline electrocardiogram was performed in each participant. If diagnostic uncertainty remained, a head-up tilt test with carotid massage, a 24-hour holter electrocardiogram and 24-hour ambulatory blood pressure monitoring were performed. Further investigations, including echocardiography, knee or hip x-rays, computed tomography of the brain, were conducted as per published guidelines. [6] Two geriatricians, who had previously received training sessions and standardized instructions, conducted the assessments. Definitions and measurements at assessment Based on previous studies, gait disorders were defined as either gait velocity (GV) equal or below 0.8 m/s or Tinetti Gait Test score below 9 [14-16]. Gait velocity was measured as the time taken to walk the middle 8 meters of 10 meters and was timed by a chronometer. The first and last meters, considered respectively warm-up and the deceleration phases, were not included in the calculation. Participants began the GV test on the word "go" and were instructed to "walk at a comfortable and secure pace." Each participant performed the task twice, with the final score being the time in seconds of the quicker of the two timed trials [14,15]. The Tinetti Gait Test is part of the performance-oriented assessment of mobility problems which assesses the following nine components: initiation of gait, step height and length, step symmetry and continuity, path deviation, trunk stability, walking stance, and turning while walking [17]. Each component was scored as 1 (normal) or 0 (abnormal) providing a final score which ranged from 0 to 12, with a higher score indicating a better gait performance. Under the domain of neurally-mediated cardiovascular falls we included orthostatic and postprandial hypotension, vasovagal syndrome and carotid sinus hypersensitivity [10,11] Orthostatic hypotension was defined as a 20-mmHg fall in systolic, or 10-mmHg fall in diastolic BP during 3 minutes of active standing, whereas vasovagal syncope was defined when a head-up tilt test induced hypotension with or without bradycardia-asystole [10]. Carotid sinus hypersensitivity was defined as more than 3 seconds asystole (cardioinhibitory type), more than 50 mmHg fall in systolic BP (vasodepressor type), or both (mixed type) during carotid massage [9,10]. Postprandial hypotension was defined as hypotension (20-mmHg fall in systolic) within 2 hours of the start of a meal documented by 24 hour ambulatory blood pressure monitoring [10,11]. Categorization after assessments Each participant was classified as having an intrinsic or extrinsic fall using the information obtained at the fall assessment interview. Extrinsic falls were related to environmental hazards (slip, trip or externally induced displacement), whereas intrinsic falls were related to mobility or balance disorder, muscle weakness, orthopedic problems, loss of consciousness, neurally mediated cardiovascular disorder or sensory impairment [4]. According to the performance in gait velocity test and Tinetti Gait Test, each participant was classified as having a gait disorder or not. After a complete evaluation, an attributable etiology of the fall was determined. Following previous categorization, the attributable etiologies were grouped into four major categories including the neuromuscular, sensory, orthopedic and cardiovascular domains [7] (Table 1). Table 1 Attributable etiology after fall assessment (n = 41) Domain Principal cause Number of patients Neuromuscular Parkinsonism syndrome 3 Unexplained Ataxia 2 Lower extremity weakness 7 Sensory Dizziness or vertigo 2 Visual impairment 4 Peripheral neuropathy 3 Orthopedics Osteoarthritis (including foot problems) 8 Hip problems or deformity 6 CV causes* Orthostatic hypotension 2 Postprandial hypotension 1 Vasovagal syndrome 2 Carotid sinus hypersensitivity 1 CV causes: neurally mediated cardiovascular cause Statistical analysis Associations between gait disorder and cardiovascular mediated diagnosis of falls were assessed using two-tailed fisher's exact test. A p-value ≤ 0.01 (two-sided) was considered statistically significant. Comparisons between intrinsic-extrinsic fall classification and gait assessment classification (presence or not of gait disorder) for diagnosing an attributable cardiovascular mediated cause were calculated with Chi-square test. Finally, sensitivity and specificity of each classification to detect a non-cardiovascular cause was evaluated. All calculations were performed with STATA 6.0 statistical package, Stata Corporation, College Station, Texas. Results Forty-one consecutive patients with recurrent and unexplained falls were included. Average age was 79.63, SD ± 4 (range 70 to 95) and 73 percent were females. All subjects were ambulatory community dwelling older persons. Activities of daily living (ADLs) were normal for 85 percent of the subjects and mini mental status exam (MMSE) was normal in 78 percent (MMSE>26). The majority of the subjects could not describe a clear cause of their falls and all denied clinical characteristics compatible with syncope. Although six participants provided a clear description of the fall episode, it was insufficient for attributing an etiology to the fall. Average number of falls was 5 in the previous six months, with a range of 3 to 8. After a complete clinical and laboratory assessment, an etiological diagnosis was attributed to every participant as shown in Table 1. A neurally mediated cardiovascular disorder (orthostatic or postprandial hypotension, vasovagal syncope or carotid sinus hypersensitivity) was identified in 6 of the 41 patients (14%) as the main cause of their falls. Two patients were diagnosed with vasovagal syncope, two with postural hypotension, one with postprandial hypotension and one with carotid sinus hypersensitivity. Of the 35 patients with a gait disorder, 34 were diagnosed with a non-cardiovascular cause as the etiological diagnosis. By contrast, of the 6 patients without a gait disorder, 5 were diagnosed with a neurally mediated cardiovascular cause of fall (Table 2) (p < 0.001, fisher exact test). Distribution of the patients according to cardiovascular cause with both traditional and gait assessment classification is shown in Figure 1. Sensitivity and specificity of the presence of a gait disorder for diagnosing a non-cardiovascular cause was 97.1% (CI95% = 85–99) and 83% (CI95% = 36–99), respectively. Sensitivity and specificity of intrinsic-extrinsic classification for diagnosing a non-cardiovascular mediated cause was 28% (CI95% = 14–46) and 83% (CI95% = 36–99) respectively (Table 2). Table 2 Diagnostic accuracy of gait disorder and intrinsic-extrinsic classification for detecting non-cardiovascular mediated falls (n = 41) Diagnostic Approach Patients with Non-CV Causes Patients with CV Causes† Sensitivity (%) Specificity (%) p-value Fisher exact test Gait Disorder + 34 1 97.1 83 <0.001‡ Gait Disorder - 1 5 Extrinsic Fall 10 1 28 83 0.9128‡ Intrinsic Fall 25 5 Non-CV cause: non-cardiovascular neurally mediated cause. †CV causes: cardiovascular neurally mediated cause. Gait Disorder +: presence of gait disorder. Gait Disorder -: absence of gait disorder. ‡p-value of the associations with non-cardiovascular mediated falls for each diagnostic approach. Note: Neurally mediated cardiovascular causes are the following: orthostatic hypotension, postprandial hypotension, vasovagal syncope and carotid sinus hypersensitivity. Figure 1 Distribution of fallers divided in cardiovascular mediated causes or not according to extrinsic-intrinsic classification (A) or gait disorders presence (B). Discussion In this sample of ambulatory patients referred with unexplained falls, the presence of a gait disorder was significantly associated with a non-cardiovascular cause of falls. Of the 35 participants with a gait disorder, 34 had a non-cardiovascular mediated fall. By contrast, 5 out of 6 participants with a cardiovascular mediated cause had a normal gait performance. This interesting observation suggests that when gait disorders are not detected in patients with unexplained falls, a cardiovascular mediated origin should be seriously considered. Although intrinsic-extrinsic categorization was intended to separate and identify multiple contributors to the fall, this classification provides limited clinical help since older people who experience an extrinsic fall often have an underlying intrinsic condition that decreases their ability to compensate for the hazardous situation [7]. Rather, falls are often related to a complex interaction among intrinsic and extrinsic factors which challenges the postural control and the ability to maintain an upright position. In our study, the intrinsic-extrinsic categorization did not provide additional information for detecting a cardiovascular mediated cause. In fact, when these 41 participants with unexplained falls were analyzed under this categorization, we found a similar pattern of distribution regarding cardiovascular cause with a high non-cardiovascular/cardiovascular cause ratio in both groups (Figure 1A). In contrast, when using the gait assessment approach, the distribution showed almost no gait disorder when falls were associated with a cardiovascular origin with an inverse ratio. (Fig 1B). For instance, 5 out of 6 patients with the attributable diagnosis of a cardiovascular cause had a normal performance in the Tinneti Gait Test and a gait velocity above 0.8 m/s. Why should gait assessment be considered a key component in the assessment of unexplained falls? Gait performance is a fascinating and complex task that depends on the normal functioning of multiple systems working in a highly coordinated and integrated manner [14,16]. As impairments in different domains can alter this delicate system, it is interesting to hypothesize that different chronic conditions such as visual or hearing problems, muscular weakness, osteoarthritis or peripheral neuropathy could be evidenced through gait performance. Moreover, the use of benzodiazepines, neuroleptics, and other drugs with central action may affect gait performance. Chronic cardiovascular diseases and systemic chronic hypertension have been also associated with continuous impairment in gait performance [18,19]. However, falls secondary to neurally-mediated cardiovascular causes may be expressed by a different mechanism, without necessarily chronically affecting gait performance [10,15]. Regulation of systemic blood pressure is important for postural control in elderly people since failure to perfuse the brain increases the risk of falls. Additionally, age-related decline in baroreflex sensitivity contributes to the vulnerability in changing posture or in the postprandial state [20,21]. Although the exact mechanism by which a neurally mediated cardiovascular problem causes a fall remains unclear, there is growing clinical evidence for its association with unexplained falls [22]. Interestingly, participants in our study who were diagnosed with a cardiovascular cause of falls denied the clinical symptoms traditionally associated with a syncopal phenomenon. This observation agrees with the increasingly detected overlap between syncope and falls in the subgroup of unexplained elderly fallers [9]. A possible explanation may be the incapacity to remember syncopal symptoms due to a retrograde amnesia of the episode [9,10,13]. Since a cardiovascular mediated cause of fall has been described in up to 20 percent of unexplained elderly fallers, and in view of the associated morbidity, mortality and the availability of treatment, careful consideration should be given when assessing patients with unexplained falls [9,10]. As well, falls related to a neurally mediated cardiovascular event could be expressed acutely and intermittently without chronically affecting gait performance, providing a potential explanation for the absence of gait disorders in these participants. As gait performance can be assessed directly, thus avoiding reporting bias, this approach could be applied even in absence of a detailed description of the episode. Finally, the simplicity of the proposed gait assessment makes it easy to perform and accessible for general clinicians and other health professionals. Our study has the strength of a systematic and comprehensive assessment of the participants who were diagnosed with a potential etiology of falls; however, important limitations should be outlined. The sample studied comprised a select and relatively small number of participants with a low proportion of them with cardiovascular mediated falls, facts which affect the generalizability of the findings. Due to the design characteristics, causality was not pursued and only an association between the cardiovascular diagnosis and the fall episode can be described. A prospective and controlled trial in a larger sample is needed to strengthen our hypothesis. Conclusion In ambulatory older persons with unexplained falls, the presence of a gait disorder was associated with non-cardiovascular mediated falls. Gait assessment was useful for the detection of non-cardiovascular causes of falls, providing an additional diagnostic yield to this assessment. This clinical based approach may help on the evaluation of elderly persons with unexplained falls. Authors' contributions MMO and MS collected the clinical data. All six authors participated in the design of the study. MMO and ERS did the analyses. MMO, ERS and GD wrote the first draft of the manuscript. All of the authors reviewed the analyses, and read and approved the final manuscript. The authors declare that they have no competing interests. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements We are indebted for the critical review of the manuscript and comments from Dr. Howard Bergman, McGill University Division of Geriatric Medicine. Our gratitude to Mrs. Rebecca Rupp for her kind help in correcting the manuscript and Ms. Nadia Sourial for her useful comments on statistical analysis, both from Solidage Research Group (McGill University and Université de Montréal Research Group on Integrated Services for Older Persons). Dr. Montero-Odasso holds a Clinical and Research Fellowship Award from McGill Division of Geriatric Medicine and Maimonides Geriatric Centre (McGill University, Montreal). Hospital Italiano de Buenos Aires has funded the study. This paper was presented in part at the 2004 American Geriatrics Society Scientific Meeting, Las Vegas, Nevada, USA. ==== Refs Tinetti ME Where is the vision for fall prevention? J Am Geriatr Soc 2001 49 676 677 11380766 10.1046/j.1532-5415.2001.49132.x Masud T Morris RO Epidemiology of falls Age Ageing 2001 30 3 7 11769786 10.1093/ageing/30.3.255 Wild D Nayak US Isaacs B Description, classification and prevention of falls in old people at home Rheumatol Rehabil 1981 20 153 159 7280490 Lach HW Reed AT Arfken CL Miller JP Paige GD Birge SJ Peck WA Falls in the elderly: reliability of a classification system J Am Geriatr Soc 1991 39 197 202 1991951 Tinetti ME Speecheleyn M Ginter SF Risk factors for falls among elderly persons living in community N Eng J Med 1988 319 1701 1707 Guideline for the prevention of falls in older persons. American Geriatrics Society, British Geriatrics Society, and American Academy of Orthopedic Surgeons panel on falls prevention J Am Geriatr Soc 2001 49 664 772 11380764 10.1046/j.1532-5415.2001.49115.x Lipsitz LA Jonson PV Kelly MM Koestler JS Causes and correlates of recurrent falls in ambulatory frail elderly J Gerontol 1991 46 M114 122 2071832 Shaw FE Kenny RA The overlap between syncope and falls in the elderly Postgrad Med J 1997 73 635 639 9497972 McIntosh SJ Lawson J Kenny RA Clinical characteristics of vasodepressor, cardioinhibitory, and mixed carotid sinus syndrome in the elderly Am J Med 1993 95 203 208 8356984 10.1016/0002-9343(93)90261-M Carey BJ Potter JF Cardiovascular causes of falls Age Ageing 2001 30 19 24 11769783 10.1093/ageing/30.1.19 Jansen RW Lipsitz LA Postprandial hypotension: epidemiology, pathophysiology, and clinical management Ann Intern Med 1995 122 286 295 7825766 Tinetti ME Preventing falls in elderly persons N Engl J Med 2003 348 42 49 12510042 10.1056/NEJMcp020719 Cummings SR Nevitt MC Kidd S Forgetting falls: the limited accuracy of recall of falls in the elderly J Am Geriatr Soc 1988 36 613 616 3385114 Montero-Odasso M Schapira M Varela C Pitteri C Soriano ER Kaplan R Camera LA Mayorga LM Gait velocity in senior people as an easy test for detecting mobility impairment in community elderly J Nutr Health Aging 2004 8 340 343 15359349 Montero-Odasso M Schapira M Soriano ER Varela M Kaplan R Camera LA Mayorga LM Gait velocity as a single predictor of adverse events in healthy seniors aged 75 and older J Gerontol A Biol Sci Med Sci 2005 60 1304 1309 16282564 Alexander N Gait Disorders in older adults J Am Geriatr Soc 1996 44 434 451 8636592 Tinetti ME Performance-oriented assessment of mobility problems in elderly patients J Am Geriatr Soc 1986 34 119 126 3944402 Hausdorff JM Herman T Baltadjieva R Gurevich T Giladi N Balance and gait in older adults with systemic hypertension Am J Cardiol 2003 91 643 645 12615286 10.1016/S0002-9149(02)03332-5 Bloem J Gussekloo AM Lagaay Remarque EJ Haan J Westendorp RG Idiopathic senile gait disorders are signs of subclinical disease J Am Geriatr Soc 2000 48 1098 1101 10983910 Lipsitz LA Orthostatic hypotension in the elderly N Engl J Med 1989 321 952 957 2674714 Maurer MS Karmally W Rivadeneira H Parides MK Bloomfield DM Upright posture and postprandial hypotension in elderly persons Ann Intern Med 2000 133 533 536 11015166 Parry SW Kenny RA Drop attacks in older adults: Systematic assessment has a high diagnostic yield J Am Geriatr Soc 2005 53 74 78 15667379 10.1111/j.1532-5415.2005.53013.x	16321159	PMC1325027	CC BY	2021-01-04 16:30:33	no	BMC Geriatr. 2005 Dec 1; 5:15	utf-8	BMC Geriatr	2,005	10.1186/1471-2318-5-15	oa_comm
==== Front BMC Cell BiolBMC Cell Biology1471-2121BioMed Central London 1471-2121-6-431633664010.1186/1471-2121-6-43Research ArticleA novel Dictyostelium RasGEF required for chemotaxis and development Arigoni Maddalena [email protected] Enrico [email protected] Daniel F [email protected] Helmut [email protected] Gerald [email protected] Salvatore [email protected] Department of Clinical and Biological Sciences, University of Torino, Regione Gonzole 10, 10043 Orbassano, Italy2 Faculty of Biology, University of Konstanz, 78457 Konstanz, Germany3 Dept. Microbiology and Immunology, University of British Columbia, Canada V6T1Z32005 7 12 2005 6 43 43 16 6 2005 7 12 2005 Copyright © 2005 Arigoni et al; licensee BioMed Central Ltd.2005Arigoni 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 Ras proteins are guanine-nucleotide-binding enzymes that couple cell surface receptors to intracellular signaling pathways controlling cell proliferation and differentiation, both in lower and higher eukaryotes. They act as molecular switches by cycling between active GTP and inactive GDP-bound states, through the action of two classes of regulatory proteins: a) guanine nucleotide exchange factor (GEFs) and b) GTP-ase activating proteins (GAPs). Genome wide analysis of the lower eukaryote Dictyostelium discoideum revealed a surprisingly large number of Ras Guanine Nucleotide Exchange Factors (RasGEFs). RasGEFs promote the activation of Ras proteins by catalyzing the exchange of GDP for GTP, thus conferring to RasGEFs the role of main activator of Ras proteins. Up to date only four RasGEFs, which are all non-redundant either for growth or development, have been characterized in Dictyostelium. We report here the identification and characterization of a fifth non-redundant GEF, RasGEFM. Results RasGEFM is a multi-domain protein containing six poly-proline stretches, a DEP, RasGEFN and RasGEF catalytic domain. The rasGEFM gene is differentially expressed during growth and development. Inactivation of the gene results in cells that form small, flat aggregates and fail to develop further. Expression of genes required for aggregation is delayed. Chemotaxis towards cAMP is impaired in the mutant, due to inability to inhibit lateral pseudopods. Endogenous cAMP accumulates during early development to a much lower extent than in wild type cells. Adenylyl cyclase activation in response to cAMP pulses is strongly reduced, by contrast guanylyl cyclase is stimulated to higher levels than in the wild type. The actin polymerization response to cAMP is also altered in the mutant. Cyclic AMP pulsing for several hours partially rescues the mutant. In vitro experiments suggest that RasGEFM acts downstream of the cAMP receptor but upstream of the G protein. Conclusion The data indicate that RasGEFM is involved in the establishment of the cAMP relay system. We propose that RasGEFM is a component of a Ras regulated pathway, which integrate signals acting as positive regulator for adenylyl cyclase and negative regulator for guanylyl cyclase. Altered guanylyl cyclase, combined with defective regulation of actin polymerization, results in altered chemotaxis. ==== Body Background The ras proto-oncogenes encode membrane-bound small monomeric GTP-binding proteins with molecular masses ranging between 20 to 40 kDa, which are highly conserved in the course of eukaryotic evolution. Ras proteins control fundamental cell processes including proliferation, differentiation, motility and polarity [1-3]. Like heterotrimeric G proteins they act as molecular switches cycling between two interconvertible forms: inactive, when bound to guanosine diphosphate (GDP), and active, when bound to guanosine triphosphate (GTP) [4]. Conversion between GTP and GDP-bound states is tightly regulated by two set of proteins: guanosine-nucleotide exchange factors (GEFs), and GTPase activating proteins (GAPs). GEF proteins cause activation by catalysing the exchange of bound GDP with GTP, whereas GAPs inactivate Ras by increasing their rate of GTP hydrolysis [5]. Systems composed by GTPases, GAPs and GEFs allow great versatility in the construction of signaling pathways. Signals can be amplified (one GEF could activate several GTPases), integrated (several pathways activate the same GEF and GAP, and the behaviour of one GTPase depends on the total effect of all its GEFs and GAPs), or split (one GTPase induces many effects) [6]. This versatility allows small GTPases to mediate a wide range of different biological functions among different organisms. For example in S. cerevisiae RasGEF CDC25 is required for Ras mediated activation of adenylyl cyclase and it is essential for proliferation and spore germination [7], whereas in Drosophila, the RasGEF Son-of-sevenless (Sos) functions upstream of Ras in R7 photoreceptor differentiation [8]. Despite its relatively small genome, Dictyostelium possesses a relatively large number of ras and rasGEF encoding genes. Dictyostelium is a lower eukaryote with a simple life cycle, consisting of growth and multicellular development, the latter being fully completed in approximately 24 hours. The amoebae live as single cells, growing by feeding on bacteria, which are taken up by phagocytosis, and dividing by binary fission. Upon starvation, cells start releasing the chemoattractant cAMP and gather by chemotaxis to form multicellular aggregates. Within each aggregate, cells differentiate into prespore and several classes of prestalk cells, while undergoing a series of morphogenetic changes, which end up in the formation of fruiting bodies. Fruiting bodies consist of slender stalks of vacuolated, dead cells, bearing on top spores encapsulated in sori. Here we report the identification and characterization of a novel RasGEF, named RasGEFM. To date only four Dictyostelium RasGEFs have been characterized, namely rasGEFA, formerly known as aimless [9], rasGEFB [10], gbpC and gbpD [11]. The latter two encode for unconventional RasGEFs. GbpC possesses a RasGEF domain coupled to a cyclic nucleotide-binding domain, which is associated with a MAPKKK-like kinase domain, Leucine Rich Repeats (LRR) and a Ras domain. GbpD is highly similar to GbpC, but it lacks the Ras, MAPKKK-like and the LRR domains [11]. These two proteins control myosin phosphorylation and, as consequence, cell motility and chemotaxis [12]. The GEF protein encoded by the rasGEFA gene is essential for cell aggregation, acting at the level of adenylyl cyclase activation [9]. Dictyostelium cells lacking rasGEFB are defective in early development, although they eventually form tiny but normally proportioned fruiting bodies. Furthermore, these cells move unusually rapidly and show severe impairment in cell growth [10]. Here we present evidences that RasGEFM is involved in a Ras regulatory network, required for cAMP receptor dependent signal transduction. Mutant cells lacking the rasGEFM gene produce very small, flat aggregates and fail to develop further. Chemotaxis is altered in the mutant, due to inability of the cells to polarize properly. The phenotype can be partially rescued by pulsing cells with cAMP. We show that RasGEFM is involved in controlling cAMP relay and cell motility. Results Identification, cloning and sequence analysis of D.d.rasGEFM To identify Ras regulator proteins in the Dictyostelium genome a bioinformatic approach was taken, based on a tBLASTn algorithm, to search the Dictyostelium genome project databases [13] and [14]. Using known RasGEFs domains as query, we identified approximately 30 genes encoding for proteins with significant homology to putative RasGEF proteins (unpublished results). Among the putative RasGEF encoding genes, one of these, designated as Dd rasGEFM, was isolated for functional studies. The rasGEFM gene, located on the chromosome 2, is organized in 4 exons, which are interrupted by 3 introns. Southern blot analyses performed under high and low stringency conditions, and extensive analysis of the Dictyostelium genomic sequence databases, indicated that the gene is present as single locus (data not shown). The rasGEFM gene encodes a relatively large multi-domain protein of 929 amino acids, with a calculated molecular mass of approximately 102 kDa, whose domain structure is displayed in Fig. 1. The amino-terminal region of the protein contains six short poly-proline stretches, which are putative consensus docking site for SH3 or WW domains [15] whereas the carboxy-terminal region possesses the putative RasGEF catalytic domain. BLAST analysis shows that the catalytic domain, with evolutionary identical amino acid residues conserved, shares high homology with the murine p140GRF (30% identity, 59% homology). Among Dictyostelium RasGEF proteins, the most similar to RasGEFM are two putative uncharacterized RasGEFs (RasGEFE and RasGEFJ) (Fig. 1). Between the catalytic domain and the amino-terminal poly-proline rich region, there are a DEP and a RasGEF-N terminal domain. The RasGEF-N terminal module is peculiar only for Ras specific GEFs, and it is likely to have a purely structural role [16]. The DEP (Dishevelled-Eglin-Pleckstrin homology) domain, located between the catalytic and the RasGEF-N terminal domain, shows significant homology to the DEP domain of the human pleckstrin 25% overall homology and 11% identity, versus a 27% overall homology and 14% identity displayed by the Dictyostelium GbpC DEP [11] when compared to the human plekstrin counterpart. DEP is a widespread motif found in proteins involved in Wnt signalling, in regulators of G protein signalling (RGS), in pleckstrin and other signalling proteins [17,18], responsible either for targeting proteins to the membrane or mediating protein-protein interaction, although the underlying molecular mechanism remains still unknown. Figure 1 Schematic representation of the different functional domains of RasGEFM identified with MotifScan. The RasGEFM protein contains the following recognizable modules: 6 proline rich regions, a RasGEF N-terminal domain, a DEP domain and a RasGEF catalytic domain. The sequence of the RasGEFM catalytic domain (aminoacids 673–863) has been aligned, using the ClustaIW program, with representative RasGEFs proteins from different organisms and others putative Dictyostelium RasGEFs. Identical and conserved aminoacid residues are boxed in dark or light gray respectively. Accession numbers are referred to GeneBank: [AAN46882 D.d.GEF M, AAN46874 D.d.GEF E, AAN46879 D.d.GEF J, A41216 D.m.Sos, P28818 R.n.p140, A38985 H.s. CDC25] rasGEFM gene expression is developmentally regulated and partially controlled by the G protein One of the prominent features of Dictyostelium life cycle is the transition from solitary amoebae to multicellular aggregates. This transition is triggered by starvation of the cells, is enhanced by periodic release of cAMP, and results in the coordinated activation and repression of aggregation-specific and growth-phase genes, respectively [19]. We examined the expression pattern of the rasGEFM gene during development of wild-type cells and in three mutant strains, which are blocked at sequential steps of development. Two transcripts were detected in Northern blots: the upper one is present during growth and at the beginning of development and disappears at 6 hours of starvation. The lower transcript is barely visible at the beginning of starvation, reaches its maximal expression at 6 hours of development and decreases thereafter, though being present up to the end of development (Fig. 2A). Figure 2 Northern blot analysis of rasGEFM expression in parental and mutant strains. (A) Total RNA was extracted from AX2 cells developed in suspension for 0 to 9 hours or on filters. In the latter case the cells were harvested at mound (12 hours), first finger (16 hours) and preculminant (20 hours) stages. The membrane was hybridized to a radiolabelled rasGEFM specific probe (probe a) corresponding to bp 760–1518 of the cDNA clone and to the actin gene used as a loading control. (B) Total RNA was extracted from different developmental Dictyostelium mutants starved in shaking suspension up to 6 hours. LW6 (G protein β subunit minus), HSB1 (PIA ts-mutant, defective in the G-protein adenylyl cyclase activation) and HSB50 (mutant blocked at mound stage). We examined rasGEFM expression in the LW6 mutant lacking the β subunit of heterotrimeric G protein [20]. Although two Gβ genes are present in the Dictyostelium genome, disruption of one of them blocks all G protein-dependent pathways in LW6 mutant cells, which fail to respond to and to relay cAMP, to express aggregation specific genes and to develop. In this strain only the rasGEFM upper transcript is detected during starvation, while the lower transcript fails to accumulate (Fig. 2B). This indicates that up-regulation of the lower transcript is under control of the heterotrimeric G protein. Gene expression was also analyzed in the HSB1 and HSB50 mutants. HSB1 is defective in cAMP relay, due to a temperature-sensitive mutation in the adapter protein PIA, which is essential for adenylyl cyclase activation [21,22]. HSB1 cells thus fail to aggregate but, in contrast to LW6, express at moderate level aggregation-specific genes and are sensitive to exogenous cAMP pulses. The HSB50 mutant undergoes chemotaxis and aggregation, but is blocked at mound stage [23]. In both mutants, the rasGEFM expression pattern was similar to that observed in the parental strain (Fig. 2B), suggesting that transcription of the lower transcript does not require cAMP relay, even though it may be enhanced by cAMP pulses, similarly to other aggregation-specific genes [24]. It is worth mentioning that the two transcripts may arise from differential splicing or different degree of polyadenylation. Screening 24 cDNA clones obtained by using mRNA from 6-h starved cells resulted in a signle sized cDNA to be present. Sequencing three such clones gave rise to a transcript containing all four exons. Assuming this transcript to correspond to the most abundant mRNA species at this time point, namely the lower band, then we should conclude that differential splicing is unlikely and the upper band is the result of extensive polyadenylation. Additional experiments are required to confirm this hypothesis and to understand this intriguing developmentally regulated changes in rasGEFM gene expression. rasGEFM null mutant is delayed in the acquisition of aggregation competence and blocked at the aggregation stage To gain insight into the function of rasGEFM, the gene was inactivated by homologous recombination. Disruption was confirmed by Southern and Northern blot analysis of the mutant strain, named HSB61 (Fig. 3). HSB61 cells grew normally both in axenic medium, or on bacterial lawn, but were defective in development. In contrast to the parental cells (Fig. 4A), which form fruiting bodies after about 24 hours of starvation, HSB61 cells formed small, flattened aggregates, unable to develop into tipped tight mounds (Fig. 4B). Aggregate formation was density-dependent. If cells were plated on non-nutrient agar at concentration of 1 or 2 × 105/cm2 no aggregates were formed, whereas control cells formed aggregates. At 1 × 106 cells per cm2 the mutant cells formed small aggregates, similar to those shown in Fig. 4B, but many cells failed to aggregates. Treating cells with cAMP pulses rescued in part the mutant phenotype; if aggregates, formed in suspension after 6–8 hours pulses, were transferred on agar, they continued to develop and to form small fruiting bodies, although many cells failed to aggregate (Fig. 4C). We have failed to observe cells forming streams, even in 6 to 8 hours pulsed cells, although elongated cells very near to aggregates are found, suggesting that some short-range chemotaxis may occur. Figure 3 Disruption of the rasGEFM gene. (A) Southern blot of genomic DNA from parental strain (AX2) and rasGEFM null mutant (HSB61). Genomic DNA was digested with EcoRI, separated in 0.8% agarose gel, blotted onto nylon membrane and probed with probe a (rasGEFM N-terminal fragment corresponding to bp 0–617 of the cDNA clone), probe b (corresponding to bp 760–1518 of the rasGEFM cDNA clone) and probe c (bsr cassette). Three different probes were used to characterize the genomic locus of the mutant and the parental strain. In the rasGEFM null strain the central part of the gene (recognized by probe b), was replaced by the blasticidin cassette (recognized by probe c), which has the same size of the replaced fragment. Because of that, the size of the locus remains unchanged, but the central part is recognized specifically by bsr (probe c) or probe b in HSB61 or AX2, respectively. Both AX2 and HSB61 genomic loci are recognized by the probe a. (B) Northern blot of total RNA extracted from HSB61 cells at the indicated time points of growth and development. The membrane was hybridized to the rasGEFM and to the actin gene. Figure 4 Developmental phenotypes of (A) wild-type or (B, C) rasGEFM null cells. (A, B) AX2 or HSB61 cells were plated at the beginning of starvation at a concentration of 1 × 107cell/ml on non-nutrient agar (approx. 6.5 × 105 cell/cm2). (C) HSB61 cells were pulsed with cAMP for 10 hours before plated on agar. The final phenotype after 24 hours is shown. The finding that HSB61 formed flat aggregates prompted us to test whether the mutant was defective in cell-cell adhesion or chemotaxis. Intercellular adhesion in Dictyostelium is developmentally regulated, as growth-phase cells are weakly adhesive, and this adhesion is completely blocked by EDTA. During the first 5–6 hours of development, cells express the adhesion glycoprotein csA on the cell surface, which is responsible for an EDTA-resistant form of adhesion (for a review see [25]), and at least in part, for post-aggregative pattern formation [26]. We tested the ability of HSB61 cells to develop EDTA-resistant adhesion over the first 8 hours of development. As shown in Fig. 5, HSB61 cells exhibited a delay of 5 to 6 hours in the appearance of EDTA-resistant adhesiveness compared to the parental strain. Cell treatment with cAMP pulses, which are known to stimulate expression of csA as well as several other aggregation-specific genes [19], accelerated acquisition of EDTA-stable adhesion, though in contrast to the parental strain the mutant cells appeared to be refractory to the pulses for at least 4 hours (Fig. 5). Figure 5 Appearance of EDTA-stable contacts in (open symbols) untreated and (closed symbols) cAMP treated cells. Cells of (squares) AX2 or (triangles) HSB61 were developed in suspension. At the developmental time indicated in the abscissa cells were taken, incubated with 10 mM EDTA and cell adhesion measured in the agglutinometer of Beug and Gerisch, as described in Methods. A representative experiment is shown. To examine whether rasGEFM null cells were impaired in their ability to chemotax towards cAMP, a micropipette-based assay was used. HSB61 cells resulted strongly impaired in chemotaxis, as will be described further below, but their behaviour improved significantly when pulsed with cAMP. Taken together these results suggest that HSB61 cells fail to aggregate properly, due to impaired periodic cAMP signalling, which is required for optimal expression of aggregation-stage specific genes. To confirm this hypothesis, we followed the expression of three such genes, namely carA, acaA and csA. CarA and acaA encode the cAMP receptor cAR1 and adenylyl cyclase A, respectively, whereas csA is, as mentioned above, the gene for the cell adhesion glycoprotein csA. Expression of acaA, csA, and to a lesser extent carA, was deranged in the mutant compared to the parental strain. The peak of expression for all three genes was reached in the parental strain between 4 and 6 hours of development followed by down-regulation as cells undergo aggregation. Pulsing with cAMP accelerated the kinetics of their expression, leading to a higher mRNA accumulation rate (Fig. 6 upper panel). In the rasGEFM mutant, mRNA expression was delayed and down-regulation was not detected even after 12 hours starvation. Cyclic AMP pulses elicited a stimulatory effect, though not as efficient as in the parental strain (Fig. 6 bottom panel). Figure 6 Early developmental gene expression in AX2 and HSB61 cells. Total RNA was extracted from cells pulsed (+ cAMP) or not (- cAMP) in suspension for the time indicated. After electrophoresis and transfer, the membranes were hybridized with radiolabelled acaA, carA, csA. Actin was used as control. Cyclic AMP receptor and G protein dependent activation of adenylyl and guanylyl cyclases is altered in rasGEFM null cells The finding that the HSB61 mutant was delayed in the acquisition of the aggregation competence but partially responded to cAMP pulses suggested that the RasGEFM protein might be involved in cAMP receptor-mediated signaling. We therefore monitored cAMP accumulation during the early hours of development and tested adenylyl and guanylyl cyclase activities in vitro. In AX2 cells, the basal cAMP is low at the beginning of development and rises sharply at around 4 hours of development, reaching a maximum by around 7–10 hours, after which it decreases (Fig. 7A) [19,24]. In HSB61 cells, cAMP started to accumulate with a delay of 2 hours and increased at a very low rate, reaching at 10 hours of starvation less than half the maximal concentration of the wild type (Fig. 7A). Cell treatment with cAMP pulses accelerated of two hours endogenous cAMP accumulation both in wild type and the mutant, restoring almost normal levels of cAMP in the latter (Fig. 7B). Figure 7 Cyclic AMP accumulation in AX2 and HSB61 during development in shaken suspension. (Squares) AX2 or (triangles) HSB61 cells were incubated in suspension for the time indicated in the abscissa. (A) control cells or (B) cells treated with cAMP pulses every 6 minutes. At the time indicated, cell aliquots were quenched with perchloric acid, neutralised with KOH, and total cAMP in the samples was determined by radioimmunoassay as described in Material and Methods. Cyclic AMP accumulation in the pre-aggregative and aggregation stage results from the periodic activation of adenylyl cyclase, which is stimulated in autocrine and paracrine loops by cAMP. This leads to both increased accumulation of the enzyme and oscillatory stimulation of its activity, with a period of about 6 minutes [19]. Exogenously supplied cAMP pulses mimic the endogenous oscillations of cAMP and give insight on potential defects in cAMP relay or other cAMP induced responses in mutant cells. We investigated cAMP and cGMP changes in HSB61 cells during a period of cAMP pulses. As shown in Fig. 8A, in response to a cAMP pulse, 5 hours starved and cAMP-treated HSB61 cells displayed a dramatically reduced increase in cAMP compared to AX2 cells. In contrast the cGMP peak in the mutant was about twice that observed in the parental strain (Fig. 8B). Remarkably, when cells were assayed after 9 hours pulsing, cGMP peaked 5 to 10 fold higher compared to AX2 cells (Fig. 8C), while the cAMP level increased slightly compared to t5 mutant cells (data not shown). The rasGEFM null mutant, therefore, displays significantly increased cGMP and reduced cAMP responses in the pre-aggregative stage and during aggregation compared to the parental strain. The HSB61 developmental defects are supported by light scattering measurements of cells in suspension, which showed that the mutant cells failed to undergo spontaneous light scattering oscillations. When pulsed with cAMP for several hours, light scattering changes were induced, but the responses were lower than in the wild type (data not shown). Figure 8 Cyclic AMP stimulation of (A) adenylyl and (B, C) guanylyl cyclase activity. (Squares) AX2 or (triangles) HSB61 cells were treated with cAMP pulses for (A, B) 5 or (C) 9 hours. In conjunction with a cAMP pulse (arrows), samples were taken at the time indicated in the abscissa for determining total concentration of (A) cAMP or (B, C) cGMP. Cyclic-nucleotides were measured using the radioimmunoassay kit as described in the section Material and Methods. Representative experiments are shown. The finding of a reduced cAMP response in the mutant prompted us to assay adenylyl cyclase activity in cell lysates upon stimulation of the cells with 2'-deoxy-cAMP or the slowly hydrolyzable GTP analog GTPγS. The assay was done with mutant cells pulsed with cAMP for 7 hours, since after such treatment the mutant cells accumulated cAMP to levels comparable to 5-h treated control cells, as shown in Fig. 7B. GTPγS stimulated adenylyl cyclase activity about 16 and 13 fold in AX2 and HSB61 cells, respectively (Fig. 9). In contrast, stimulation with 2'-deoxy-cAMP increased the level of cAMP 4 fold in the mutant compared to 44 fold for AX2 cells (Fig. 9). The finding that GTPγS stimulated adenylyl cyclase to about the same level in both strains suggests that basal adenylyl cyclase activity is roughly comparable in mutant and parental strain, and this was confirmed by assaying adenylyl cyclase activity in the presence of Mn2+ (Fig. 9). The strongly reduced adenylyl cyclase activation by 2'-deoxy-cAMP in the mutant could be due to a lower number of cAMP receptors expressed on the cell surface. To exclude this we performed cAMP binding experiments in both cell lines AX2 and HSB61 pulsed for 5 and 7 hours, respectively. The total amount of cAMP receptors was comparable, as both maximum amount of ligand bound to the receptor (Bmax), and dissociation constant (Kd) were in the same range in wild type and mutant cells (Fig. 10A). Figure 9 GTPγS and 2'-deoxy-cAMP induced adenylyl cyclase activation. AX2 or HSB61 cells were pulsed with cAMP for 5 or 7 hours, respectively. Cell lysates were prepared in the presence of (grey bars) GTPγS, (black bars) 2'-deoxy-cAMP, (black stripes) MnCl2 and assayed for adenylyl cyclase activity. Plotted values were normalized relative to the (open bars) unstimulated activity obtained in the absence of GTPγS or 2'-deoxy-cAMP (0.7 pmol/mg/min and 0.6 pmol/mg/min for AX2 and HSB61 respectively). Values for AX2 and HSB61 cell lysates are the means ± sd of two indipendent experiments run in duplicate. Figure 10 cAMP binding to the cell surface and its inhibition by GTPγS. (A) Scatchard analysis of cAMP binding in AX2 and HSB61 cells pulsed with cAMP for 5 or 7 hours, respectively. The [3H] cAMP binding was determined over a range of 700–19700 nM cAMP by incubation for 5 min at 0°C. A fitted line, Bmax and Kd are shown in each panel. (B) Inhibitory effect of GTPγS (black bar) on the binding of cAMP to their cognate receptors. Crude membranes were incubated with [3H] cAMP in the absence or presence of GTPγS. Values are indicated as the percentage of cAMP binding in treated normalized to the untreated membranes (white bar). In wild type strain cAMP binding dropped from 10.4 ± 0.3 nM to 5.26 ± 0.04 nM for GTPγS untreated and treated membranes, respectively. HSB61 strain displayed comparable absolute values: 16.23 ± 0.25 and 16.00 ± 0.3 nM for GTPγS untreated and treated membranes, respectively. Means of three independent experiments, each run in duplicate, are shown. These data strengthen the hypothesis that the strongly reduced adenylyl cyclase activation in HSB61 cells is G protein independent, and suggest that Ras GEFM may be located upstream of the heterotrimeric G protein and downstream of the cAMP receptor. If this hypothesis is correct, cAMP binding to membranes upon treatment with GTPγS should be altered in the mutant compared to the wild type, since the affinity of cAMP receptors differs when they are complexed with G proteins [27]. Consistent with this hypothesis, cAMP binding to HSB61 membranes was not affected by GTPγS, whereas it was inhibited cAMP binding approximately 50% in wild type membranes (Fig. 10B). Taken together, these results lead us to conclude that Ras GEFM very likely acts between the cAMP receptor and the heterotrimeric G protein. RasGEFM is not the activator of RasC or RasG Two previously characterized Ras proteins, RasC and RasG, have been shown to be involved in cAMP mediated signalling events [28-30]. Therefore, it was tempting to hypothesize that RasGEFM could function as the putative exchange factor for either of these Ras proteins. Activated Ras can be measured by using a Ras Binding Domain (RBD) to affinity purify Ras proteins. Given that the RBD-Ras interaction is dependent on Ras being in a GTP bound form, one can selectively measure activated Ras from cellular lysates. An assay has recently been described for Dictyostelium Ras proteins employing the RBD from S. pombe Byr2, and it has been shown that both RasC and RasG are transiently activated by cAMP [30]. To determine whether RasGEFM mediates this activation, we challenged 6 hours pulsed AX2 and HSB61 cells with cAMP and observed the kinetics of Ras activation (Fig. 11). While the levels of both activated RasC and RasG increased upon cAMP stimulation in HSB61, the maximum level of RasG-GTP was greatly reduced relative to AX2, whereas RasC activation was largely unaffected in the mutant. This suggested the possibility that RasGEFM may be partly responsible for mediating the activation of RasG. However, when 10-hour pulsed cells were stimulated with cAMP, RasG activation was restored. Interestingly, the total level of RasG remained at a very high level relative to AX2. This data shows that RasGEFM is unlikely to be an activator of RasC or RasG, and the partial loss of RasG activation seen in 6 hours cells may be a secondary effect of reduced gene expression. In addition, the finding that RasG is activated after prolonged cAMP treatment favours the notion that a developmentally regulated component, regulated by cAMP signalling, is directly or indirectly required for RasG activation. Figure 11 RasC and RasG activation in HSB61 mutant. AX2 or HSB61 cells were pulsed with cAMP for the time indicated, concentrated ten times, treated with cAMP and immediately lysed. The lysates were incubated with GST-Byr2 (RBD) as described in Material and Methods. The precipitate was subjected to electrophoresis and Western blot and hybridized with antibodies against RasC or RasG. Time (in seconds) after cAMP treatment is indicated. Total RasC and RasG in cell lysates from AX2 or HSB61 are shown on the right. Ca2+ influx in rasGEFM null mutant is reduced In addition to the effects on cAMP and cGMP levels, the second messenger Ca2+ is rapidly regulated by cAMP binding to the receptor [31]. The finding that chemotaxis and cGMP response were altered in the mutant, prompted us to test chemoattractant-induced Ca2+ entry in HSB61 cells. Ca2+ entry is negatively regulated by cGMP [32], and both Ca2+ and cGMP have been implicated in regulating chemotactic motility [33]. The basal levels of intracellular Ca2+ concentration were comparable in parental and mutant cells and elevation in response to a cAMP pulse was only slightly reduced in the mutant (data not shown). Cyclic AMP-induced Ca2+influx depends on the extracellular Ca2+concentration and on the dose of the cAMP stimulus. The influx is increased with elevation of the concentration of both parameters [34]. Ca2+ influx is maximal in AX2 cells after 4 to 6 hours of development and remains constant thereafter [34]. We determined maximal Ca2+ influx in unpulsed and pulsed HSB61 cells. As shown in the saturation curve of Fig. 12A, left, maximal Ca2+ influx was reduced to about 50% of the control in unpulsed HSB61 cells, but influx was restored when the mutant cells were pulsed with cAMP (Fig. 12A, right). Surprisingly, both in unpulsed or cAMP pulsed HSB61 cells, the kinetics of single Ca2+response was as rapid as the wild type in terms of influx and efflux (Fig. 12B, a-b). The increased concentration level of cGMP in the HSB61 mutant did not affect the timing of Ca2+ influx. This suggests that either cGMP level must be constantly high to affect Ca2+ entry [27] or that cGMP has no effect on Ca2+ influx [35]. Figure 12 Ca2+ influx in HSB61 mutant. A) Dependence of cAMP-induced Ca2+-influx from [Ca2+]e measured in unpulsed and pulsed HSB61 cell suspensions. Dose-response curves are shown for (left) untreated cells and (right) cells treated with 20 nM cAMP pulses. In both cases, cells were tested after 4–5 hours starvation. White circles depict the HSB61 cells and black circles the AX2 cells. Data are presented for at least 4 independent experiments for each cell type. P values = 0.00068 and 0.8368 for left and right graphs, respectively (Wilcoxon test). B) cAMP-induced Ca2+-influx in HSB61 and wild-type cells. Single electrode recordings are shown for (a, b) mutant cells either (a) unpulsed or (b) pulsed overnight with 20 nM cAMP. For comparison (c, d) wild-type cells are shown after 5 hours of development in (c) and after 7.5 hours in (d). Addition of cAMP is indicated. As shown in Fig. 12Bd, AX2 cells at 8 hours of development, which corresponds to late aggregation stage, displayed a rapid Ca2+ influx in response to cAMP, followed by a very slow efflux. Later on in development, no efflux could be detected [34]. In HSB61 cells, even when pulsed for as long as 12 hours, the kinetics of efflux was similar to 4–6 hours AX2 pulsed cells (Fig. 11B, b and c). Thus, these data further confirm that the mutant development is delayed even after 12 hours pulsing with cAMP. In addition, the finding that maximal Ca2+ entry in response to cAMP is reduced in the mutant may account, at least in part, for the chemotactic defect of the mutant. Mutant HSB61 is impaired in chemotactic, but not in spontaneous cell motility When cell motility was analysed, differences between mutant and parental strains were only detected during chemotaxis. Spontaneous cell motility was indistinguishable between both cell types. Directional cell migration in response to external chemoattractant gradients implies at least three steps: a) sensing the chemoattractant, b) formation of a leading front and c) cell polarization, with suppression of lateral pseudopods, followed by forward movement and uropod detachment [33]. As depicted in Fig. 13 and additional file 1, aggregation competent wild type cells, when challenged with the cAMP-loaded micropipette, become highly polarized, with a clearly defined leading front and a posterior uropod, and rapidly move towards the cAMP source. In their movement, AX2 cells eventually adhere to each other into streams, due to outward relay of the signal (Fig. 13, upper panel). Five hours-starved HSB61 null cells behaved differently: they sensed the gradient and extended a membrane lamella towards the micropipette, but they also extended several lateral pseudopodia with high frequency, thus displaying a severe polarization defect. In contrast to the parental strain, the mutant cells exhibited a rather flattened shape and an apparently increased cell-substratum adhesion. As a result, their chemotactic orientation and motility were strongly reduced (Fig. 13, middle panel and additional file 2). Pulsing with cAMP for at least 6–8 hours partially rescued the mutant cell phenotype, in that these cells were now better polarized and displayed an organized leading front, moving towards the capillary in a way similar to AX2, though the cell population was somewhat heterogeneous in that respect (Fig. 13, lower panel and additional file 3). The chemotactic speed of mutant cells changed from 1.43 ± 0.49 μm/min for untreated cells to 6.2 ± 2.1 μm/min for cAMP pulsed cells. Both values were below the 9.7 ± 1.81 μm/min observed for 5 hours starved AX2 cells. Figure 13 Chemotaxis of wild-type and HSB61 cells. Cells were developed in shaken suspension, either in the presence or absence of exogenous cAMP pulses, plated on coverslips and tested for chemotaxis towards a microcapillary diffusing cAMP. Upper and middle panels show AX2 or HSB61 cells starved for 5 hours, bottom panel shows HSB61 cells treated with cAMP pulses for 10 hours (see additional files 1, 2, 3). Higher magnifications of each cell sample are shown on the right. Numbers: time in minutes, starting after positioning of the microcapillary (0' time). Cyclic AMP pulses failed to induce cell streaming: the HSB61 cells chemotaxed towards the capillary mostly as single cells, suggesting that spontaneous cAMP relay was still reduced in the mutant. Stimulation with chemoattractants causes polymerization and reorganization of actin, and this has been correlated with the extension of new pseudopods during chemotaxis [36]. We investigated the levels of F-actin in HSB61 cells following cAMP pulses. Wild-type cells showed the expected biphasic response with a sharp peak of F-actin at about 5 seconds after stimulation and a lower second peak at about 50 seconds (Fig. 14A). A very low actin response was detected in HSB61 cells pulsed for 4 hours (Fig. 14B), but when cells were pulsed for at least 6 hours the intensity of actin response was similar to that of the parental strain (Fig. 14C). Figure 14 Actin polymerization assay in response to chemoattractant stimulation. The cells concentrated at 2 × 107/ml were stimulated with 1.0 μm cAMP (time 0) and the F-actin formation measured with the phalloidin binding assay as described in Material and Methods. Starving (A) AX2 and (B, C) HSB61 cells were treated with cAMP pulses and tested for F-actin polymerization after (A, B) 4 or (C) 6 hours of starvation. Discussion In this study we have reported the isolation and characterization of a novel RasGEF, named RasGEFM, which is required for proper Dictyostelium development. RasGEFM is peculiar in that it contains 6 poly-proline stretches that represent putative interacting sites for SH3, WW and/or EVH1 containing proteins [37], and a DEP1 motif. Disrupting rasGEFM results in cells to be blocked at the aggregation stage, forming rather small, flattened aggregates that fail to develop further. The RasGEFM-null phenotype could be partially rescued by pulsing cells with cAMP. Gene expression studies and functional assays indicate that the mutant phenotype is due to defective developmental expression of cAMP relay and are consistent with a role of RasGEFM in regulating cAMP signalling. The mutant phenotype differs, however, from the phenotype displayed by other mutants that are strictly defective in adenylyl cyclase activation, such as mutants in RasGEFA [9], PIA [21,22] CRAC [38] and RasC [29]. PIA and CRAC are cytosolic proteins that couple the G protein to adenylyl cyclase, and RasGEFA as well as RasC have been shown to also be essential for adenylyl cyclase activation. In contrast to rasGEFM null cells, which form small flat aggregates, all these other mutants fail to aggregate, although they respond to cAMP pulses. Their chemotactic motility in response to cAMP diffusing from a microcapillary is also much less impaired, and guanylyl cyclase activation is rather normal and not enhanced as in the rasGEFM null mutant. These findings indicate that these mutants display different alterations in signal transduction downstream of the cAMP receptor. The RASGEFM protein, in contrast to RasGEFA, PIA and CRAC, appears to be located downstream of the cAR1 receptor, but upstream of the G protein, thus possibly affecting both G protein dependent and G protein independent pathways. RasGEFM null cells have defects in both adenylyl cyclase activation and in chemotaxis, characteristics similar to those of rasC null cells. It might have been expected, therefore, that RasGEFM would act upstream of RasC. However, we now have direct evidence that RasGEFA directly activates RasC and is the only GEF responsible for the activation of RasC (Kae et al, unpublished). Consistent with this result, there was no major effect on the cAMP dependent activation of both RasC and RasG in the rasGEFM null cells. Several attempts have been performed to try to rescue the HSB61 phenotype but so far none of them succeeded, although a RasGEFM GFP-fused protein of the correct size was expressed (data not shown). This observation, combined with the rather peculiar rasGEFM mRNA expression, in which two transcripts are detected and regulated independently, suggests that the failed rescue might be due to the necessity for the RasGEFM protein to be expressed in a regulated way and not under the control of a constitutive promoter, such as the actin 15 promoter. Experiments are in progress in this direction and they will be crucial to confirm that the phenotype is due to direct disruption of the rasGEFM gene and not to additional detrimental effects of the transformation. The possibility of a second random mutation, which independently of rasGEFM disruption could be responsible for the observed phenotype, cannot be excluded but it seems unlikely to us. The GTPγS inhibition of cAMP binding, together with the stimulating effects of GTPγS on adenylyl cyclase, and the rescuing effect of cAMP, clearly locate the mutant defect upstream of the G protein and suggest a single event to be involved. When rasGEFM null cells are challenged with a cAMP filled micropipette, the chemotactic response is impaired. Apparently the cells sense the gradient but they seem incapable of organizing a leading front. Moreover, several lateral pseudopodia are continuously extended and retracted. As a consequence, the cell fails to polarize, maintains a rather flattened shape and moves "spastically" towards the micropipette. If mutant cells are treated with cAMP pulses for at least 6 to 8 hours, chemotactic orientation and speed are improved, and many cells become elongated and move faster towards the micropipette. Formation of a leading front, acquisition of polarity and suppression of lateral pseudopodia are processes characterized by the redistribution of cytoskeletal components, with F-actin and numerous actin-binding proteins being enriched at the front and myosin II assembled in filaments at the back of cells [33]. These processes are regulated by the balanced action of several sub-cellular components, including second messengers (e.g. Ca2+, cGMP) and proteins such as PI3K and PTEN [39]. A mechanism by which localized Ras activation mediates leading edge formation through activation of PI3K and other Ras effectors required for chemotaxis has been recently proposed for Dictyostelium cells by Sasaki et al. [28]. Cyclic AMP elicits in the mutant a biphasic actin polymerization response comparable to the parental strain, but the absolute peak values are strongly reduced. Strikingly, when mutant cells are pulsed for additional 2–4 hours the actin response is comparable to that of the wild type. These results suggest to us that RasGEFM is not directly involved in mediating a putative Ras-induced actin polymerization. They are instead consistent with the notion that a developmentally regulated component, required for proper actin recruitment in response to chemoattractant, is absent or expressed under a threshold level in the mutant and is induced by prolonged cAMP treatment. Taken together our findings support a role for RasGEF M in developmental acquisition of chemotactic efficiency and aggregation competence. The proposed localization of Ras GEFM downstream the cAMP receptor lead us to suggest that Ras GEFM and its cognate Ras may act as dissociating components between G protein coupled receptors and the G protein or molecular switches acting in a G protein independent way. A physical interaction between β 1 adrenergic receptors and a RasGEF has been recently reported in mammalian cells [40]. A peculiarity of the RasGEFM null mutant is the elevated cyclic GMP response to cAMP stimulation both at early and late starvation times, which suggests a role for RasGEFM as negative regulator of cAMP receptor-induced guanylyl cyclase activation. Cyclic GMP accumulation has been proposed to be regulated by adaptation [41] and to inhibit pseudopodia formation at the back of the cell by inducing myosin filament formation in the cell cortex [42-45]. Myosin filament formation at the presumptive leading front would be counteracted by myosin phosphorylation due to myosin heavy chain kinase A (MHCK-A), which is selectively recruited at the leading front [45]. As a result, a high cGMP concentration leads to improved orientation in a chemoattractant gradient, as shown in mutant cells lacking cGMP phosphodiesterase [46]. RasGEFM null cells, challenged with cAMP diffusing from a microcapillary, show, however, reduced orientation and increased lateral pseudopodia, despite a stronger cGMP response to cAMP. It may be argued that cGMP is rapidly hydrolysed in the HSB61 mutant, in contrast to the cGMP phosphodiesterase null cells. However, if cGMP is down-regulated by adaptation, it must be assumed that as long as cells are exposed to a chemoattractant gradient, as occurs when they are stimulated with cAMP diffusing from a microcapillary, adaptation of guanylyl cyclase is switched off [41], and this should lead in RasGEF M null cells to constantly higher levels of cGMP. Why is then chemotactic orientation reduced in the mutant? A possible explanation for this discrepancy is offered by the reduced actin response to chemoattractant. We have suggested that a developmentally regulated component, which is required for actin recruitment at the presumptive leading front, is lacking/down-regulated in the mutant. Strongly reduced actin polymerization at the front would result in impaired translocation of myosin heavy chain kinase A (MHCK-A), which has been shown to require F-actin binding [47]. As a consequence, a high cGMP level in the mutant, as in wild type cells or PDE null mutants, would induce myosin filaments all over the cell, consistent with its proposed role as global inhibitor [33,45], but myosin filaments in the front would not be dissociated by MHCK-A, due to its impaired recruitment. Thus we propose that in the RasGEFM null mutant chemotaxis is inhibited due firstly to altered F-actin polymerization at the presumptive leading front and secondarily to impaired recruitment of MHCK-A. Conclusion All the defects observed in the rasGEFM null mutant can be explained by assuming a modulatory role of the RasGEFM close to the cAMP receptor, which regulates guanylyl and adenylyl cyclases. RasGEFM appears to act as a negative regulator of guanylyl cyclase and a positive regulator of adenylyl cyclase. Currently we have no obvious explanation for this opposite activity. It must be kept in mind, however, that in contrast to adenylyl cyclase, which is stimulated by the Gβ subunit of the heterotrimeric G protein, receptor-dependent stimulation of guanylyl cyclase is more complex and less well understood. Several lines of evidence suggest a role for small GTPases in its activation, independently and in addition to the heterotrimeric G protein [48-50]. If RasGEFM and its putative Ras target act as receptor-linked molecular switches, this may lead to differential, albeit opposite, effects on heterotrimeric G-protein-dependent or independent pathways. Identifying the putative RasGEFM regulated Ras may help in understanding these complex transduction pathways. Methods Cell cultures, growth, and developmental conditions Wild type strain AX2-214 and rasGEFM null mutant, referred as HSB61, were grown either in liquid nutrient medium at 23°C under shaking at 150 rpm [51] or on nutrient agar plates with Escherichia coli B/2 [52]. When cultured in liquid medium, HSB61 cells were supplemented with blasticidin (ICN) at a final concentration of 10 μg/ml. For development on solid substratum, cells were grown to a density of 2–3 × 106 cell/ml, washed three times in 0.017 M Na+/K+ Soerensen phosphate buffer, pH 6.0 and, once deposited on non nutrient 1.5% (w/v) agar plates, allowed to develop at 23°C. For development in suspension, cells were incubated at a concentration of 1 × 107 cell/ml in Soerensen phosphate buffer. For cAMP treatment, pulses of 20 nM cAMP were applied every 6 minutes using a Braun perfusor VI equipped with 10 ml-syringe [53]. Measurement of EDTA-stable contacts EDTA-stable contacts were measured as described by using the agglutinometer of Beug and Gerisch [52]. Chemotaxis assay Chemotaxis was studied by using the microcapillary assay [54]. Briefly, cells were seeded onto 35 mm glass base dishes (Iwaki) at a density of approximately 1 × 105/cm2 and local stimulation of chemotaxis was obtained by passive diffusion of cAMP from a microcapillary (Femtotips1, Eppendorf), filled with a 1.0 mM solution of cAMP. The microcapillary was positioned with an automated Zeiss micromanipulator and cells observed either with a 20× or with a Neofluar 100x/1.3 oil immersion objective, equipped with DIC filter. Images were captured at an interval of 0.66 sec. and recorded in a Panasonic videorecorder (AG-TL700) with a ZVS-47DE camera (Zeiss) mounted on a microscope (Zeiss, Axiovert HAL100). The recorded time-lapse movies were transferred to a computer using USB Instant Video package (ADS Technologies). Actin polymerization assay Actin polymerization assays were carried out as previously described [36,55]. Briefly, cells were starved at 2 × 107cells/ml for 4 and 6 hours and pulsed with 1 μm cAMP. At the indicated time points, 100 μl samples were taken and transferred to 1 ml of actin buffer (20 mM K2PO4, 10 mM PIPES, 5 mM EGTA, 2 mM MgCl2, 3.7% formaldehyde, 0.1% Triton X-100, 0.25 μM TRITC-phalloidin, pH6.8). After shaking for 1 hour at room temperature, samples were centrifuged for 10 minutes at 16000 g and the resulting pellet was resuspended in 1 ml methanol. After shaking overnight, the amount of F-actin was determined by measuring the fluorescence with a fluorimeter (Kontron SFM 25). Setting: excitation wavelength 540 nm, emission 570 nm. Measurement of cAMP-induced Ca2+-influx Net Ca2+-influx after agonist stimulation was done as described previously [34,56]. Cells were developed in suspension as described above. For accelerating developmental gene expression cells were pulsed with 20 nM cAMP every 6 minutes, overnight in case of HSB61 cells, and for 2 hours in case of wild-type cells. At appropriate time points the cells were washed in nominally Ca2+ free tricine buffer (tricine pH 7.0, supplemented with 5 mM KCl) and resuspended at a cell density of 5 × 107 cells/ml. The cell suspension was then stimulated with cAMP and Ca2+-influx measured with a Ca2+-sensitive electrode (Möller) and a voltmeter (Metrohm). Statistical analysis was performed with "Wilcoxon test". Light scattering measurements of cells in suspension were done has described by Gerisch and Hess [57]. The cell suspension (2 × 107cell/ml) was aerated in a cuvette and extinction was concomitantly monitored at 500 nm in a Zeiss PM6 spectrophotometer. In vitro stimulation of adenylyl cyclase Adenylyl cyclase in Dictyostelium lysates was assayed in the presence of 2'-deoxy-cAMP or GTPγS as described by Lilly and Devreotes [58]. Briefly starving cells were treated with 20 nM cAMP pulses every 6 min for different hours. A total of 1 × 108 cells were pelleted and resuspended in 1 ml of Soerensen phosphate buffer. An equal volume of ice cold lysis buffer containing either 4 mM MgCl2 or 4 mM MnCl2, 20 mM Tris pH 8.0 was added. Cells were lysed by passage through a 3-μm pore size Nucleopore membrane in the absence or presence of 30 μM GTPγS or 50 μM 2'-deoxy-cAMP and the lysates incubated on ice for 5 min. A 40 μl aliquot of cell lysate was added to a 40 μl assay mix (20 mM DTT, 1 mM ATP, 2 mM MgCl2 or 2 mM MnCl2 in 10 mM Tris pH 8.0, 0.4 mM IBMX) and incubated at 20°C for 5 min. The reaction was stopped, by adding 40 μl 0.1 M EDTA pH 8.0 and boiling the sample for 2 min. The total concentration of cAMP in the samples was determined by using the "Biotrak cAMP assay Kit" according to manufacturer's instructions (Amersham Pharmacia Biotech). cAMP binding assays Binding of cAMP to cell surface receptors was determined as previously described by Van Haastert [59]. Briefly, wild-type and mutant cells were starved by shaking in Soerensen phosphate buffer, without or with cAMP pulses for 5 or 8 hours respectively, washed and resuspended at a density of 1 × 108 cell/ml. Aliquots of 80 μl of cells were incubated with a radioactive binding mixture, containing 300 nM of [3H]cAMP (Amersham Pharmacia Biotech), 50 mM dithiothreitol in 90% saturated ammonium sulphate, and a variety of cAMP concentration ranging between 700 to 19700 nM. Specific binding was obtained by subtracting non specific binding determined in the presence of 1 mM cAMP. After 5 min. incubation at 0°C, cells were collected by centrifugation at 14000 × g for 2 min, the pellet resuspended in 100 μl of 0.1 M acetic acid and dissolved in 1.3 ml scintillation fluid. Scatchard plots of cAMP binding were done with GraphPad software (GraphPad Inc.). GTP-inhibition of cAMP binding to plasma membranes The assay was performed as described earlier [27]. Brieflly, GTP-inhibition of cAMP binding was measured in a total volume of 100 μl containing PB (10 mM KH2PO4/Na2HPO4 pH 6.5), 5 nM [3H] cAMP, 10 mM dithiothreitol, GTPγS (300 μM when present) and 70 μl membranes. Samples were incubated 5 min. at 0°C, centrifuged for 2 min. at 14000 × g, the supernatant was aspirated and the pellet dissolved in 100 μl acetic acid. Radioactivity was determined after the addition of 1.3 ml of liquid scintillation. Cyclic AMP and cGMP determination For the determination of cyclic nucleotides, cells aliquots were taken at different time points before and after a cAMP pulse, and quenched with 1 vol of 2 N perchloric acid [24]. After centrifugation, neutralization of the supernatant with potassium carbonate, and acetylation, the concentration of cAMP and cGMP in the extract was measured using the 125I radioimmunoassay kit according to manufacturer's instructions (Amersham Pharmacia Biotech). RBD binding assay The RBD binding assay was performed as described elsewhere [30]. Briefly, 6 h or 10 h pulsed cells were washed twice and resuspended at 5 × 107 cell/ml in Soerensen phosphate buffer. Cells were stimulated with 200 nM cAMP, aliquots (0.5 ml) were selected at the indicated time points and lysed in an equal volume of ice-cold 2× HK-LB (20 mM sodium phosphate pH 7.2, 2% Triton X-100, 20% glycerol, 300 mM NaCl, 20 mM MgCl2, 2 mM EDTA, 2 mM Na3VO4, 10 mM NaF, with protease inhibitor Roche), and incubated on ice for 5 min. The lysates were cleared by centrifugation for 10 min and protein concentrations were determined using DC Protein Assay (Bio-Rad). A 0.8 mg portion of protein lysates was incubated with 100 μg of GST-Byr2 (RBD) and the mixture was incubated at 4°C for 1 h. Beads were harvested by centrifugation and washed three times in 1× HK-LB. A volume of 40 μl of 1× SDS gel loading buffer was added to the pellet beads and the mixture was boiled for 5 min. Samples were fractionated by SDS-PAGE, blotted, blocked and probed with RasC or RasG antibody. Bands were determined by enhanced chemiluminescence reaction (Amersham Pharmacia Biotech). Molecular cloning and sequence analysis The rasGEFM gene was isolated using PCR based method. Through the sequence derived from DNA database screening [13,14] a pair of primers was designed (5'-ATGATGAATGAAGTTTCTTCAAATTC-3' and 5'-CCATCGATAATTATCTAAATAATGGATTTGA-3') and used for PCR amplification. As template, cDNA isolated with "First strand cDNA synthesis kit" (Amersham Pharmacia Biotech), from RNA of cells developed for 5 hours on solid substrata, was used. The DNA fragment, of approximately 2.7 kb, was then ligated into "pGEM-T easy vector" (PROMEGA) and cloned into DH5α E. coli strain. PCR products were purified from gel with "High Pure PCR Product Purification" kit (Roche) and verified by sequencing. Construction of the D.d. rasGEFM null strain (HSB61) To construct the rasGEFM null strain the rasGEFM locus was disrupted via homologous recombination. The blasticidin resistance gene (bsr), used as selectable marker, was excised from pUCBsrΔ Bam [60] with HindIII and XbaI and subsequently inserted into the JC2a86b07.s1 clone (Dictyostelium genome project) digested with HindIII and XbaI. Subsequently the N-terminal portion (from position 0 to 617 bp) of the gene was ligated into KpnI site of the above vector, using the "DNA ligation kit" (Amersham Pharmacia Biotech). The vector, carrying the selectable marker was electroporated [61] into parental strain and transformed cells selected for blasticidin resistence. Resistant cells were cloned, and clones subsequently screened via Southern blot in order to identify Dictyostelium clones in which the RasGEF M locus was disrupted. Southern and Northern hybridization analysis Genomic DNA was extracted and purified by CsCl gradient centrifugation as previously described by Nellen, et al. [62], digested with EcoRI, run onto 0.8% agarose gel, blotted onto Hybond-N membrane (Amersham Pharmacia Biotech), and subjected to Southern assay [63]. The membrane was probed with a 800 bp RasGEF M cDNA specific probe, corresponding to bp. 760–1518 of the cDNA clone, previously radiolabelled by the "Megaprimer™DNA Labelling System" using [α32] dATP (Amersham). For Northern blots, total RNA was prepared using TRIZOL reagent (GIBCO) according to manifacturer's instructions. RNA was then resuspended in DEPC treated water, quantified, and 15 μg were size separated on 1.2% agarose gel in presence of formaldehyde. Equal loading of samples, was checked by probing membranes with the actin gene. The radiolabelled DNA fragments used as probes were as follow: car1, csA, acaA (fragment from 2.7 kb to 4.2 kb of cDNA clone), RasGEF M. After being hybridised with the first cDNA probe, Northern blots were stripped with 0.1% SDS in boiling water and then re-hybridised with a second probe. Authors' contributions M. A. and E. B. carried most of the experiments and were involved in drafting the manuscript. D. L. was responsible for light scattering oscillations and Ca2+ analysis, H. K. and G. W. for pull down experiments with RasC and RasG. S. B. supervised the work and revised the manuscript, with contributions by all authors. Supplementary Material Additional File 1 Chemotaxis of AX2 after 5 hours of starvation.mov: 6.5 MB. Cells were seeded onto 35 mm/glass base dish (Iwaki) and subjected to a cAMP filled micropipette assay, as described in Material and Methods. Images were recorded in a Panasonic videorecorder (AG-TL700) with a ZVS-47 DE camera (Zeiss) mounted on an Axiovert HAL100, using Neofluar 100X/1.3 oil immersion objective and DIC filter. The recorded time-lapse movie was transferred to a computer using USB Istant Video (ADS Technologies). Click here for file Additional File 2 Chemotaxis of HSB61 after 5 hours of starvation.mov: 6.9 MB Click here for file Additional File 3 Chemotaxis of HSB61 starved and pulsed with cAMP for 10 hours.mov: 9.9 MB Click here for file Acknowledgements We thank Barbara Peracino for technical help and D. Malchow for helpful suggestions. This work was supported by funds of MIUR (cofin 2002) to S.B. and local funds of the University of Turin to E.B. ==== Refs Campbell SL Khosravi-Far R Rossman KL Clark GJ Der CJ Increasing complexity of Ras signaling Oncogene 1998 17 1395 1413 9779987 10.1038/sj.onc.1202174 Lim CJ Spiegelman GB Weeks G Cytoskeletal regulation by Dictyostelium Ras subfamily proteins Journal of Muscle Research & Cell Motility 2002 23 729 736 12952071 Takai Y Sasaki T Matozaki T Small GTP-binding proteins Physiological Reviews 2001 81 153 208 11152757 Bourne HR Sanders DA McCormick F The GTPase superfamily: conserved structure and molecular mechanism Nature 1991 349 117 127 1898771 10.1038/349117a0 Boguski MS McCormick F Proteins regulating Ras and its relatives Nature 1993 366 643 654 8259209 10.1038/366643a0 Wilkins A Insall RH Small GTPases in Dictyostelium: lessons from a social amoeba Trends Genet 2001 17 41 48 11163921 10.1016/S0168-9525(00)02181-8 Broek D Toda T Michaeli T Levin L Birchmeier C Zoller M Powers S Wigler M The S. cerevisiae CDC25 gene product regulates the RAS/adenylate cyclase pathway Cell 1987 48 789 799 3545497 10.1016/0092-8674(87)90076-6 Simon MA Bowtell DD Dodson GS Laverty TR Rubin GM Ras1 and a putative guanine nucleotide exchange factor perform crucial steps in signaling by the sevenless protein tyrosine kinase Cell 1991 67 701 716 1934068 10.1016/0092-8674(91)90065-7 Insall RH Borleis J Devreotes PN The aimless RasGEF is required for processing of chemotactic signals through G-protein-coupled receptors in Dictyostelium Current Biology 1996 6 719 729 8793298 10.1016/S0960-9822(09)00453-9 Wilkins A Chubb JR Insall RH A novel Dictyostelium RasGEF is required for normal endocytosis, cell motility and multicellular development Current Biology 2000 10 1427 1437 11102804 10.1016/S0960-9822(00)00797-1 Goldberg JM Bosgraaf L Van Haastert PJ Smith JL Identification of four candidate cGMP targets in Dictyostelium Proceedings of the National Academy of Sciences of the United States of America 2002 99 6749 6754 12011437 10.1073/pnas.102167299 Bosgraaf L Russcher H Smith JL Wessels D Soll DR van Haastert PJM A novel cGMP signalling pathway mediating myosin phosphorylation and chemotaxis in Dictyostelium EMBO J 2002 21 4560 4570 12198158 10.1093/emboj/cdf438 Dictyostelium genome project database: Dictyostelium genome project database: Macias MJ Wiesner S Sudol M WW and SH3 domains, two different scaffolds to recognize proline-rich ligands FEBS Letters 2002 513 30 37 11911877 10.1016/S0014-5793(01)03290-2 Lai CCBM Broek D Powers S Influence of Guanine Nucleotides on complex formation between Ras AND CDC25 proteins Molecular and Cellular Biology 1993 13 1345 1352 8441380 Ponting CP Bork P Pleckstrin's repeat performance: a novel domain in G-protein signaling? Trends in Biochemical Sciences 1996 21 245 246 8755244 10.1016/0968-0004(96)30027-3 Wong HC Mao J Nguyen JT Srinivas S Zhang W Liu B Li L Wu D Zheng J Structural basis of the recognition of the dishevelled DEP domain in the Wnt signaling pathway Nature Structural Biology 2000 7 1178 1184 11101902 10.1038/82047 Gerisch G Cyclic AMP and other signals controlling cell development and differentiation in Dictyostelium Annu Rev Biochem 1987 56 853 879 3039908 10.1146/annurev.bi.56.070187.004225 Lilly P Wu L Welker DL Devreotes PN A G-protein beta-subunit is essential for Dictyostelium development Genes & Development 1993 7 986 995 8099335 Chen MY Long Y Devreotes PN A novel cytosolic regulator, Pianissimo, is required for chemoattractant receptor and G protein-mediated activation of the 12 transmembrane domain adenylyl cyclase in Dictyostelium Genes Devel 1997 11 3218 3231 9389653 Pergolizzi B Peracino B Silverman J Ceccarelli A Noegel A Devreotes P Bozzaro S Temperature-sensitive inhibition of development in Dictyostelium due to a point mutation in the piaA gene Developmental Biology 2002 251 18 26 12413895 10.1006/dbio.2002.0809 Ceccarelli A Bozzaro S Selection of mutants defective in binding to immobilized carbohydrates in Dictyostelium discoideum Anim Biol 1992 1 59 68 Bozzaro S Hagmann J Noegel A Westphal M Calautti E Bogliolo E Cell differentiation in the absence of intracellular and extracellular cyclic AMP pulses in Dictyostelium discoideum Dev Biol 1987 123 540 548 10.1016/0012-1606(87)90412-X Bozzaro S Ponte E Cell adhesion in the life cycle of Dictyostelium Experientia 1995 51 1175 1188 8536805 10.1007/BF01944735 Queller DC Ponte E Bozzaro S Strassmann JE Single-gene greenbeard effects in the social amoeba Dictyostelium discoideum Science 2003 299 105 106 12511650 10.1126/science.1077742 Van Haastert PJM Alteration of Receptor/G-protein Interaction by Putative Endogenous Protein Kinase Activity in Dictyostelium discoideum Membranes J Biol Chem 1987 262 3239 3243 3029106 Sasaki AT Chun C Takeda K Firtel RA Localized Ras signaling at the leading edge regulates PI3K, cell polarity, and directional cell movement Journal of Cell Biology 2004 167 505 518 15534002 10.1083/jcb.200406177 Lim CJ Spiegelman GB Weeks G RasC is required for optimal activation of adenylyl cyclase and Akt/PKB during aggregation EMBO Journal 2001 20 4490 4499 11500376 10.1093/emboj/20.16.4490 Kae H Lim CJ Spiegelman GB Weeks G Chemoattractant-induced Ras activation during Dictyostelium aggregation EMBO Reports 2004 5 602 606 15143344 10.1038/sj.embor.7400151 Schaloske RH Lusche DF Bezares-Roder K Happle K Malchow D Schlatterer C Ca2+ regulation in the absence of the iplA gene product in Dictyostelium discoideum BMC Cell Biol 2005 6 13 15760480 10.1186/1471-2121-6-13 Lusche DF Malchow D Developmental control of cAMP-induced Ca2+-influx by cGMP: influx is delayed and reduced in a cGMP-phosphodiesterase D deficient mutant of Dictyostelium discoideum Cell Calcium 2005 37 57 67 15541464 10.1016/j.ceca.2004.06.005 Van Haastert PJ Devreotes PN Chemotaxis: signalling the way forward Nature Reviews Molecular Cell Biology 2004 5 626 634 15366706 10.1038/nrm1435 Bumann J Wurster B Malchow D Attractant-induced changes and oscillations of the extracellular Ca++ concentration in suspensions of differentiating Dictyostelium cells J Cell Biol 1984 98 173 178 6323484 10.1083/jcb.98.1.173 Veltman DM de Boer JS van Haastert PJM Chemoattractant-stimulated calcium influx in Dictyostelium discoideum does not depend on cGMP Biochim Biophys Acta 2003 1623 129 134 14572910 Hall AL Schlein A Condeelis J Relationship of pseudopod extension to chemotactic hormone-induced actin polymerization in amoeboid cells J Cell Biochem 1988 37 285 299 3410887 10.1002/jcb.240370304 Zarrinpar A Bhattacharyya RP Lim WA The structure and function of proline recognition domains Science's Stke [Electronic Resource]: Signal Transduction Knowledge Environment 2003 2003 RE8 12709533 Insall R Kuspa A Lilly PJ Shaulsky G Levin LR Loomis WF Devreotes P CRAC, a cytosolic protein containing a pleckstrin homology domain, is required for receptor and G protein-mediated activation of adenylyl cyclase in Dictyostelium J Cell Biol 1994 126 1537 1545 8089184 10.1083/jcb.126.6.1537 Franca-Koh J Devreotes PN Moving forward: mechanisms of chemoattractant gradient sensing Physiology 2004 19 300 308 15381759 10.1152/physiol.00017.2004 Pak Y Pham N Rotin D Direct binding of the beta1 adrenergic receptor to the cyclic AMP-dependent guanine nucleotide exchange factor CNrasGEF leads to Ras activation Molecular & Cellular Biology 2002 22 7942 7952 12391161 10.1128/MCB.22.22.7942-7952.2002 van Haastert PJM van der Heijden PR Excitation adaptation and deadaptation of the cAMP mediated cGMP response in Dictyostelium discoideum J Cell Biol 1983 96 347 353 6300135 10.1083/jcb.96.2.347 Uchida KSK Kitanishi-Yumura T Yumura S Myosin II contributes to the posterior contraction and the anterior extension during the retraction phase in migrating Dictyostelium cells J Cell Sci 2003 116 51 60 12456715 10.1242/jcs.00195 Moores SL Sabry JH Spudich JA Myosin dynamics in live Dictyostelium cells Proc Natl Acad Sci USA 1996 93 443 446 8552657 10.1073/pnas.93.1.443 Stites J Wessels D Uhl A Egelhoff T Shutt D Soll DR Phosphorylation of the Dictyostelium myosin II heavy chain is necessary for maintaining cellular polarity and suppressing turning during chemotaxis Cell Motil Cytoskel 1998 39 31 51 10.1002/(SICI)1097-0169(1998)39:1<31::AID-CM4>3.0.CO;2-J Postma M Bosgraaf L Loovers HM van Haastert PJM Chemotaxis: signalling modules join hands at front and tail EMBO Rep 2004 5 35 40 14710184 10.1038/sj.embor.7400051 Bosgraaf L Russcher H Snippe H Bader S Wind J van Haastert PJM Identification and characterization of two unusual cGMP-stimulated phosphodiesterases in Dictyostelium Mol Biol Cell 2002 13 3878 3889 12429832 10.1091/mbc.E02-05-0302 Steimle PA Licate L Cote GP Egelhoff TT Lamellipodial localization of Dictyostelium myosin heavy chain kinase A is mediated via F-actin binding by the coiled-coil domain FEBS Lett 2002 516 58 62 11959103 10.1016/S0014-5793(02)02494-8 Roelofs J van Haastert PJM Characterization of two unusual guanylyl cyclases from Dictyostelium J Biol Chem 2002 277 9167 9174 11777934 10.1074/jbc.M111437200 Wu LJ Valkema R van Haastert PJM Devreotes PN The G protein beta subunit is essential for multiple responses to chemoattractants in Dictyostelium J Cell Biol 1995 129 1667 1675 7790362 10.1083/jcb.129.6.1667 Roelofs J Loovers HM van Haastert PJM GTPgammaS regulation of a 12-transmembrane guanylyl cyclase is retained after mutation to an adenylyl cyclase J Biol Chem 2001 276 40740 40745 11522784 10.1074/jbc.M105154200 Watts DJ Ashworth JM Growth of myxameobae of the cellular slime mould Dictyostelium discoideum in axenic culture Biochemical Journal 1970 119 171 174 5530748 Bozzaro S Merkl R Gerisch G Cell adhesion: its quantification, assay of the molecules involved, and selection of defective mutants in Dictyostelium and Polysphondylium Meth Cell Biol 1987 28 359 385 Devreotes P Fontana D Klein P Sherring J Theibert A Transmembrane signaling in Dictyostelium Methods in Cell Biology 1987 28 299 331 3037274 Bozzaro S Roseman S Adhesion of Dictyostelium discoideum cells to carbohydrates immobilized in polyacrylamide gels. II. Effect of D-glucoside derivatives on development J Biol Chem 1983 258 13890 13897 6315710 Peracino B Borleis J Jin T Westphal M Schwartz JM Wu LJ Bracco E Gerisch G Devreotes P Bozzaro S G protein beta subunit-null mutants are impaired in phagocytosis and chemotaxis due to inappropriate regulation of the actin cytoskeleton J Cell Biol 1998 141 1529 1537 9647646 10.1083/jcb.141.7.1529 Schaloske R Sordano C Bozzaro S Malchow D Stimulation of calcium influx by platelet activating factor in Dictyostelium J Cell Sci 1995 108 1597 1603 7615678 Gerisch G Hess B Cyclic-AMP-controlled oscillations in suspended Dictyostelium cells: Their relation to morphogenetic cell interactions Proc Natl Acad Sci USA 1974 71 2118 2122 4365764 Lilly PJ Devreotes PN Identification of CRAC, a cytosolic regulator required for guanine nucleotide stimulation of adenylyl cyclase in Dictyostelium Journal of Biological Chemistry 1994 269 14123 14129 8188693 Van Haastert PJM The modulation of cell surface cAMP receptors from Dictyostelium discoideum by ammonium sulphate Biochim Biophys Acta 1985 845 254 260 2986720 10.1016/0167-4889(85)90184-3 Sutoh K A transformation vector for dictyostelium discoideum with a new selectable marker bsr Plasmid 1993 30 150 154 8234487 10.1006/plas.1993.1042 Pang KM Lynes MA Knecht DA Variables controlling the expression level of exogenous genes in Dictyostelium Plasmid 1999 41 187 197 10366524 10.1006/plas.1999.1391 Nellen W Datta S Reymond C Sivertsen A Mann S Crowley T Firtel RA Molecular biology in Dictyostelium: tools and applications Meth Cell Biol 1987 28 67 100 Southern EM Detection of specific sequences among DNA fragments separated by gel electrophoresis Journal of Molecular Biology 1975 98 503 517 1195397	16336640	PMC1325028	CC BY	2021-01-04 16:31:30	no	BMC Cell Biol. 2005 Dec 7; 6:43	utf-8	BMC Cell Biol	2,005	10.1186/1471-2121-6-43	oa_comm
==== Front Nutr Metab (Lond)Nutrition & Metabolism1743-7075BioMed Central London 1743-7075-2-341631863710.1186/1743-7075-2-34ResearchA low-carbohydrate, ketogenic diet to treat type 2 diabetes Yancy William S [email protected] Marjorie [email protected] Allison M [email protected] Mary C [email protected] Eric C [email protected] Center for Health Services Research in Primary Care, Department of Veterans' Affairs Medical Center (152), 508 Fulton Street, Durham, NC, USA 277052 Department of Medicine, Duke University Medical Center, Durham, NC, USA3 Private Bariatric and Family Practice, and Clinical Faculty, University of Kansas School of Medicine, Lawrence, KS, USA2005 1 12 2005 2 34 34 10 8 2005 1 12 2005 Copyright © 2005 Yancy et al; licensee BioMed Central Ltd.2005Yancy 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 low-carbohydrate, ketogenic diet (LCKD) may be effective for improving glycemia and reducing medications in patients with type 2 diabetes. Methods From an outpatient clinic, we recruited 28 overweight participants with type 2 diabetes for a 16-week single-arm pilot diet intervention trial. We provided LCKD counseling, with an initial goal of <20 g carbohydrate/day, while reducing diabetes medication dosages at diet initiation. Participants returned every other week for measurements, counseling, and further medication adjustment. The primary outcome was hemoglobin A1c. Results Twenty-one of the 28 participants who were enrolled completed the study. Twenty participants were men; 13 were White, 8 were African-American. The mean [± SD] age was 56.0 ± 7.9 years and BMI was 42.2 ± 5.8 kg/m2. Hemoglobin A1c decreased by 16% from 7.5 ± 1.4% to 6.3 ± 1.0% (p < 0.001) from baseline to week 16. Diabetes medications were discontinued in 7 participants, reduced in 10 participants, and unchanged in 4 participants. The mean body weight decreased by 6.6% from 131.4 ± 18.3 kg to 122.7 ± 18.9 kg (p < 0.001). In linear regression analyses, weight change at 16 weeks did not predict change in hemoglobin A1c. Fasting serum triglyceride decreased 42% from 2.69 ± 2.87 mmol/L to 1.57 ± 1.38 mmol/L (p = 0.001) while other serum lipid measurements did not change significantly. Conclusion The LCKD improved glycemic control in patients with type 2 diabetes such that diabetes medications were discontinued or reduced in most participants. Because the LCKD can be very effective at lowering blood glucose, patients on diabetes medication who use this diet should be under close medical supervision or capable of adjusting their medication. ==== Body Background Prior to the advent of exogenous insulin for the treatment of diabetes mellitus in the 1920's, the mainstay of therapy was dietary modification. Diet recommendations in that era were aimed at controlling glycemia (actually, glycosuria) and were dramatically different from current low-fat, high-carbohydrate dietary recommendations for patients with diabetes [1,2]. For example, the Dr. Elliot Joslin Diabetic Diet in 1923 consisted of "meats, poultry, game, fish, clear soups, gelatin, eggs, butter, olive oil, coffee, tea" and contained approximately 5% of energy from carbohydrates, 20% from protein, and 75% from fat [3]. A similar diet was advocated by Dr. Frederick Allen of the same era [4]. Recently, four studies have re-examined the effect of carbohydrate restriction on type 2 diabetes. One outpatient study enrolled 54 participants with type 2 diabetes (out of 132 total participants) and found that hemoglobin A1c improved to a greater degree over one year with a low-carbohydrate diet compared with a low-fat, calorie-restricted diet [5,6]. Another study enrolled 8 men with type 2 diabetes in a 5-week crossover outpatient feeding study that tested similar diets [7]. The participants had greater improvement in glycohemoglobin while on the low-carbohydrate diet than when on a eucaloric low-fat diet. The third study was an inpatient feeding study in 10 participants with type 2 diabetes [8]. After only 14 days, hemoglobin A1c improved from 7.3% to 6.8%. In the fourth study, 16 participants with type 2 diabetes who followed a 20% carbohydrate diet had improvement of hemoglobin A1c from 8.0% to 6.6% over 24 weeks [9]. Only these latter three studies targeted glycemic control as a goal, and two of these were intensely-monitored efficacy studies in which all food was provided to participants for the duration of the study [7,8]. Three of the studies [6,8,9] mentioned that diabetic medications were adjusted but only one of them provided detailed information regarding these adjustments [9]. This information is critical for patients on medication for diabetes who initiate a low-carbohydrate diet because of the potential for adverse effects resulting from hypoglycemia. The purpose of this study was to evaluate the effects of a low-carbohydrate, ketogenic diet (LCKD) in overweight and obese patients with type 2 diabetes over 16 weeks. Specifically, we wanted to learn the diet's effects on glycemia and diabetes medication use in outpatients who prepared (or bought) their own meals. In a previous article, we reported the results observed in 7 individuals [10]; this report includes data from those 7 individuals along with data from additional participants enrolled subsequently. Methods Participants Participants were recruited from the Durham Veterans Affairs Medical Center (VAMC) outpatient clinics. Inclusion criteria were age 35–75 years; body mass index (BMI) >25 kg/m2; and fasting serum glucose >125 mg/dL or hemoglobin A1c >6.5% without medications, or treatment with oral hypoglycemic agents (OHA) and/or insulin. Exclusion criteria were evidence of renal insufficiency, liver disease, or unstable cardiovascular disease by history, physical examination, and laboratory tests. All participants provided written informed consent approved by the institutional review board. No monetary incentives were provided. Intervention At the first visit, participants were instructed how to follow the LCKD as individuals or in small groups, with an initial goal of ≤20 g carbohydrate per day. Participants were taught the specific types and amounts of foods they could eat, as well as foods to avoid. Initially, participants were allowed unlimited amounts of meats, poultry, fish, shellfish, and eggs; 2 cups of salad vegetables per day; 1 cup of low-carbohydrate vegetables per day; 4 ounces of hard cheese; and limited amounts of cream, avocado, olives, and lemon juice. Fats and oils were not restricted except that intake of trans fats was to be minimized. Participants were provided a 3-page handout and a handbook [11] detailing these recommendations. Participants prepared or bought all of their own meals and snacks following these guidelines. In addition, on the day the diet was initiated, diabetes medications were reduced – generally, insulin doses were halved, and sulfonylurea doses were halved or discontinued. Due to the possible diuretic effects of the diet soon after initiation, diuretic medications were discontinued if of low dosage (up to 25 mg of hydrochlorothiazide or 20 mg of furosemide) or halved if of higher dosage. Participants were also instructed to take a standard multivitamin and drink 6–8 glasses of water daily, and were encouraged to exercise aerobically for 30 minutes at least three times per week. Participants returned every other week for 16 weeks for further diet counseling and medication adjustment. When a participant neared half the weight loss goal or experienced cravings, he or she was advised to increase carbohydrate intake by approximately 5 g per day each week as long as weight loss continued. Participants could choose 5 g carbohydrate portions from one of the following foods each week: salad vegetables, low-carbohydrate vegetables, hard or soft cheese, nuts, or low-carbohydrate snacks. Diabetes medication adjustment was based on twice daily glucometer readings and hypoglycemic episodes, while diuretic and other anti-hypertensive medication adjustments were based on orthostatic symptoms, blood pressure, and lower extremity edema. Measurements Participants completed take-home food records (4 consecutive days, including a weekend) collected at baseline and at weeks 2, 8, and 16 during the study. Participants were given handouts with examples of how to complete the records. A registered dietician analyzed the food records using a nutrition software program (Food Processor SQL, ESHA Research, Inc., Salem, OR). The following measurements were made every other week: anthropometric and vital sign measurements; urine testing for ketones; and assessment for hypoglycemic episodes and other symptomatic side effects. Weight was measured on a standardized digital scale while the participant was wearing light clothes and shoes were removed. Skinfold thickness was measured at 4 sites – the average of 2 measurements at each site was entered into an equation to calculate percent body fat [12]. Waist circumference was measured at the midpoint between the inferior rib and the iliac crest using an inelastic tape; 2 measurements were averaged in the analysis. Blood pressure and heart rate were measured after the participant had been seated quietly without talking for 3 minutes. Certified laboratory technicians assessed urine ketones from a fresh specimen using the following semi-quantitative scale: none, trace (up to 0.9 mmol/L [5 mg/dL]), small (0.9–6.9 mmol/L [5–40 mg/dL]), moderate (6.9–13.8 mmol/L [40–80 mg/dL]), large80 (13.8–27.5 mmol/L [80–160 mg/dL]), large160 (>27.5 mmol/L [160 mg/dL]). Hypoglycemic episodes and symptomatic side effects were assessed by direct questioning of the participant and by self-administered questionnaires. Blood specimens were obtained at weeks 0, 8, and 16 after the participant had fasted overnight. The following serum tests were performed in the hospital laboratory using standardized methods: complete blood count, chemistry panel, lipid panel, thyroid-stimulating hormone, and uric acid. A non-fasting specimen was also drawn at weeks 4 and 12 to monitor electrolytes and kidney function. The primary outcome was the change from baseline to week 16 in hemoglobin A1c. Changes in all variables were analyzed by the paired t-test or Wilcoxon signed-ranks test, as appropriate. Linear regression analysis was used to examine predictors of change in hemoglobin A1c. A p value of 0.05 or less was considered statistically significant. Statistical analysis was performed using SAS version 8.02 (SAS Institute, Cary, NC). Results Of the 28 participants enrolled in the study, 21 completed the 16 weeks of follow-up. Reasons for discontinuing the study included unable to adhere to study meetings and unable to adhere to the diet; no participant reported discontinuing as a result of adverse effects associated with the intervention. All but one of the 21 participants were men; 62% (n = 13) were Caucasian, 38% (n = 8) were African-American (Table 1). The mean age was 56.0 ± 7.9 years. Table 1 Baseline characteristics (n = 21) Characteristic Summary Age, years, mean (SD) 56.0 (7.9) Gender, male, n (%) 20 (95%) Race, White, n (%) 13 (62%) African-American, n (%) 8 (38%) Weight, kg, mean (SD) 131.4 (18.3) BMI, kg/m2, mean (SD) 42.2 (5.8) Adequate food records were available for analysis in a proportion of participants at each of the 4 timepoints (Table 2). Participants completed food records at a mean of 2.5 and a median of 3 timepoints. In general, comparing baseline to subsequent timepoints, mean carbohydrate intake decreased substantially and energy intake decreased moderately while protein and fat intake remained fairly constant. Table 2 Diet composition Nutrient Week 0 Mean (SD) Week 2 Mean (SD) Week 8 Mean (SD) Week 16 Mean (SD) n 14 15 15 8 Carbohydrate, g 204.4 (118.4) 44.6 (27.4) 44.0 (29.1) 33.8 (24.6) Protein, g 95.8 (23.9) 111.7 (38.6) 114.8 (57.0) 98.5 (52.5) Fat, g 95.5 (27.3) 95.1 (47.2) 106.6 (47.6) 93.5 (63.7) Energy, kcal 2031.5 (521.4) 1515.5 (587.2) 1603.4 (713.0) 1418.7 (756.9) From baseline to week 16, the mean body weight decreased significantly from 131.4 ± 18.3 kg to 122.7 ± 18.9 kg, BMI decreased from 42.2 ± 5.8 kg/m2 to 39.4 ± 6.0 kg/m2, and waist circumference from 130.0 ± 10.5 cm to 123.3 ± 11.3 cm (Table 3). The percent change in body weight was -6.6%. The mean percent body fat decreased from 40.4 ± 5.8% to 37.0 ± 6.0%. Systolic and diastolic blood pressures did not change significantly over the 16 weeks. The mean heart rate decreased from 81.2 ± 12.9 beats per minute to 74.6 ± 14.0 beats per minute (p = 0.01). Table 3 Anthropometric and vital sign measurements (n = 21) Measurement Week 0 Mean (SD) Week 16 Mean (SD) Change % p value* Body weight, kg 131.4 (18.3) 122.7 (18.9) -6.6 <0.001 Body mass index, kg/m2 42.2 (5.8) 39.4 (6.0) -6.6 <0.001 Waist circumference, cm 130.0 (10.5) 123.3 (11.3) -5.2 <0.001 Percent body fat, % 40.4 (5.8) 37.0 (6.0) -8.4 <0.001 Systolic blood pressure, mm Hg 135.1 (14.8) 135.4 (17.6) 0.2 0.9 Diastolic blood pressure, mm Hg 79.2 (14.9) 74.1 (13.0) -6.4 0.1 Heart rate, beats/min 81.2 (12.9) 74.6 (14.0) -8.1 0.01 By paired t-test or Wilcoxon signed-ranks test. Urine ketone data were missing in a median of 4 participants (range 0–8) at any given visit. The proportion of participants with a urine ketone reading greater than trace was 1 of 17 participants at baseline, 5 of 17 participants at week 2, and similar frequencies at subsequent visits until week 14 when 2 of 18 participants had readings greater than trace and week 16 when 2 of 21 participants had readings greater than trace. During the study, only 27 of 151 urine ketone measurements were greater than trace, with one participant accounting for all 7 occurrences of the highest urine ketone reading (large160). In regard to serum measurements, the mean fasting glucose decreased by 17% from 9.08 ± 4.09 mmol/L at baseline to 7.57 ± 2.63 mmol/L at week 16 (p = 0.04) (Table 4). Serum sodium and chloride levels increased significantly, but only by 1% and 3%, respectively. Uric acid level decreased by 10% (p = 0.01). Serum triglyceride decreased 42% from 2.69 ± 2.87 mmol/L to 1.57 ± 1.38 mmol/L (p = 0.001). Increases occurred in both high-density lipoprotein (HDL) cholesterol (8%) and low-density lipoprotein (LDL) cholesterol (10%) but these changes were of borderline statistical significance (p = 0.08 and p = 0.1, respectively). The following blood tests did not change significantly: total cholesterol, potassium, bicarbonate, urea nitrogen, creatinine, calcium, thyroid-stimulating hormone, and hemoglobin. Table 4 Serum test results (n = 21) Measurement Week 0 Mean (SD) Week 16 Mean (SD) Change % p value Hemoglobin A1c, % 7.5 (1.4) 6.3 (1.0) -16.0 <0.001 Glucose, mmol/L 9.08 (4.09) 7.57 (2.63) -16.6 0.04 Total cholesterol†, mmol/L 4.61 (1.40) 4.54 (1.26) -1.5 0.7 Triglyceride†, mmol/L 2.69 (2.87) 1.57 (1.38) -41.6 0.001 HDL-C†, mmol/L 0.92 (0.20) 0.99 (0.22) 7.6 0.08 LDL-C†, mmol/L 2.51 (0.64) 2.77 (0.89) 10.4 0.1 Sodium, mmol/L 138.2 (3.4) 140.2 (3.4) 1.4 0.02 Potassium, mmol/L 4.2 (0.4) 4.2 (0.3) 0 0.2 Chloride, mmol/L 101.0 (3.4) 103.8 (2.8) 2.8 0.001 Bicarbonate, mmol/L 28.8 (2.0) 28.2 (2.5) -2.1 0.3 Urea nitrogen, mg/dL 5.90 (1.90) 6.27 (1.81) 6.3 0.3 Creatinine, μmol/L 82.8 (20.4) 79.9 (19.1) -3.5 0.3 Calcium†, mmol/L 2.32 (0.11) 2.33 (0.08) 0.4 0.8 Uric acid†, μmol/L 403.5 (94.6) 352.1 (85.4) -10.3 0.01 TSH†, mIU/L 1.6 (1.0) 1.4 (0.7) -12.7 0.2 Hemoglobin, g/L 142 (11) 141 (10) -0.7 0.8 Legend: TSH = thyroid-stimulating hormone, HDL-C = high-density lipoprotein cholesterol, LDL-C = low-density lipoprotein cholesterol. * By paired t-test or Wilcoxon signed-ranks test. † Serum calcium, TSH, total cholesterol, triglyceride, and HDL-C results are based on n of 20 because one participant did not have these tests at baseline. Uric acid is based on n of 19. LDL-C is based on n of 17 because triglyceride levels were too high at baseline to calculate LDL-C in 3 participants. The primary outcome, hemoglobin A1c, decreased from 7.5 ± 1.4% at baseline to 6.3 ± 1.0% at week 16 (p < 0.001), a 1.2% absolute decrease and a 16% relative decrease (Table 4). All but two participants (n = 19 or 90%) had a decrease in hemoglobin A1c (Figure 1). The absolute decrease in hemoglobin A1c was at least 1.0% in 11 (52%) participants. The relative decrease in hemoglobin A1c from baseline was greater than 10% in 14 (67%) participants, and greater than 20% in 6 (29%) participants. In regression analyses, the change in hemoglobin A1c was not predicted by the change in body weight, waist circumference, or percent body fat at 16 weeks (all p > 0.05). Figure 1 Hemoglobin A1c for each participant. *Red line is the group mean. P value is for the mean change from baseline. The improvement in glycemic control occurred while medications for diabetes were discontinued or reduced in most participants (Table 5). During the study, hypertension and hyperlipidemia medication doses were not increased from baseline nor were new agents added, except in 3 individuals. No serious adverse effects related to the diet occurred. One participant had a hypoglycemic episode requiring assistance from emergency services after he skipped a meal but the episode was aborted without need for transportation to the emergency room or hospitalization. Table 5 Diabetes medication changes Participant Week 0 Total daily dose Week 16 Total daily dose Participants with diabetes medications discontinued (n = 7 of 21) 5 glipizide 10 mg metformin 1000 mg none 6 glipizide 20 mg metformin 1500 mg none 7 metformin 2000 mg rosiglitazone 8 mg none 9 metformin 1000 mg none 15 metformin 1000 mg none 22 metformin 1000 mg none 24 metformin 1000 mg none Participants with diabetes medications reduced (n = 10 of 21) 3 70/30 insulin 50 units metformin 1000 mg metformin 1000 mg 11 metformin 2000 mg glyburide 20 mg metformin 2000 mg 16 metformin 2000 mg pioglitazone 45 mg glipizide 20 mg metformin 2000 mg 21 metformin 1500 mg pioglitazone 30 mg metformin 1000 mg 8 NPH 145 units metformin 1000 mg NPH 25 units metformin 1000 mg 13 70/30 insulin 70 units metformin 2550 mg 70/30 insulin 35 units metformin 2550 mg 23 70/30 insulin 110 units pioglitazone 45 mg 70/30 insulin 80 units pioglitazone 45 mg metformin 1000 mg 25 NPH 70 units, R 30 units metformin 2000 mg pioglitazone 45 mg NPH 8 units metformin 2000 mg pioglitazone 45 mg 27 70/30 insulin 86 units metformin 2000 mg 70/30 insulin 18 units metformin 2000 mg 28 NPH insulin 110 units lispro insulin 90 units glipizide 20 mg NPH insulin 30 units glipizide 20 mg Participants with diabetes medications unchanged (n = 4 of 21) 1 none none 2 metformin 1700 mg metformin 1700 mg 10 none none 26 metformin 2000 mg metformin 2000 mg Discussion In this single-arm, 4-month diet intervention, an LCKD resulted in significant improvement of glycemia, as measured by fasting glucose and hemoglobin A1c, in patients with type 2 diabetes. More importantly, this improvement was observed while diabetes medications were reduced or discontinued in 17 of the 21 participants, and were not changed in the remaining 4 participants. Participants also experienced reductions in body weight, waist circumference, and percent body fat but these improvements were moderate and did not predict the change in hemoglobin A1c in regression analyses. Several recent studies indicate that a low-carbohydrate diet is effective at improving glycemia. A few studies have shown that in non-diabetic individuals, low-carbohydrate diets were more effective than higher carbohydrate diets at improving fasting serum glucose [13,14] and insulin [6,14-16], and at improving insulin sensitivity as measured by the homeostasis model [6]. One of these studies also included diabetic patients and noted a comparative improvement in hemoglobin A1c after 6 months (low fat diet: 0.0 ± 1.0%; low carbohydrate diet: -0.6 ± 1.2%, p = 0.06) [6] and 12 months (low fat diet: -0.1 ± 1.6%; low carbohydrate diet: -0.7 ± 1.0%, p = 0.019) duration [5]. In a 5-week crossover feeding study, 8 men with type 2 diabetes had greater improvement in fasting glucose, 24-hour glucose area-under-the-curve (AUC), 24-hour insulin AUC, and glycohemoglobin while on the low-carbohydrate diet than when on a eucaloric low-fat diet [7]. In a 14-day inpatient feeding study, 10 participants with type 2 diabetes experienced improvements in hemoglobin A1c and insulin sensitivity as measured by the euglycemic hyperinsulinemic clamp method [8]. Hemoglobin A1c also improved in an outpatient study of 16 participants who followed a 20% carbohydrate diet for 24 weeks [9]. Similar to our results, three studies noted that diabetes medications were reduced in some participants[6,8,9], although details were provided in only one study. We also discontinued diuretic medications during diet initiation because of concern for additional diuresis incurred by the diet. This concern was based on the theoretical effects of the diet [17], observed effects of the diet on body water by bioelectric impedance [18], and practical experience with the diet [19]. Until we learn more about using low carbohydrate diets, medical monitoring for hypoglycemia, dehydration, and electrolyte abnormalities is imperative in patients taking diabetes or diuretic medications. While body weight decreased significantly (-8.5 kg) in these 21 diabetic participants, the mean weight loss was less compared with what we observed in the LCKD participants of an earlier trial (-12.0 kg) [18]. Given that the diabetic participants had a higher baseline mean weight than the LCKD participants of our previous trial (131 kg vs. 97 kg), this translates into an even more dramatic disparity in percent change in body weight (-6.6% vs. -12.9%). This lesser weight loss might result from several factors. First, in the current study, most of the participants were taking insulin and/or oral hypoglycemic agents that are known to induce weight gain[20,21] Second, these same agents, particularly insulin, inhibit ketosis, which is strived for in the earliest phases of the LCKD; while it remains unclear whether ketones actually play a role in weight loss on the LCKD, previous research in non-diabetic patients has shown a positive correlation between level of ketonuria and weight loss success [22]. Lastly, compared with our previous study the participants in the current study had more comorbid illness, lower socioeconomic status, and a shorter duration of follow-up (16 weeks versus 24 weeks), all of which are associated with reduced success on any weight loss program [23]. The main limitations of our study are its small sample size, short duration, and lack of control group. That the main outcome, hemoglobin A1c, improved significantly despite the small sample size and short duration of follow-up speaks to the dramatic and consistent effect of the LCKD on glycemia. For other effects, however, such as the rises in serum LDL and HDL cholesterol, the small sample size might be the reason statistical significance was not reached. Future studies of larger samples and containing a control group are needed to better address questions about the effect of the LCKD on serum lipids in patients with type 2 diabetes. Conclusion In summary, the LCKD had positive effects on body weight, waist measurement, serum triglycerides, and glycemic control in a cohort of 21 participants with type 2 diabetes. Most impressive is that improvement in hemoglobin A1c was observed despite a small sample size and short duration of follow-up, and this improvement in glycemic control occurred while diabetes medications were reduced substantially in many participants. Future research must further examine the optimal medication adjustments, particularly for diabetes and diuretic agents, in order to avoid possible complications of hypoglycemia and dehydration. Because the LCKD can be very effective at lowering blood glucose, patients on diabetes medication who use this diet should be under close medical supervision or capable of adjusting their medication. Competing interests Dr. Vernon has held a consulting relationship with Atkins Nutritionals, Inc. Authors' contributions WY conceived, designed, and coordinated the study; participated in data collection; performed statistical analysis; and drafted the manuscript. MF assisted with study design, performed data collection, and helped to draft the manuscript. AC analyzed the food records. MV assisted with study/intervention design and safety monitoring. EW participated in the conception and design of the study, and assisted with the statistical analysis. All authors read and approved the final manuscript. Acknowledgements Dr. Yancy is supported by a VA Health Services Research Career Development Award (RCD 02-183-1). ==== Refs Klein S Sheard NF Pi-Sunyer S Daly A Wylie-Rosett J Kulkarni K Clark NG Weight management through lifestyle modification for the prevention and management of type 2 diabetes: rationale and strategies. A statement of the American Diabetes Association, the North American Association for the Study of Obesity, and the American Society for Clinical Nutrition Am J Clin Nutr 2004 80 257 263 15277143 Arora SK McFarlane SI The case for low carbohydrate diets in diabetes management Nutr Metab (Lond) 2005 2 16 16018812 10.1186/1743-7075-2-16 Osler W McCrae T The Principles and Practice of Medicine 1923 New York, Appleton and Co. Allen FM Stillman E Fitz R Total dietary regulation in the treatment of diabetes: monograph No. 11 1919 New York, The Rockefeller Institute for Medical Research Stern L Iqbal N Seshadri P Chicano KL Daily DA McGrory J Williams M Gracely EJ Samaha FF The effects of low-carbohydrate versus conventional weight loss diets in severely obese adults: one-year follow-up of a randomized trial Ann Intern Med 2004 140 778 785 15148064 Samaha FF Iqbal N Seshadri P Chicano K Daily D McGrory J Williams T Williams M Gracely EJ Stern L A low-carbohydrate as compared with a low-fat diet in severe obesity N Engl J Med 2003 348 2074 2081 12761364 10.1056/NEJMoa022637 Gannon MC Nuttall FQ Effect of a high-protein, low-carbohydrate diet on blood glucose control in people with type 2 diabetes Diabetes 2004 53 2375 2382 15331548 Boden G Sargrad K Homko C Mozzoli M Stein TP Effect of a low-carbohydrate diet on appetite, blood glucose levels, and insulin resistance in obese patients with type 2 diabetes Ann Intern Med 2005 142 403 411 15767618 Nielsen JV Jonsson E Nilsson AK Lasting improvement of hyperglycaemia and bodyweight: low-carbohydrate diet in type 2 diabetes. A brief report Ups J Med Sci 2005 110 179 183 16075898 Yancy WS JrVernon MC Westman EC Brief report: a pilot trial of a low-carbohydrate, ketogenic diet in patients with type II diabetes Metabolic Syndrome and Related Disorders 2003 1 239 244 10.1089/154041903322716723 The Atkins Trial Kit Handbook: A Simple Guide to Doing Atkins 2001 Ronkonkoma, NY, Atkins Nutritionals, Inc. Durnin JVGA Womersley J Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years Br J Nutr 1974 32 77 97 4843734 10.1079/BJN19740060 Sharman MJ Gomez AL Kraemer WJ Volek JS Very low-carbohydrate and low-fat diets affect fasting lipids and postprandial lipemia differently in overweight men J Nutr 2004 134 880 885 15051841 Volek JS Sharman MJ Gomez AL DiPasquale C Roti M Pumerantz A Kraemer WJ Comparison of a very low-carbohydrate and low-fat diet on fasting lipids, LDL subclasses, insulin resistance, and postprandial lipemic responses in overweight women J Am Coll Nutr 2004 23 177 184 15047685 Meckling KA O'Sullivan C Saari D Comparison of a low-fat diet to a low-carbohydrate diet on weight loss, body composition, and risk factors for diabetes and cardiovascular disease in free-living, overweight men and women J Clin Endocrinol Metab 2004 89 2717 2723 15181047 10.1210/jc.2003-031606 Volek JS Sharman MJ Love DM Avery NG Gomez AL Scheett TP Kraemer WJ Body composition and hormonal responses to a carbohydrate-restricted diet Metabolism 2002 51 864 870 12077732 10.1053/meta.2002.32037 Denke MA Metabolic effects of high-protein, low-carbohydrate diets Am J Cardiol 2001 88 59 61 11423059 10.1016/S0002-9149(01)01586-7 Yancy WS JrOlsen MK Guyton JR Bakst RP Westman EC A low-carbohydrate, ketogenic diet versus a low-fat diet to treat obesity and hyperlipidemia: a randomized, controlled trial Ann Intern Med 2004 140 769 777 15148063 Vernon MC Mavropoulos J Yancy WS JrWestman EC Brief report: clinical experience of a carbohydrate-restricted diet: effect on diabetes mellitus Metabolic Syndrome and Related Disorders 2003 1 233 238 10.1089/154041903322716714 Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group Lancet 1998 352 837 853 9742976 10.1016/S0140-6736(98)07019-6 Purnell JQ Hokanson JE Marcovina SM Steffes MW Cleary PA Brunzell JD Effect of excessive weight gain with intensive therapy of type 1 diabetes on lipid levels and blood pressure: results from the DCCT. Diabetes Control and Complications Trial JAMA 1998 280 140 146 9669786 10.1001/jama.280.2.140 Westman EC Yancy WS Edman JS Tomlin KF Perkins CE Effect of six-month adherence to a very-low-carbohydrate diet program Am J Med 2002 113 30 36 12106620 10.1016/S0002-9343(02)01129-4 Yancy WS JrBoan J Adherence to Diet Recommendations Patient Treatment Adherence: Concepts, Interventions, and Measurement 2005 Mahwah, NJ, Lawrence Erlbaum Associates, Inc.	16318637	PMC1325029	CC BY	2021-01-04 16:24:11	no	Nutr Metab (Lond). 2005 Dec 1; 2:34	utf-8	Nutr Metab (Lond)	2,005	10.1186/1743-7075-2-34	oa_comm
==== Front Chiropr OsteopatChiropractic & Osteopathy1746-1340BioMed Central London 1746-1340-13-251631863210.1186/1746-1340-13-25ResearchAdolescent idiopathic scoliosis screening for school, community, and clinical health promotion practice utilizing the PRECEDE-PROCEED model Mirtz Timothy A [email protected] Mark A [email protected] Leon [email protected] Lawrence A [email protected] Cynthia G [email protected] Department of Health Sport and Exercise Science, University of Kansas, Lawrence, Kansas2 Division of Clinical Sciences, Texas Chiropractic College, Pasadena, Texas2005 30 11 2005 13 25 25 18 8 2005 30 11 2005 Copyright © 2005 Mirtz et al; licensee BioMed Central Ltd.2005Mirtz 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 Screening for adolescent idiopathic scoliosis (AIS) is a commonly performed procedure for school children during the high risk years. The PRECEDE-PROCEDE (PP) model is a health promotion planning model that has not been utilized for the clinical diagnosis of AIS. The purpose of this research is to study AIS in the school age population using the PP model and its relevance for community, school, and clinical health promotion. Methods MEDLINE was utilized to locate AIS data. Studies were screened for relevance and applicability under the auspices of the PP model. Where data was unavailable, expert opinion was utilized based on consensus. Results The social assessment of quality of life is limited with few studies approaching the long-term effects of AIS. Epidemiologically, AIS is the most common form of scoliosis and leading orthopedic problem in children. Behavioral/environmental studies focus on discovering etiologic relationships yet this data is confounded because AIS is not a behavioral. Illness and parenting health behaviors can be appreciated. The educational diagnosis is confounded because AIS is an orthopedic disorder and not behavioral. The administration/policy diagnosis is hindered in that scoliosis screening programs are not considered cost-effective. Policies are determined in some schools because 26 states mandate school scoliosis screening. There exists potential error with the Adam's test. The most widely used measure in the PP model, the Health Belief Model, has not been utilized in any AIS research. Conclusion The PP model is a useful tool for a comprehensive study of a particular health concern. This research showed where gaps in AIS research exist suggesting that there may be problems to the implementation of school screening. Until research disparities are filled, implementation of AIS screening by school, community, and clinical health promotion will be compromised. Lack of data and perceived importance by school/community health planners may influence clinical health promotion practices. ==== Body Background The PRECEDE-PROCEDE (PP) model is a planning model that provides structure for applying theories so the most appropriate intervention strategies can be identified and implemented [1,2]. The PP model has been utilized in public health and holds potential for use by school, community and clinical health promotion practice in preventing a number of disorders [3]. Adolescent idiopathic scoliosis (AIS) is defined as a lateral curvature of the spine greater than 10 degrees accompanied by vertebral motion and is the most common spinal deformity affecting children with the most common form being the idiopathic form [4-6]. The clinical diagnosis of AIS has not been assessed using the PP model. The lack of use of the PP model for program planning and implementation for AIS indicates an area of need. Increased public awareness and screening clinics has resulted in an increased number of children referred for orthopedic opinion yet, over-diagnosis and unnecessary treatment occurs [7-9]. Inconsistencies about AIS may be indicative of the potential value that the PP model has in fully evaluating the diagnosis for the purposes of program planning and implementation [9]. The purpose of this research is to study AIS in the school age population using the PP model and its relevance for community, school, and clinical health promotion [1]. Although community, school, and clinical health promotion are arguably distinct in their mission and approach, their commonalities bring them together under one model for analysis. Methods Green and Kreuter's textbook Health Promotion Planning: an Educational and Ecological Approach (Table 1) provides a framework for each of the phases involved in the PP model. The PP model guided the literature and pertinent resource search [1]. A literature search using the National Library of Medicine's MEDLINE aided in locating pertinent studies using keywords scoliosis, prevalence, quality of life, and screening, epidemiology, and policy. An evidence-based approach, i.e. the usage of the best available evidence from the scientific literature, was utilized in determining the value of the literature. Where quantitative literature was lacking usage of qualitative reports from what the authors concluded as credible sources were utilized. In areas where insufficient data existed, expert opinion from the author(s) was utilized as directed by the PP model's framework. Table 1 Phases and Descriptions of the PRECEDE-PROCEED model Phase Description I. Social assessment Identify and evaluate the social problems which impact the quality of life of a target population II. Epidemiological assessment Defined as program objectives which define the target population (WHO), the desired outcome (WHAT), and HOW MUCH benefit the target population should benefit, and by WHEN that benefit should occur III. Behavioral/environmental assessment Focuses on the systematic identification of health practices and other factors which seem to be linked to health problems IV. Educational/ecological assessment Selection of the factors which if modified, will be most likely to result in behavior change V. Administration/policy assessment Analysis of policies, resources and circumstances prevailing organizational situations that could hinder or facilitate the development of the health program; Assessment of the compatibility of program goals and objectives with those of the organization and its administration and its fit into the mission statements, rules and regulations. VI. Implementation of the program VII. Process evaluation Used to evaluate the process by which the program is being implemented. VIII. Impact evaluation Measures the program effectiveness in terms of intermediate objectives and changes in predisposing, enabling, and reinforcing factors. IX. Outcome evaluation Measures change in terms of overall objectives and changes in health and social benefits or the quality of life. It takes a very long time to get results and it may take years before an actual change in the quality of life is seen. Source: [1] Green & Kreuter, 1999; [14] Brown, 1999 Results Use of keywords scoliosis, scoliosis prevalence, and scoliosis screening along with terms PRECEDE-PROCEED and health promotion found no relevant studies using this model in relation to each other. The literature on social assessment of AIS related to the quality of life (QOL) is limited. Very few studies have approached the long-term effects of AIS as it relates to the person's QOL. Epidemiological reports suggest that AIS is the most common form of scoliosis seen in the childhood population and is the leading orthopedic problem seen in school age children. Behavioral and environmental reports related to AIS tend to focus on finding an etiologic relationship yet this data is confounded by the fact that AIS is not a behavioral diagnosis nor is it perceived as having any direct relationship to behavior. The Health Belief Model, the preferred method of the PP model, has not been utilized in any AIS research. The types of health behaviors that can be appreciated are illness behavior and parenting health behavior. These two behaviors may be identified with AIS in the attempt to promote positive self-esteem and good posture. The educational diagnosis is confounded because AIS is considered an orthopedic disorder instead of being related to a specific behavior. The administration and policy diagnosis is compromised by the fact that scoliosis screening programs have been determined to not be cost-effective. There is a wide variance of opinion of what the cost is to screen a child to rule out AIS. Policies are determined in some schools as based on the fact that in the United States, only 26 states mandate school scoliosis screening. The policy diagnosis is confused due to the varying opinions from professional research societies as to when to perform AIS screening. Furthermore, there exists the potential for error with the most common screening tool (Adam's test) resulting in over-diagnosis and inappropriate referral. Overall, the utilization of the PP model demonstrated where gaps exist in the research record. Discussion Throughout school health, community health, and clinical health practice, the quality of health care delivery needs to be provided within a framework using structure, process, and outcome measures [1]. The PRECEDE-PROCEED (PP) model was developed as a planning framework from which health and health promotion programs could be designed [1]. The PRECEDE model is based on the premise that an education diagnosis should precede an intervention just as a medical diagnosis precedes a treatment plan [10]. PRECEDE stands for "predisposing, reinforcing, and enabling factors in educational diagnosis and evaluation." PROCEED stands for "policy, regulatory, and organizational constructs in education and environmental development." This aspect of the model acknowledges the importance of environmental factors in determining behaviors [10]. The PP framework defines intervention development as a systematic process involving nine phases with the first five (social diagnosis, epidemiological diagnosis, behavioral and environmental diagnosis, educational and organizational diagnosis, and administrative diagnosis) involving the identification of health problems and their determinants through a series of diagnostic steps [11]. The PP model can be a useful guide to facilitate the characterization of the resources, educational and behavioral barriers, and organizational factors in a community and can serve to ensure that interventions and other learning opportunities are tailored to the needs and cultural values of the community [12]. The PP model can provide a useful way to organize research varying from clinician attitudes, beliefs, knowledge, and skills to a model for program planning, serving as a framework for the planning and implementation of health education activities aimed at priority areas [11,13]. In summary, the PRECEDE-PROCEED model begins with the outcome of interest and is used to design an intervention for achieving the desired outcome [10]. Social assessment of AIS Phase I of the PP model focuses on the identification and evaluation of a possible social problem, which may influence the quality of life (QOL) of the target population [14]. Quality of life is defined as the perception of individuals/groups that their needs are being satisfied and that they are not being denied opportunities to pursue happiness and fulfillment [1]. Bridwell et al reinforced the idea that AIS is of great concern to parents [15]. The parental concern was about surgery creating neurological defect, to reduce future pain, and disability as an adult [15]. Parental concern appeared to be potential dissatisfaction to see their child spend the rest of their life "as is" [15]. Cosmesis has been an important factor to consider in the treatment of adolescent patients with scoliosis [16,17]. Nonetheless, Edgar and Mehta noted that while surgery can improve appearance of patients with AIS, some patients were not always completely satisfied with the cosmetic result [18]. Untreated scoliosis affects the quality of life and can be a disabling disease in the adult [19]. Although back pain is a prominent concern, especially after age 30, the majority of adults are embarrassed by their deformity if left untreated [19]. Patients with AIS ten years after an orthopedic referral have experienced difficulty in lifting, walking and socializing [20]. Women, in particular, are less likely to marry if deformity is not corrected [19]. Females who have been given the diagnosis of AIS during adolescence have a poorer overall perception of health than do women without such a diagnosis [21]. Approximately 9% of girls will discontinue therapeutic brace wearing because of psychological distress related to the deformity around the hips [16]. Ryan and Nachemson discovered that scoliosis is more common in higher socioeconomic groups and indicates AIS as an equal opportunity disorder [22]. It has been suggested that AIS may lead to multiple physical and psychosocial impairments depending on its severity [23]. Previous studies have only assessed generic health measures, functional status, body image, and self-image [23]. Before 2001, no results of long-term outcomes in terms of health-related QOL had been executed for patients treated for AIS [24]. Most research had been directed to determine whether pain becomes a problem with significant scoliosis in individuals as they age [25]. Climent and Sanchez in their study of adolescents with spinal deformities contended that QOL variables include the Risser sign, clinical diagnosis, duration of brace treatment, and degree of correction [26]. These variables do not constitute a significant measurement of patient well-being, are more related to the diagnostic evaluation and do nothing to alter one's perception of happiness. Health educators, school nurses, and clinicians need to be aware of social well-being factors, and how these factors relate to psychosocial functioning. Epidemiological assessment of AIS The epidemiological diagnosis of the PP model constitutes phase II of the PRECEDE portion [14]. The primary task in this phase is to determine for a given target population which health problems, measured objectively, pose the greatest threat to health and QOL [1]. Planners use epidemiologic data to identify and rank the health problems [1]. This ranking is critical for planners because there is rarely enough resources to deal with all or multiple problems [14]. Scoliosis is classified by its etiology [27]. The most common form is the idiopathic form because no cause has been determined or associated with an already existing pathological state. Table 2 lists the classification of scoliosis by etiology. Idiopathic scoliosis can be classified into categories as based on age (Table 3). AIS is present in 2 to 4% of children between 10 and 16 years of age [6]. Classification of AIS can be based on the patient's age at onset, the rate of the curve progression as measured via radiographic by the Cobb angle, the magnitude of the scoliosis at the end of the growth phase, and the curve pattern [6]. Table 2 Types of scoliosis 1. Idiopathic scoliosis 2. Neuromuscular scoliosis 3. Congenital scoliosis 4. Neurofibromatosis 5. Connective tissue scoliosis 6. Osteochondrodystrophy 7. Metabolic scoliosis 8. Non-structural scoliosis Source: [27] Hu et al, 2000 Table 3 Types and age range of adolescent idiopathic scoliosis (AIS) 1. Infantile (under age 3 years of age) 2. Juvenile (from 3 to 10 years of age) 3. Adolescent (from 10 years of age to skeletal maturity) 4. Adult Source: [27] Hu et al, 2000 Table 4 Epidemiology of AIS Prevalence (curve of 10 degrees) 1.7% 1436 of 82,901 subjects Prevalence (curve of 10–19 degrees or more) 1.5% 1255 of 82,901 subjects Girls to boys (overall) 2.1:1 Girls to boys (curves less than 10 degrees) 1.5:1 Girls to boys (curves 10 to 19 degrees) 2.7:1 Girls to boys (curves 20 to 29 degrees) 7.5:1 Girls to boys (curves 30 to 39 degrees) 5.5:1 Girls to boys (curves 40+) 1.2:1 Most common curve of at least 10 degrees Thoracolumbar (34.3%) (n = 493) Second most common curve Lumber (33.1%) (n = 475) Third most common curve Thoracic (18.2%) (n = 261) Fourth most common curve Double curve (14.4%) (n = 207) Source: [30] Soucacos et al, 2000 Table 5 Risk factors for AIS Risk factor Curve progression Female gender Large curve magnitude Skeletal immaturity Female body type Much thinner Decreased body weight Decreased chest girth Decreased body mass index Source: [32] Remes et al, 2001; [39] Sugita, 2000; Historically, it was believed that the female to male ratio of AIS was 9:1 [28]. Al-Arjani et al found that 59% of all cases of scoliosis were idiopathic with the mean age of discovery at 12.5 years of age [29]. The mean Cobb angle was 58 degrees with 74% of the curves constituting the adolescent type of idiopathic curve with a thoracic curve being the most common with a female to male ratio of 3.8 to 1 [29]. Soucacos et al found that the prevalence of scoliosis was 1.7% of the population with most cases appearing at ages 13 and 14 with a small scoliotic curve in 1.5% of 10 to 19 degrees [30]. The etiology of idiopathic scoliosis may involve genetic, neuromuscular, hormonal, biomechanical, and other abnormalities [31]. The exact pathogenic mechanism of scoliosis is unknown [32]. Three possible etiologic theories still exist; (1) possible bone malformation during development; (2) asymmetric muscle weakness, and (3) abnormal postural control because of possible dysfunction of the vestibular system [33]. These classifications are not mutually exclusive; nor can any of these abnormalities directly produce lateral curves in the spine yet, whatever the underlying etiological factors are they eventually express themselves in the biomechanical changes associated with lateral curve progression [31]. Cummings et al suggested that the etiologic classification system is substantially reproducible but is only moderately reliable [34] while Lenke et al believed that the current system does not appear to have sufficient intra-observer or inter-observer reliability among scoliosis surgeons [35]. However, the diagnosis in pediatrics is directed or established, sometimes exclusively, by an extensive personal and family history and adequate interpretation, which in the end depends on the skill of the clinician [36]. The etiology classification AIS is still in doubt leading to the conclusion that it is multifactorial [37]. The main risk factors for curve progression are: (1) large curve magnitude; (2) skeletal immaturity; and (3) female gender [32]. A high risk of curve progression is usually associated with the following: girls before the onset of menses at around the time of pubertal growth spurt, right thoracic and double curves with magnitude of > or = 30 degrees [30]. Approximately 30% of the scoliotic deformities involve the thoraco-lumbar region whereas 48% and 22% of curves are confined to the thoracic or lumbar spines [38]. Sugita noted that females with scoliosis were much thinner, had a decreased body weight, chest girth, and body mass index [39]. LeBlanc et al supported this finding when it was found that adolescent girls with progressive AIS have a morphologic somatotype (less mesomorphic) that is different from the normal adolescent population [40]. Of adolescents diagnosed with scoliosis, only 10 percent have curves that progress and require medical intervention [6]. Although a small percentage of scoliotic curves undergo progression, the pattern of curve direction and the sex of the child has played a significant role in the ability to differentiate which curves will progress [30]. Behavioral and environmental assessment of AIS Phase III of the PP model focuses on the systematic identification of health practices and other related factors which appear to be linked to the identified health problem [14]. Confounding the behavioral and environmental diagnosis of is the fact that the etiology of scoliosis remains unknown in most cases despite extensive research [41]. Two behaviors (Table 6) can readily be identified and include the illness behavior and the parenting health behavior. Illness behavior is defined as an activity undertaken by an individual, who perceives they are ill, to define the state of their health and discover a suitable remedy [1]. Parenting health behavior is defined as wellness, prevention, at-risk, illness, self-care, or sick-role actions performed by an individual for the purpose of ensuring, maintaining, or improving the health of a child for whom the individual has responsibility [1]. Table 6 Health-related behaviors analogous to AIS Behavior Definition Illness behavior Activity undertaken by an individual, who perceives they are ill, to define the state of their health and discover a suitable remedy. Parenting health behavior Wellness, prevention, at-risk, illness, self-care, or sick-role actions performed by an individual for the purpose of ensuring, maintaining, or improving the health of a child for whom the individual has responsibility. Source: [1] Green & Kreuter, 1999 These behaviors (parenting health and illness behavior) may be applied to children with AIS. With reference to illness behavior, Frediel et al found that juvenile patients with AIS were unhappier with their lives due to more physical complaints and lower self-esteem [23]. Perceptions of body image, happiness, and satisfaction of adolescents with scoliosis is significant. Danielsson et al noted that those treated for AIS felt they were limited because of difficulties participating physically in activities and/or self-conscious about their appearance [24]. Sapountzi-Krepia et al found that females with scoliosis have a poorer perception of body image [42]. Only 5% of those with scoliosis declared that they had opportunities to discuss their feelings and problems with health professionals, while 90% of the declared that they wanted to have more opportunities to do this [42]. The potential mental anguish a child may endure from AIS is possible because the child can interpret that others perceive him differently as where he/she can see that he/she deviates in posture from his/her peers. The potential problem fully defining this as an illness behavior is the inability of the child to fully define their state of health and thus seek a remedy. Since AIS is commonly identified in children indicates a parental role (parenting health behavior). Bridwell et al found that when it came to their child about scoliosis surgery the parents concerns were naturally higher and expectations greater than that of the child undergoing the procedure [15]. Table 7, devised by expert opinion since the data is unavailable, establishes the behavioral prioritization and possible treatment focus in AIS. In addition to carefully selected surgical candidates, good family support, a proper educational environment, and promotion of independence at an early age are required to achieve maximal adult function [43]. Interventions should be devised to maintain or improve the person's orthopedic, pulmonary, and functional status as it applies to activities of daily living (ADL). [44] A great deal of experience, patience and the consideration of the patient's individual demands are inevitable for successful treatment [45,46]. Table 7 Behavioral prioritization in AIS Prevention Treatment 1. Prevent low self-esteem. 1. Cognitive therapy possibly. 2. Proper compliance if brace treated. 2. Compliance education. 3. Family low self-esteem. 3. Family support intervention. Growth and proper posture is an important environment in which spinal curves progress and peak prevalence rates occur at the ages of 11 and 13 years [47]. Changes in the intervertebral disc and endplate composition have been implicated as possible etiologic factors in the pathogenesis of AIS suggesting that scoliotic changes are due to an altered and ineffective synthetic response to a pathologic mechanical environment [48]. Proper posture improves QOL by stability in seating and standing by correction of pelvic obliquity and truck instability [49]. However, there is a strong association between back pain and smoking in people with scoliosis suggesting that smoking may have a greater impact on persons with injured spines [50,51]. Although smoking would aptly be described as a behavioral/lifestyle decision, with AIS it can be classified as an environmental factor as per the potential damage to an already injured spine. No studies have assessed adolescent smoking use and/or peer influence on those with AIS. Industrial environmental factors have not significantly influenced the prevalence of AIS [52]. Table 8, since the data is limited and established by expert opinion, offers the environmental prioritization for AIS which offers a possible role of preventive treatment. Table 8 Environmental prioritization in AIS Prevention Treatment 1. Proper posture 1. Spinal education 2. Reduce incidence of delayed menarche 2. Possible role of birth control; athletic education to avoid female triad 3. Proper compliance if brace treated 3. Compliance education of orthosis 4. Infection control if surgically corrected 4. Home education 5. Prevent future back pain 5. Back safety; child back pack education; tobacco prevention/cessation; Educational and ecological assessment of AIS The educational and ecological phase attempts to identify factors that necessitate change to initiate and sustain the process of behavioral change and thus become the immediate targets/objectives of programming intervention priorities [1]. Three areas are important to the educational assessment: predisposing factors, enabling factors, and reinforcing factors [1,14]. Predisposing factors include knowledge, attitudes, beliefs, personal preferences, existing skills, and self-efficacy toward the desired behavior change. Reinforcing factors include factors that reward or reinforce the desired behavior change [6]. Enabling factors are psychological/emotional or physical factors that facilitate motivation to change behavior [6]. The most widely used measure in the PP model, the Health Belief Model, has failed to be utilized in any scoliosis research. The authors cannot account for the lack of research with this model for the diagnosis of AIS. A research question that may need to be asked is "does AIS create behavioral problems? And if so, can the behavior be changed?" The Child Health Questionnaire Outcomes Data Collection instrument and the American Academy of Orthopedic Surgeons Pediatric Outcomes Data Collection instrument may need to be interpreted with a differing view [53]. Predisposing factors: AIS may lead to a more negative social affect as based on societal standards from a cosmetic point of view. Since females are mainly effected and society places emphasis on superficial appearance this may preclude towards the negative social affect. Enabling factors: Currently, twenty-six states have laws that mandate scoliosis screening, and other states without such laws may still provide state-supported screening programs or have screening programs conducted voluntarily in communities by local agencies [54]. Sugita believed it is necessary for school teachers and school nurses to pay more attention to a student's lifestyles since body mass index, chest girth, and body weight is significantly lower in females with AIS [39]. Yet, those with contact with potential AIS patients need to determine the value the community has for AIS. Those in school health settings need to be aware of mandatory screening statutes. For the patient with AIS and the nurse/clinician it is particularly important to define the early therapeutic prognosis because treatment can be long and difficult [55]. Lantz and Chen found that postural and lifestyle counseling has no discernable effect on the severity of curves as a function of age, initial curve severity, frequency of care, or attending physician [56]. Enabling factors confound the treatment of AIS and/or make difficult any type of positive bio-behavioral change [56]. Reinforcing factors: Confounding the reinforcing factor for reward is the negative connotation as being "labeled" as scoliotic thus leading to a psychosocial effect [57]. The need for increasing self-esteem for those afflicted with AIS, as determined from the behavioral assessment, appears validated. The factors noted previously might help or hinder health behavior(s). The question that needs to be forwarded is "what priorities for intervention(s) should be listed on these issues?" Three areas as it relates to AIS include the components of prevalence, immediacy, and necessity [1]. Prevalence can be answered from the epidemiological assessment, nonetheless, immediacy and necessity needs to be determined. The priorities must take into account what is known as it applies to the categories of importance and changeability. Table 9, established by expert opinion since the data in this area is unavailable, represents the prevalence, immediacy and necessity of AIS as a clinical diagnosis in the educational assessment. Changeability may be interpreted to not only mean change in the scoliotic presentation by orthosis or surgery but to potential change in negative perception of self. As per necessity, the evidence thus far, points to the mandatory need for screening of AIS in this population. Although learning and resource objectives have not been covered in AIS research, a developmental model may be arbitrarily advanced which targets the three categories of knowledge, beliefs, and skills. Table 10 (established by expert opinion) represents a model covering the target groups and may include not only the person with AIS, but the primary parental care giver, school nurse/health educator, and/or healthcare provider. Table 9 Priorities within categories of AIS Importance Priority Prevalence Females, age 9–11; 2 to 4% of general population Immediacy When females reach this stage; delayed menarche Necessity Mandatory depending on community and parental concern Changeability Priority Behavior Parents as caregivers; teachers as health educators; Increased awareness as to potential of problem in age group; social concern via cosmesis; Scoliotic deviation Dependent upon severity at time of diagnosis; treatment compliance; cosmetic effect; Table 10 Learning and resource objectives for AIS Problem: Teaching health educators on AIS; Problem: Clinical review for school nurses/physicians; Knowledge Understanding of natural history of AIS Increase appreciation for self-esteem issues as per children Identify high-risk groups Beliefs Elimination of prior misconceptions about AIS Development of a proper posture attitude Skills Identify and comprehend current policies, statutes, and research concerns Be able to perform the basic screening assessment and be able to utilize such assessments proficiently Administrative and policy assessment Phase V of the PP model takes into consideration the administration and policy aspects. This phase focuses on the administrative and organizational concerns which must be addressed prior to program implementation [14]. Green and Kreuter defined administrative diagnosis as the analysis of polices, resources and circumstances, and prevailing organizational situations that could hinder or facilitate the development of the health program [1]. Bott noted four basic criteria that are important when determining if AIS is amenable to detection and treatment by screening assessment methods. First, the diagnostic test used must be highly sensitive. Second, AIS should cause substantial morbidity or mortality. Third, early detection and treatment should eliminate the condition or prevent its progression. Finally, the incidence of AIS should be high enough to warrant the utilization of time, resources, and money [58]. The most notable policy in AIS pertains to the screening of children while they are in the age-range of vulnerability. Scoliosis detection through screening of school children is a technique that has been popularized over the last three decades and has been instituted in schools during the child's attendance in fifth to ninth grades. This effort was to achieve early detection of progressive AIS [59,60]. However, there is conflict among various authors and professional groups in formulating policy from the research record [61]. Nussinovitch et al concluded that screening programs for school age children coupled with subsequent follow-up procedures are worthwhile [62]. Wong et al suggested that screening of 11- to 12- and 13- to 14-year-old girls can identify a significant number who could benefit from treatment" [63]. The Scoliosis Research Society has recommended annual screening of all children age 10 to 14 years; the American Academy of Orthopedic Surgeons has recommended screening girls at the ages of 11 and 13 years and boys age 13 or 14; the American Academy of Pediatrics has recommended screening routinely at ages 10, 12, 14, and 16 years [64-67]. The Bright Futures guidelines recommend noting the presence of scoliosis during the physical examination of adolescents and children 8 years of age [68]. On the other hand, the US Preventive Services Task Force has concluded that there is insufficient evidence to recommend routine screening [69]. Evidently, there are interpretations from the research record as to what policy recommendations should be enacted. From an administrative perspective, conflicting opinion on recommendations can hinder implementation. Budgetary concerns for screening of AIS includes the concern for medico-legal issues, cost concerns, outcome measures, and patient preference issues that have not been completely accounted for previously require they be included in a school policy [70]. Health care costs as well as costs in loss of production can increase with the introduction of a clinical screening program [71]. This increase in health care costs can be due to factors such as over-diagnosis, inappropriate referral, and/or misinterpretation of the findings as based on the use of the testing protocol [72,73]. School scoliosis screenings have been reported to cost from as little as 6 cents per child to as much as $194 per child [74-76]. The lower estimate of cost only considered the cost to the school to implement and the higher cost attributed to children with curves of 5 degrees or more [74]. Koukourakis found that the cost of screening each child was estimated to be approximately $10 [38]. Soucacos et al confounded the financial cost figures by suggesting that screening has a negligible cost estimated at 30 cents per child [30]. Renshaw suggested that much valuable data can be secured from screening but costs of screening are not inconsequential and costs in follow-up procedures are high [73]. Table 11 and 12 provide the difference in costs for screening as well as other significant administrative factors. Clearly there is a considerable cost discrepancy and an accurate price tag on school-based screening programs in order that health care systems can allocate resources on a rational basis [77]. Table 11 Cost estimate for AIS Procedure Cost 1. Screening per child US $24.66 2. Child with curve of 20 degrees or more US $3,386.25 3. Child treated for scoliosis US $10,836.00 Source: [38] Koukourakis et al, 1997; [74] Yawn & Yawn, 2000; [75] Morais et al, 1985; [76] Lonstein et al, 1982; Table 12 Overview of AIS as it pertains to the administrative diagnosis Resource Procedure Time For children starting in the fifth grade Budget US6 cents to US$24.66 per child Personnel School nurse; physician Source: [38] Koukourakis et al, 1997 The screening tool of choice has been the Adam's forward bending test which tests for asymmetry via visual inspection [78,79]. Yawn and Yawn noted that school screenings identified some children who went on to receive treatment but referred many more who did not [74]. Challenges in scoliosis screening include the low prevalence rate of clinically significant scoliosis, the inverse relationship of sensitivity and specificity in the screening process because of the poor correlation of clinical deformity and radiographic abnormality, and the inflated cost of these programs because of overreferral [80]. Thus, Type I and Type II errors from the assessment measure can be implicated as the culprit in AIS screening. Before implementation of a health plan, it is important to know how it fits with the existing organizational mission, policies and regulations and possible recommendations [1]. Brown defined policy diagnosis as the assessment of the compatibility of a program's goals and objectives with those of the organization and its administration thus asking "does it fit into the mission statements, rules and regulations?" [14]. As noted, scoliosis screening is still under debate [81]. In the United States, only 26 states have laws that mandate scoliosis screening, and other states without laws may still provide state-supported screening programs or have screening programs conducted voluntarily in communities by local agencies [54]. In states where there are no mandates, individual schools may formulate policy voluntarily. Individual communities have the ability to decide what is appropriate and feasible for their schools based on the best available data [74]. It has been argued that that scoliosis screening is not cost-effective, but some still favor it as an integral part of preventative medicine [54]. Spinal screening appears to be effective in reducing the need for surgical treatment, but does not decrease the total cost of care for AIS [81]. Administratively, a redefinition of what actually constitutes a "significant" scoliosis for screening as well as the use of objective referral criteria and re-screening patients rather than referring those who have borderline cases can pose problems for local policy development [82]. Nonetheless, an informed physician/school nurse can make this assessment efficiently with a minimum cost to the family and hopefully reduce radiation exposure and an unwarranted referral [83]. The physician should be well aware of the local school's policy as well as the concerns that such a school has as per resources and budget as well as the community's concern. Table 13, established by expert opinion, demonstrates how the PP model can be utilized to ask questions concerning significant inter- and intra-professional factors. Table 13 Policy factors as it pertains to AIS Factor Problem Intra-professional Whose recommendation carries more weight? Who is most qualified to perform screenings? Who has the best available data? Inter-professional Unknown as research has not compiled political forces. Dependant upon community value/agenda; state law. Health promotion is inherently political [84]. Notwithstanding, political concerns must be based on accuracy of the research record and to not on unwarranted extrapolations of the research. Higginson demonstrated the politics of AIS screening from a physician's standpoint: "The legislative process is not necessarily a logical one in which good ideas turn into law; rather, success is usually based on relationships, timing, hard work, and luck. Parents also may perceive that screening is effective and insist that their children not be denied something they believe is valuable. Proactive work to educate and change opinion, such as a parent information campaign using the media, PTA's, and school officials can go a long way to reduce or remove potential grassroots opposition" [54]. Conclusion Health educator's, school nurses, as well a clinicians awareness of AIS and associated complications may possibly permit more effective patient surveillance, which may afford those at high risk the opportunity for an improved QOL [85]. However, this research using the PP model has found not only gaps in the research but conflicting views as to the value of AIS screening, the time to screen, cost concerns, as well as reliability of the most common screening tool (Adam's test). Despite adequate epidemiological research that suggests a problem the etiology of AIS remains in debate. Although AIS is the most common orthopedic disorder affecting children the previous problems noted also include state mandates and conflicting recommendations from the research record. The physician interested in such service to a school should have knowledge that screenings are ineffective due to examiner error, an assessment tool that is prone to error, and concern that cost-effectiveness for gaining an accurate outcome, as well as professional organizations lack of consistent recommendations exist to aid the decision-making. For effective implementation of a program a detailed evaluation of the pertaining literature and use of expert opinion is needed to fully appreciate the value of the PP model. Further research using the PP model's preference for the Health Belief Model with those afflicted with AIS along with the other gaps found by the PP model may need to become priorities in successfully developing pertinent learning and resource objectives for successful implementation of AIS programs. Competing interests The author(s) declare that they have no competing interests. Authors' contributions All authors contributed equally to this research. ==== Refs Green LW Kreuter MW Health Promotion Planning: An Educational and Ecological Approach 1999 3 Mountain View, CA: Mayfield Publishing Gielen AC McDonald EM Glanz K, Rimer BK, Lewis FM Using the PRECEDE-PROCEED planning model to apply health behavior theories Health Behavior and Health Education 2002 San Fransisco, CA: Jossey-Bass Publishers Antoniadis A Lubker BB Epidemiology as an essential tool for establishing prevention programs and evaluating their impact and outcome J Comm Disord 1997 30 269 283 10.1016/S0021-9924(97)00003-8 Goldberg CJ Dowling FE Fogarty EE Moore DP Adolescent idiopathic scoliosis as developmental instability Genetica 1995 96 247 255 8522164 10.1007/BF01439579 Been HD van Oiij A Veraart BE Slo GH One hundred years of orthopedics in the Netherlands. IV: spinal abnormalities Ned Tijdschr Geneeskd 1998 142 1009 1115 9623201 Reamy BV Slakey JB Adolescent idiopathic scoliosis: review and current concepts Am Fam Phys 2001 64 111 116 Weinstein SL Adolescent idiopathic scoliosis: prevalence and natural history Instruct Cours Lect 1989 38 115 128 Bollini G Jouve JL Lecoq C Garron E Idiopathic scoliosis: evaluation of the results Bull Acad Natl Med 1999 183 757 67 10437298 Dvonch VM Siegler AH Cloppas CC Bunch WH The epidemiology of "schooliosis" J Ped Orthop 1990 10 206 207 Ransdell RB Using the PRECEDE-PROCEED model to increase productivity in health education faculty Inter Elec J Health Educ 2001 4 276 282 Hecker EJ Feria de Salud: implementation and evaluation of a community-wide health fair Pub Health Nurs 2000 17 247 256 10943772 10.1046/j.1525-1446.2000.00247.x Watson MR Horowitz AM Garcia I Canto MT A community participatory oral health promotion program in an inner-city Latino community J Pub Health Dent 2001 61 34 41 11317603 Smith PG Danis M Helmick L Changing the health care response to battered women: a health education approach Fam Comm Health 1998 20 1 18 Brown KM The PRECEDE/PROCEDE Model. Community and Family Health 1999 University of South Florida Retrieved February 5, 2001 Bridwell KH Shufflebarger HL Lenke LG Lowe TG Betz RR Bassett GS Parents' and patients' preferences and concerns in idiopathic adolescent scoliosis: a cross-sectional pre-operative analysis Spine 2000 25 2392 2399 10984794 10.1097/00007632-200009150-00020 Korovessis P Stamatakis M Baikousis A Kirkos C Kavouris A Vertical transmission of the hip rolls due to wearing of TLSO for scoliosis J Spinal Disord 1996 9 326 333 8877961 Theologis TN Jefferson RJ Simpson AH Turner-Smith AR Fairbank JC Quantifying the cosmetic defect of adolescent idiopathic scoliosis Spine 1993 18 909 912 8316892 Edgar MA Mehta MH Long-term follow-up of fused and unfused idiopathic scoliosis J Bone Joint Surg (Brit) 1988 70 712 716 3192566 Fowles JV Drummond DS L'Ecuyer S Roy L Kassab MT Untreated scoliosis in the adult Clin Orthop 1978 134 212 217 729247 Mayo NE Goldberg MS Poitras B Scott S Hanley J The Ste-Justine adolescent idiopathic scoliosis cohort study: Part III: back pain Spine 1994 19 1573 1581 7939993 Goldberg MS Mayo NE Poitras B Scott S Hanley J The Ste-Justine Adolescent Idiopathic Scoliosis Cohort Study. Part I: Description of the study Spine 1994 19 1551 1561 7939991 Ryan MD Nachemson A Thoracic adolescent idiopathic scoliosis: perinatal and environmental aspects in a Swedish population and their relationship to curve severity J Ped Orthop 1987 7 72 77 Freidal K Petermann F Reichel D Steiner A Warschburge P Weiss HR Quality of life in women with idiopathic scoliosis Spine 2002 27 87 91 10.1097/00007632-200202150-00013 Danielsson AJ Wiklund I Pehrsson K Nachemson AL Health-related quality of life in patients with adolescent idiopathic scoliosis: a matched follow-up at least 20 years after treatment with brace or surgery Euro Spine J 2001 10 278 288 10.1007/s005860100309 Boninger ML Saur T Trefler E Hobson DA Burdett R Cooper RA Postural changes with aging in tetraplegia: effects on life satisfaction and pain Arch Phys Med Rehabi 1998 79 1577 1581 10.1016/S0003-9993(98)90424-5 Climent JM Sanchez J Impact of the type of brace on the quality of life of adolescents with spine deformities Spine 1999 24 1903 1908 10515014 10.1097/00007632-199909150-00007 Hu SS Carlson GD Tribus CB Disorders, Diseases, and Injuries of the Spine Skinner HB: Current Diagnosis & Treatment in Orthopedics 2000 New York: Lange Medical Books, McGraw-Hill Cailliet R Scoliosis: Diagnosis and Management 1986 Philadelphia: FA Davis Al-Arjani AM Al-Sebai NW al-Khawashki HM Saadeddin MF Epidemiological patterns of scoliosis in a spinal center in Saudi Arabia Saudi Med J 2000 21 554 557 11500705 Soucacos PN Zacharis K Soultanis K Gelalis J Xenakis T Beris AE Risk factors for idiopathic scoliosis: review of a 6-year prospective study Orthop 2000 23 833 838 Schultz AB Jacobs RR Biomechanical studies of possible causes for the progression of idiopathic scoliosis Pathogenesis of Idiopathic Scoliosis 1984 Chicago: Scoliosis Research Society Remes V Poussa M Peltonen J Scoliosis in patients with diastrophic dysplasia: a new classification Spine 2001 26 1689 1697 11474356 10.1097/00007632-200108010-00011 Kisner C Colby LA Therapeutic Exercise: Foundations and Techniques 1985 Philadelphia: FA Davis Cummings RJ Loveless EA Campbell J Samelson S Mazur JM Interobserver reliability and intraobserver reproducibility of the system of King et al for the classification of adolescent idiopathic scoliosis J Bone Joint Surg (Am) 1998 80 1107 1111 9730119 Lenke LG Betz RR Bridwell KH Clements DH Harms J Lowe TG Shufflebarger HL Intraobserver and interobserver reliability of the classification system of thoracic adolescent idiopathic scoliosis J Bone Joint Surg (Am) 1998 80 1097 1106 9730118 Lopez-Pison J Baldellou A Rebage V Arana T Lobera MO Pena-Sequra JL Demand for neuropediatric services at a general referral hospital IV: psychomotor development and physical examination Rev Neurol 1997 25 1905 1907 9580292 Nowakowski A Labaziewicz K Skrzypek H Idiopathic scoliosis: epidemiology and etiology Chir Narzadow Ruchu Ortop Pol 1998 63 317 320 9857548 Koukourakis I Giaourakism G Kouvidis G Kivernitakis E Blazos J Koukourakis M Screening school children for scoliosis on the island of Crete J Spinal Disord 1997 10 527 531 9438820 Sugita K Epidemiological study on idiopathic scoliosis in high school students: prevalence and relation to physique, physical strength and motor ability Nippon Koshu Eisei Zasshi 2000 47 320 325 10835893 LeBlanc R Labelle H Rivard CH Poitras B Relation between adolescent idiopathic scoliosis and morpholigic somatotypes Spine 1997 22 2532 2536 9383860 10.1097/00007632-199711010-00013 Tanchev PI Dzherov AD Parushev AD Dikov DM Todorov MB Scoliosis in rythmic gymnasts Spine 2000 25 1367 1372 10828918 10.1097/00007632-200006010-00008 Sapountzi-Krepia DS Valavanis J Panteleakis GP Zangana DT Vlachojiannis PC Sapkas GS Perceptions of body image, happiness and satisfaction in adolescents wearing a Boston brace for scoliosis treatment J Adv Nurs 2001 35 683 690 11529970 10.1046/j.1365-2648.2001.01900.x Carlson WO Speck GJ Vicari V Wenger DR Arthrogryposis multiplex congenita: a long term follow-up study Clin Orthop 1985 194 115 123 3978904 Bach JR Comprehensive rehabilitation of the severely disabled ventilator-assisted individual Monaldi Arch Chest Dis 1993 48 331 345 8257976 Hu SS Holly EA Lele C Averbach S Kristiansen J Schiff M Bueff HU Bradford DS Patient outcomes after spinal re-constructive surgery in patients >40 or = 40 years of age J Spinal Disord 1996 9 460 469 8976485 Heller KD Forst R Forst J Hengstler K Scoliosis in Duchenne muscular dystrophy: aspects of orthotic treatment Prosth Orthot Intl 1997 21 202 209 Smyrnis T Antoniou D Valavanis J Zachariou C Idiopathic scoliosis: characteristics and epidemiology Orthop 1987 10 921 926 Antonious J Arlet V Goswami T Aebi M Alini M Elevated synthetic activity in the convex side of scoliotic intervertebral discs and endplates compared with normal tissues Spine 2001 26 198 206 10.1097/00007632-200105150-00002 Gotze CM Hackenbergm L Liljenqvist U Halm H Surgical correction and stabilization of neuromuscular scoliosis: 2 to 4 year results of dorsal and one-stage ventro-dorsal operated patients Z Orthop Ihre Grenzgeb 2001 139 31 39 11253520 10.1055/s-2001-11868 Scott SC Goldberg MS Mayo NE Stock SR Poitras B The association between cigarette smoking and back pain in adults Spine 1999 24 1090 1098 10361658 10.1097/00007632-199906010-00008 Weis JC Betz RR Clements DH Balsara RK Prevalence of perioperative complications after anterior spinal fusion for patients with idiopathic scoliosis J Spinal Disord 1997 10 371 375 9355051 Grivas TB Samelis P Polyzois BD Giourelis B Polyzois D School screening in the heavily industrialized area – Is there any role of industrial environmental factors in idiopathic scoliosis prevalence? Stud Health Technol Inform 2002 91 76 80 15457698 Vitale MG Levy DE Moskowitz AJ Gelijns AC Spellman M Verdisco L Roye DP Capturing quality of life in pediatric orthopedics: two recent measures compared J Ped Orthop 2001 21 629 635 10.1097/00004694-200109000-00015 Higginson G Political considerations for changing medical screening programs J Am Med Assoc 1999 282 1472 1474 10.1001/jama.282.15.1472 Coillard C Leroux MA Zabjck KF Rivard CH Reducibility of idiopathic scoliosis during orthopedic treatment Ann Chir 1999 53 781 791 10584390 Lantz CA Chen J Effect of chiropractic intervention on small scoliotic curves in younger patients: a time-series cohort design J Manipul Physiolog Therap 2001 24 385 393 10.1067/mmt.2001.116419 Oh KS Chuah SL Harwant S The need for scoliosis screening in Malaysia Med J Malaysia 2001 56 26 30 11814244 Bott J Scoliosis J Flor Med Assoc 1989 76 465 468 McCarthy RE Prevention of the complications of scoliosis by early detection Clin Orthop 1987 222 73 78 2957138 Emans JB Scoliosis: diagnosis and current treatment Womens Health 1984 9 81 102 Skaggs DL Referrals from scoliosis screenings Am Fam Phy 2001 64 32 Nussinovitch M Finkelstein Y Amir J Greenbaum E Volovitz B Adolescent screening for orthopedic problems in high school Pub Health 2002 116 30 32 11896633 Wong HK Hui JH Rajan U Chia HP Idiopathic scoliosis in Singapore schoolchildren: a prevalence study 15 years into the screening program Spine 2005 30 1188 1196 15897834 10.1097/01.brs.0000162280.95076.bb Scoliosis Research Society Scoliosis: a handbook for patients 1986 Park Ridge, IL: Scoliosis Research Society American Academy of Orthopedic Surgeons Position statement: school screening programs for the early detection of scoliosis 1993 6 Park Ridge, IL: American Academy of Orthopedic Surgeons Bulletin 97 American Academy of Pediatrics Guidelines for health supervision II 1988 Elk Grove Village, IL: American Academy of Pediatrics U.S. Preventive Services Task Force Guide to Clinical Preventive Services 1996 2 Washington, DC: U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion Green M Bright Futures: guidelines for health supervision of infants, children, and adolescents 1994 Arlington, VA: National Center for Education in Maternal and Child Health Gilmer-Scott M Newton W Patients with mild scoliosis have good prognosis J Fam Pract 2003 52 451 2 12791226 Berg AO Screening for adolescent idiopathic scoliosis: a report from the United States Preventive Services Task Force J Am Bd Fam Pract 1993 6 497 501 Montgomery F Persson U Benoni G Willner S Lindgren B Screening for scoliosis: a cost-effective analysis Spine 1990 15 67 70 2109353 Weinstein MA McCabe JP Cammisa FP Postoperative spinal wound infection: a review of 2391 consecutive index procedures J Spinal Disord 2000 13 422 426 11052352 10.1097/00002517-200010000-00009 Renshaw TS Screening school children for scoliosis Clin Ortho 1988 229 26 33 Yawn BP Yawn RA The estimated cost of school scoliosis screening Spine 2000 25 2387 2391 10984793 10.1097/00007632-200009150-00019 Morais T Bernier M Turcotte F Age- and sex-specific prevalence of scoliosis and the value of school screening programs Am J Pub Health 1985 75 1377 1380 4061707 Lonstein JE Bjorklund S Wanninger MH Nelson RP Voluntary school screening for scoliosis in Minnesota J Bone Joint Surg (Am) 1982 64 481 488 6802853 Anonymous School scoliosis screening: new questions about its cost-effectiveness The Back Letter 2000 15 122 Katz DR A population-based study of school scoliosis screening Phyl Ther 2000 80 315 Katz DE Durrani AA Factors that influence outcome in bracing large curves in patients with adolescent idiopathic scoliosis Spine 2001 26 2354 2361 11679821 10.1097/00007632-200111010-00012 Bunnel WP Outcome of spinal screening Spine 1993 18 1572 1580 8235833 David R Jamal A Soudry M Screening for scoliosis in western Galilee schools Harefuah 1996 130 297 300 8707169 Bunnell WP Selective screening for scoliosis Clin Orthop Relat Res 2005 434 40 5 15864030 10.1097/01.blo.0000163242.92733.66 Staheli LT Spinal deformity J Fam Pract 1980 10 1071 1075 7373257 Nutbeam D Promoting health and preventing disease: an international perspective on youth health promotion J Adoles Health 1997 20 396 402 10.1016/S1054-139X(97)00009-8 Parisi MT Fahmy JL Kaminsky CK Malagolowkin MH Complications of cancer therapy in children: a radiologist's guide Radio 1999 19 283 297	16318632	PMC1325030	CC BY	2021-01-04 16:24:26	no	Chiropr Osteopat. 2005 Nov 30; 13:25	utf-8	Chiropr Osteopat	2,005	10.1186/1746-1340-13-25	oa_comm
==== Front BMC Womens HealthBMC Women's Health1472-6874BioMed Central London 1472-6874-5-111631367710.1186/1472-6874-5-11Research ArticleWomen and postfertilization effects of birth control: consistency of beliefs, intentions and reported use Dye Huong M [email protected] Joseph B [email protected] Stephen C [email protected] Han S [email protected] Patricia A [email protected] Department of Family and Preventive Medicine, University of Utah School of Medicine, Salt Lake City, UT, USA2 College of Nursing, University of Utah School of Medicine, Salt Lake City, UT, USA2005 28 11 2005 5 11 11 28 4 2005 28 11 2005 Copyright © 2005 Dye et al; licensee BioMed Central Ltd.2005Dye et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background This study assesses the consistency of responses among women regarding their beliefs about the mechanisms of actions of birth control methods, beliefs about when human life begins, the intention to use or not use birth control methods that they believe may act after fertilization or implantation, and their reported use of specific methods. Methods A questionnaire was administered in family practice and obstetrics and gynecology clinics in Salt Lake City, Utah, and Tulsa, Oklahoma. Participants included women ages 18–50 presenting for any reason and women under age 18 presenting for family planning or pregnancy care. Analyses were based on key questions addressing beliefs about whether specific birth control methods may act after fertilization, beliefs about when human life begins, intention to use a method that may act after fertilization, and reported use of specific methods. The questionnaire contained no information about the mechanism of action of any method of birth control. Responses were considered inconsistent if actual use contradicted intentions, if one intention contradicted another, or if intentions contradicted beliefs. Results Of all respondents, 38% gave consistent responses about intention to not use or to stop use of any birth control method that acted after fertilization, while 4% gave inconsistent responses. The corresponding percentages for birth control methods that work after implantation were 64% consistent and 2% inconsistent. Of all respondents, 34% reported they believed that life begins at fertilization and would not use any birth control method that acts after fertilization (a consistent response), while 3% reported they believed that life begins at fertilization but would use a birth control method that acts after fertilization (inconsistent). For specific methods of birth control, less than 1% of women gave inconsistent responses. A majority of women (68% or greater) responded accurately about the mechanism of action of condoms, abstinence, sterilization, and abortion, but a substantial percentage of women (between 19% and 57%) were uncertain about the mechanisms of action of oral contraceptives, intrauterine devices (IUDs), Depo-Provera, or natural family planning. Conclusion Women who believe that life begins at fertilization may not intend to use a birth control method that could have postfertilization effects. More research is needed to understand the relative importance of postfertilization effects for women in other populations, and in relation to other properties of and priorities for birth control methods. However, many women were uncertain about the mechanisms of action of specific methods. To respect the principles of informed consent, some women may need more education about what is known and not known about the mechanisms of action of birth control methods. ==== Body Background Many studies have investigated factors or barriers that are associated with women's use of birth control methods [1,2]. Using the theoretical framework of the Theory of Planned Behavior, some studies have investigated beliefs, attitudes, norms, and intentions that contribute to the use of specific contraceptive methods, either for contraception or the prevention of sexually transmitted infection [3-7]. One issue that has received little attention but that may be important for some women is understanding or beliefs and intentions about mechanisms of action of birth control methods. While there is dispute among medical experts about whether pregnancy begins at fertilization or implantation [8-10], up to half of women in national polls in the United States believe that human life begins at fertilization [11]. For these women, a birth control method that could act occasionally after fertilization may conflict with personal, ethical or religious beliefs [12]. As a result, these women may wish to refrain from using birth control methods that may exhibit a postfertilization effect, even if the effect were to occur prior to the woman recognizing that she was pregnant. A recent study found that Latina women in the United States who believed that emergency contraception acted after fertilization were less likely to use it [13]. However, we are unaware of any study that has systematically investigated how women's beliefs about the beginning of human life and the mechanisms of action of birth control methods may relate to their intentions to use and actual use of specific birth control methods. The primary purpose of our study was to determine among women beliefs about the mechanisms of action of birth control methods in relation to postfertilization effects and how this understanding related to their stated intentions to use methods of birth control, and reported use of specific methods. We aimed to do this by assessing among women: 1) beliefs about the mechanisms of action of select birth control methods, 2) personal belief of when human life begins, 3) intention to use birth control methods that may exhibit postfertilization effects, and 4) reported actual use of specific birth control methods. We assessed these by analyzing the consistency of responses to relevant questions of a questionnaire designed to address women's understanding and attitudes about the potential for postfertilization effects. Methods Questionnaire development The questionnaire used was a 4-page, 37-item survey assessing the following in women: 1) past, present and future use of birth control methods, 2) personal decision to use a birth control method if it acted at a certain stage of the reproductive process (described further below), 3) personal belief about what stage(s) specific birth control methods may act, and 4) personal belief about when human life begins. An introduction to the early stages of human reproduction was provided within the questionnaire and briefly explained in three stages. These stages were explained as follows: "Stage 1 (Before fertilization – before the uniting of the sperm and the egg), Stage 2 (After fertilization but before implantation – after the egg is fertilized but before it implants in the uterus; usually this time period is 5–7 days), and Stage 3 (After implantation in the uterus)." Key questions were focused around these stages. In constructing the questionnaire, we investigated all temporary methods of preventing unwanted birth (birth control), including abstinence, contraception, and abortion. For the purpose of assessing women's understanding of the early stages of reproduction described in the questionnaire, methods of birth control with unambiguous stages of action were designated in advance by the developers of the questionnaire. These were abstinence, condoms, sterilization, natural family planning (acting at Stage 1, prior to fertilization), and surgical or medical abortion (acting at Stage 3, after implantation). The questionnaire contained no information about how specific birth control methods work in relation to Stages 1, 2, or 3, but did contain questions asking women's own beliefs of how specific methods of birth control may act. The questionnaire was compiled and assessed for face validity through the collaboration of experts in reproductive medicine, demography, and population issues. Pilot testing of the questionnaire was performed with 21 patients in Salt Lake City, Utah, and 15 patients and 10 nurses in Tulsa, Oklahoma. The pilot testing consisted of having the patients or nurses fill out the questionnaire in writing (as it was designed to be used), followed by an interview to discuss their understanding of the reproductive stages and concerns about the questionnaire and specific understanding of key questions. Those completing the pilot testing reported understanding of the reproductive stages and issues addressed by the questionnaire. As a result of pilot testing, wording was clarified in the questions about intentions to use birth control methods, questions were added to include women's understanding of Depo-Provera, and whether her partner had or planned to have a vasectomy. The questionnaire is available from the authors on request. Study population The primary sites for this study included clinics in Salt Lake City, Utah and Tulsa, Oklahoma. In Salt Lake City, the sites included the Obstetrics and Gynecology clinic within the University of Utah Medical Center, a University of Utah family medicine clinic in the community, two private Obstetrics and Gynecology practices, and a Community Health Center. In Tulsa, Oklahoma, the site included was a teaching clinic associated with a family medicine residency program sponsored by a Protestant religious organization. None of the clinics in Utah were religiously affiliated or sponsored. The study questionnaire was presented by research assistants (male medical students or female clinic staff) to English-speaking women between the ages of 18 and 50 while they waited for appointments during regular clinic hours in early 2002. The questionnaire was also presented to women below the age of 18 if the appointment involved obstetrical care or birth control counseling. The research assistant explained the study and asked whether the woman would like to participate by filling out the questionnaire. After each woman agreed or declined, she had no further interaction with the research assistant. The questionnaire was self-administered in writing and took approximately 10–15 minutes to complete. The questionnaire contained a statement that filling out the questionnaire constituted voluntary participation in the study and did not affect clinical care provided. No identifying information was collected on participants in the study. The study was approved by the University of Utah Institutional Review Board (IRB), which did not require written consent of patients, or of parents for patients under 18, since these patients were presenting for clinical care that did not require parental consent. Statistical analyses Statistical analyses consisted of frequencies and cross-tabulations among the key questions. The intent was to determine the understanding, beliefs, and intentions of women concerning the postfertilization effects of specific methods of birth control in relation to their understanding and concern for postfertilization effects. Except when otherwise specified, all percentages reported in this paper are based on the denominator of all women included in the study. Results Demographics A total of 928 self-administered questionnaires were distributed at participating clinics in the study. Of these 928, a total of 748 were filled out and returned, thus resulting in an 81% response rate. Of these 748, a total of 130 women were removed from analyses because they were over the age of 50, or had a condition or surgery (menopause, hysterectomy, tubal ligation, etc.) that made them unable to become pregnant for the rest of their lives. This resulted in 618 eligible participants that were included in the study. About half (54%) of the participants came from clinics in Salt Lake City, UT, while the other half (46%) came from Tulsa, Oklahoma. The mean age of the study participants was 29 years. Most of the study participants were educated: 78% had a college degree or some college education, and 15% had a high school diploma. The majority (75%) were White, 4% were Black, 6% Hispanic, and 3% each were American Indian and Asian. A majority of the study participants were married (58%) and had a previous pregnancy (70%) or were currently pregnant (29%). A majority were religiously affiliated (71%), attended church at least once a week (50%), and many responded that their faith was a very important influence in their life (40%). The most commonly used methods of birth control by study participants were condoms and oral contraceptives (Table 1). Table 1 Current Use of Selected Birth Control Methods (n = 618) Birth Control Methods Percentage Using Now (or Near Future if Pregnant) (a) Condom 33.0 (29.3,36.7) Vasectomy 9.4 (7.1,11.7) Oral contraceptives 38.0 (36.7,44.5) Intrauterine device 3.6 (2.1,5.1) Depo-Provera 6.1 (4.2,8.0) Withdrawal 7.9 (5.8,10.0) Spermicide 2.4 (1.2,3.6) (a)Parentheses include 95% confidence interval limits. To assess the understanding of the mechanisms of action of birth control methods, women were asked to give their beliefs regarding at which stage(s) of human reproduction various methods of birth control may act (Table 2). In their responses to this question, women were reminded that some methods may act at more than one stage and encouraged to check all stages at which each method acts. A majority of women provided the designated single correct response for the following methods: 71% responded that abstinence acts at Stage 1, 84% responded that condoms act only at Stage 1, 68% responded that sterilization acts only at Stage 1, and 69% responded that abortion acts only at Stage 3 (note that percentages in Table 2 are higher because of inclusion of women that responded for more than one stage). There was less familiarity with medical abortion (RU-486), with only 39% of women responding that it acts at Stage 3, and natural family planning, with only 50% of women responding that it acts at Stage 1. With regard to beliefs about methods of birth control that were not designated in advance to have a single correct answer for mechanisms of action, a majority of women (70%) thought oral contraceptives act at Stage 1, but some (22%) also responded they act at Stage 2. Over half of women (56%) responded that emergency contraception acts at Stage 2. Other birth control methods that had high percentages that were marked "Unsure" by many women as to their mechanisms of action included progestin-only pills (57%), intrauterine devices (41%), and Depo-Provera (35%). Table 2 Personal Belief about Stages at which Birth Control Methods May Act (n = 618) Birth Control Methods Percentage Responding (a) Stage 1 (b) Stage 2 (b) Stage 3 (b) Unsure Abstinence 82.0 (79.0,85.0) 11.2 (8.7,13.7) 5.5 (3.7,7.3) 9.1 (6.8,11.4) Condoms 86.3 (83.6,89.0) 2.8 (1.5,4.1) 1.9 (0.8,3.0) 7.8 (5.7,9.9) Sterilization 75.6 (77.2,79.0) 7.8 (5.7,9.9) 5.5 (3.7,7.3) 15.7 (12.8,18.6) Abortion 6.2 (4.3, 8.1) 14.7 (11.9,17.5) 83.7 (80.8,86.7) 8.4 (6.2,10.6) RU-486 4.7 (3.0,6.4) 20.4 (17.2,23.6) 54.5 (50.6,58.4) 33.3 (29.6,37.0) Oral Contraceptives 70.2 (66.6,73.8) 22.3 (19.0,25.6) 3.6 (2.1,5.1) 18.5 (15.4,21.6) Progestin-only pills 27.4 (23.9,30.9) 13.9 (11.2,16.6) 1.8 (0.8,2.8) 57.3 (53.4,61.2) Intrauterine Devices 39.5 (35.6–43.4) 20.2 (17.0–23.4) 6.0 (4.1–17.9) 41.1 (37.2–45.0) Depo-Provera 54.9 (51.0,58.8) 9.1 (6.8,11.4) 2.4 (1.2, 3.6) 35.1 (31.3,38.9) Emergency Contraception 11.8 (9.3,14.3) 56.6 (52.7,60.5) 18.1 (15.1,21.1) 26.9 (23.4,30.4) Natural Family Planning 59.1 (55.2,63.0) 8.3 (6.1,10.5) 7.8 (5.7,9.9) 30.9 (27.3,34.5) (a) Percentages do not add to 100% due to the selection of more than one stage as a response, missing responses and rounding. Parentheses include 95% confidence interval limits. (b) Stage 1 – Before fertilization – before the uniting of the sperm and the egg; Stage 2 – After fertilization but before implantation; Stage 3 – After implantation in the uterus. Women were asked if they would consider using a birth control method that acts at Stage 2 or Stage 3. 53% responded "No" to using a birth control method that acts at Stage 2, and 74% responded "No" to using a birth control method that acts at Stage 3. To assess the underlying understanding of these questions about the stages of human reproduction, we analyzed the consistency of responses regarding intention to use a birth control method that acts at Stage 2 and Stage 3. If a woman answered that she would not use a birth control method that acts at Stage 2 of human reproduction, she would most likely not use a birth control method that acts at Stage 3. Half (51%) of all women responded consistently with "No" to using a birth control method acting at Stage 2 and "No" to using a birth control method acting at Stage 3. Less than 1% of women responded inconsistently with "No" to using a birth control method acting at Stage 2 and "Yes" to using a birth control method acting at Stage 3. Further evidence suggesting that a postfertilization effect may affect women's intention to use a birth control method is reflected in responses regarding intention to stop the use of birth control methods. Women were asked how their decision to use a birth control method would be affected if they were using a birth control method and learned that it acts at Stage 2 or Stage 3. 44% of women responded if there was even the remote possibility of the method acting at Stage 2, they would stop using it. 69% of women responded if there was even the remote possibility of the method acting at Stage 3, they would stop using it. If a woman answered she would not use a birth control method that acts at Stage 2 or Stage 3, then if she were to learn that she was using a birth control method that acts at Stage 2 or Stage 3 respectively, we expected that she would intend to stop using it. This would be a consistent response pattern. Of the entire group of respondents, 38% responded consistently that they would not use a method that acts at Stage 2 and they would stop using a method they were currently using if they learned it acts at Stage 2. An even higher percentage of respondents (64%) responded consistently that they would not use a method that acts at Stage 3 and that they would stop using a method they were currently using if they learned it acts at Stage 3. If a woman, however, answered that she would not use a birth control method that acts at Stage 2 or Stage 3, then if she were to learn that she was using a birth control method that acts at Stage 2 or Stage 3 respectively, but still answered she would continue using it, we considered this to be an inconsistent response pattern. Another inconsistent response pattern would be if a woman answered that she would use a birth control method that acts at Stage 2 or Stage 3, then if she were to learn that she was using a birth control method that acts at Stage 2 or Stage 3 respectively, she would stop using it. Of the respondents, 4% responded inconsistently to questions pertaining to Stage 2, and 2% responded inconsistently to questions pertaining to Stage 3. The personal belief of when human life begins for each woman is at the core of the question of using birth control methods that may act after fertilization. To assess belief of when human life begins, women were asked the open-ended question: "In your personal opinion, when does human life begin?" Most women reported the personal belief that human life begins at fertilization (48%). Some women had the belief that life begins after fertilization, but before implantation (5%), and some at implantation (5%) or later (14%). If a woman answered that human life begins at fertilization, she would most likely answer "No" to using a birth control method that acts at Stage 2 and at Stage 3. Of all the women in the study, 34% reported they believed that life begins at fertilization and would not use a birth control method that acts at Stage 2 nor use a method that acts at Stage 3, thus responding consistently. Conversely, 3% of all women responded to these questions inconsistently. If a woman answered that human life begins at implantation, she would most likely answer "Yes" to using a birth control method that acts at Stage 2 but "No" to using a birth control method that acts at Stage 3. Of all the women in the study, 2% reported they believed that life begins at implantation and that they would use a birth control method that may act at Stage 2 but not at Stage 3, thus responding consistently. Inconsistent responses to these questions were given by 1% of all women in the study. These women reported they believed that life begins at implantation but would not use a birth control method that may act at Stage 2 or Stage 3. Finally, we assessed the consistency of responses among women who stated the intention to not use any birth control method that acts at Stage 2 or Stage 3 with her personal belief of how specific birth control methods may act and her reported actual use of these same methods. Actual use was assessed in response to a question of which methods the woman was currently using or if she was pregnant, planning to use in the near future. In the Theory of Planned Behaviour (TPB), the most important determinant of behaviour is behavioural intention, which in turn, is affected by behavioural beliefs [14]. If a woman believed that a birth control method acts at Stage 2 or Stage 3 and did not intend to use a birth control method that acts at Stage 2 or Stage 3, she would most likely not be currently using or planning to use this method. This would be consistent. If a woman believed that a birth control method acts at Stage 2 or Stage 3 and answered "No" to using a birth control method that acted at Stage 2 or Stage 3, then if she was currently using or planning to use this method, this would be inconsistent with her beliefs and intentions. If a woman, however, did not believe that a birth control method acts at Stage 2 or Stage 3 and did not intend to use a birth control method that acted at Stage 2 or Stage 3, and then if she was currently using or planning to use this method, this would be consistent also. A higher percentage of women responded in this manner. The results to these analyses are given in Table 3, where the percentages reported are of all women analyzed in the study. The largest percentage of both inconsistent and consistent responses was reported for oral contraceptives pertaining to Stage 2. For oral contraceptives, 2% of respondents gave inconsistent responses while 24% of respondents gave consistent responses, 8% from women who were not using oral contraceptives and 16% from women who were using oral contraceptives. Table 3 Consistency of Beliefs, Intention to Use and Actual Use of Birth Control Methods Percentage Responding (a) Belief of how a specific method acts Acts at Stage 2 (b) Does not act at Stage 2 Intention to use any method that acts at Stage 2 (b) No Actual use of the specific method No Yes Response Consistency Consistent (c) Inconsistent (d) Consistent (e) Condoms 1.0 (0.2,1.8) 0.8 (0.1,1.5) 14.6 (11.8,17.4) Sterilization 4.1 (2.5,5.7) 0.3 (0.0,0.7) 4.7 (3.0,6.4) Oral Contraceptives 8.3 (6.1,10.5) 1.5 (0.5,2.5) 16.2 (13.3,19.1) Progestin-only pills 7.3 (5.2,9.4) 0.2 (0.0,0.6) 2.4 (1.2,3.6) Intrauterine Devices 10.4 (8.0,12.8) 0.0 (0.0,0.0) 1.1 (0.3,1.9) Depo-Provera 4.4 (2.8,6.0) 0.5 (0.0,1.1) 1.8 (0.7,2.9) Percentage Responding (a) Belief of how a specific method acts Acts at Stage 3 (b) Does not act at Stage 3 Intention to use any method that acts at Stage 3 (b) No Actual use of the specific method No Yes Response Consistency Consistent (c) Inconsistent (d) Consistent (e) Condoms 0.5 (0.0,1.1) 0.8 (0.1,1.5) 23.0 (19.7,26.3) Sterilization 4.5 (2.9,6.1) 0.2 (0.0,0.6) 6.8 (4.8,8.8) Oral Contraceptives 1.9 (0.8,3.0) 0.5 (0.0,1.1) 27.4 (23.9,30.9) Progestin-only pills 1.5 (0.5,2.5) 0.2 (0.0,0.6) 4.4 (2.8,6.0) Intrauterine Devices 5.7 (3.9,7.5) 0.0 (0.0,0.0) 2.8 (1.5,4.1) Depo-Provera 1.6 (0.6,2.6) 0.2 (0.0,0.6) 4.4 (2.8,6.0) (a) Percentages are of all respondents to questionnaire who gave the combined responses indicated in the particular column for each specific birth control method. Parentheses include 95% confidence interval limits. (b) Stage 2 – After fertilization but before implantation; Stage 3 – After implantation in the uterus. (c) This column includes respondents who gave the following internally consistent responses: they would not use a method that acted at Stage 2 or 3, believed that the specific method acted at Stage 2 or 3, and reported no actual use of the method. (d) This column includes respondents who gave the following internally inconsistent responses: they would not use a method that acted at Stage 2 or 3, believed that the specific method acted at Stage 2 or 3, and reported actual use of the method. (e) This column includes respondents who gave the following internally consistent responses: they would not use a method that acted at Stage 2 or 3, believed that the specific method did not act at Stage 2 or 3, and reported actual use of the method. Discussion These data demonstrate a strong consistency in the women studied for beliefs about postfertilization effects of specific birth control methods, beliefs about when human life begins, and the intention to use specific birth control methods. These findings are consistent with the Theory of Planned Behavior, which proposes a direct relationship between beliefs, intentions, and behavior [14], and specifically with previous studies of the Theory of Planned Behavior that have shown a consistency of beliefs, attitudes, intentions, norms, and behavior for other issues affecting contraceptive use, including overall attitudes about specific contraceptive methods, beliefs about the benefits, utility, and disadvantages of specific contraceptive methods, subjective norms about contraception, beliefs about self-control, and partners' attitudes [3-7,15]. In the Theory of Planned Behavior (TPB), an individual's behavioural beliefs and evaluations of behavioural outcomes are contributing factors that affect an individual's attitude toward the behaviours. This attitude toward the behaviour has an impact on behavioural intention, which in turn, affects behaviour. Applying the TPB, a women's personal beliefs concerning when human life begins may influence her intention to use birth control methods that may exhibit postfertilization effects, and therefore, her actual use or lack of use of methods that exhibit such effects. The strength of the association in our study suggests that for many of these women, the belief that human life begins at fertilization, the belief that a birth control method may sometimes work after fertilization or after implantation may be an important factor in intentions about which method of birth control to use, and in actual use. However, there are components of the Theory of Planned Behavior that our study did not directly address. These include attitudes, subjective norms, and self-efficacy. Contraceptive decision making in women is a complex process. A recent study has suggested a model for contraceptive decision making in women that involves three steps that are consistent with and complementary to the Theory of Planned Behavior [16]. These three steps are becoming aware, navigating a course and weighing what is best personally for each individual. This could be applied to the specific issues of postfertilization effects addressed by our study. As women become aware of the possibility of postfertilization effects, their intention to use a birth control method that may exhibit these effects may decrease if this conflicts with their moral and ethical beliefs. Similarly, the Health Belief Model, which is closely related to the Theory of Planned Behaviour, can be applied to this situation [14]. Applying the Health Belief Model, women who determine that specific birth control methods may have postfertilization effects that are in conflict with their moral or ethical beliefs (perceived severity) and are using such methods (perceived susceptibility) may develop negative attitudes toward these methods and discontinue their use (self-efficacy). In our questionnaire, the specific beliefs, intentions, and behaviors we measured were all single-item measures. Thus, this questionnaire was not a psychometric instrument, where different responses are presumed to be measuring the same underlying psychometric construct (such as stress or pain). Therefore standard approaches to assessing reliability (such as Cronbach's alpha) are not appropriate in this context. We also did not address the issues of test-retest reliability or response shift in this study. Future studies should attempt to elucidate more thoroughly the nature of the underlying beliefs, women's exact understanding of reproductive physiology, test-retest reliability, and response shift. In addition, future studies should examine other components of the Theory of Planned Behavior that we did not directly address: namely norms, attitudes, and self-efficacy. As presented in Table 2, we have used the term "birth control" as an umbrella term for methods that are used to prevent unwanted birth, including abstinence, natural family planning, contraception, emergency contraception, and abortion. While some advocates of reproductive rights consider each of these options and services related to them as legitimate and essential means to prevent unwanted births [17-20], others would accept some but reject others as being included appropriately in the term "birth control." Our use of the term "birth control" in this study is not meant to discount differences between these methods or approaches, nor to imply that they are all equally acceptable to women. Our purpose in this study was rather to investigate how women understand the mechanisms of action of the various interventions that can be undertaken to prevent unwanted birth, and the influence of this understanding on intentions. Because we were interested in studying birth control intentions among women who actually may face those choices, we excluded women who had undergone tubal ligation, who had undergone menopause, or who were otherwise sterile. The literature has focused on many properties considered by women when selecting a contraceptive method, such as whether a contraceptive method is highly effective, has a prolonged duration of action, is rapidly reversible, protects against sexually transmitted diseases, is easily accessible, is convenient/easy to use, and poses little to no side effects, risks, or disadvantages [21,22]. Some studies have considered that a disadvantage of a contraceptive method could be if it contradicted a woman's moral or religious beliefs, but have not specified such issues in more detail [23,24]. As noted, a recent study did find that many women were unwilling to use emergency contraception if they believed that its mechanism of action was after fertilization because this contradicted their moral beliefs [13]. In our study, this issue not only held true for emergency contraception, but for any birth control method that women believed had postfertilization effects as one of its mechanism of action. However, we did not study the relative weight that the potential for postfertilization effects may carry in relation to other contraceptive properties that have been studied more extensively. In our study there were few women who reported they did not intend to use a method with postfertilization effects, believed that a specific method had postfertilization effects, and still reported actual use of that specific method. Since the potential for postfertilization effects is important to many women in this study, it is also remarkable that many of them were unsure about the mechanisms of action of specific methods of birth control. Patients today have emphasized a need for more information from physicians and healthcare providers [25,26]. Though information concerning birth control methods and the potential for postfertilization effects may be difficult to provide due to uncertainty about the exact mechanisms of action and the exact frequency of these mechanisms of action for many methods of birth control, an attempt to present what is known and not known about the mechanisms of action of specific methods may be important for adequate informed consent for some women [27-33]. Without reasonably complete information, women may be denied the opportunity to make a fully informed decision during contraceptive selection. The present study has several limitations. The women included in the study are not a representative sample of all the women in the United States. A majority of women in our study were well educated and White. The religiosity of women in this particular clinic-based sample may not reflect that of the general United States population, because of the States in which the clinics are located (Utah and Oklahoma), and the religious affiliation of the clinic in Oklahoma. We recognize the limitations of the questionnaire used in the study. The questionnaire was developed for face validity by consultation with experts in the field and was pilot-tested to determine if the issues addressed were understood by women but has not been validated against other measures because we are unaware of other measures that specifically address these questions. We also recognize the controversy in asking the open-ended question "When does human life begin?" since this is a question that involves moral and religious values and a personal perception of life itself. Nevertheless, our research supports the proposal that beliefs about the beginning of human life are highly relevant to some women and their reproductive choices. In the pilot testing of the questionnaire, we found that women understood the early stages of human reproduction as presented, but we did not directly assess women's understanding of these issues in the main study sample. Without a prior knowledge of the stages of human reproduction, it may be a challenging task to assimilate and absorb new information and then attempt to apply it during the short time period of filling out the questionnaire. Patients may not have previously considered these issues prior to receiving the questionnaire. Some of them may have needed additional time to contemplate such sensitive issues prior to answering the questions that address such topics. Because of these issues, the rate of unsure and inconsistency of responses may be inflated in these data. Despite this limitation, the level of consistency in key responses within this questionnaire was high. Conclusion The strong consistency of responses for beliefs, intentions, and actual use of birth control methods in this study indicate that the potential for postfertilization effects may be important in some women's intention to use and actual use of specific birth control methods. Further research is needed to understand the relative importance of this issue in different populations and in relation to other priorities and properties of birth control methods. In the meantime, we believe that it is important that physicians and health care providers who prescribe birth control methods be willing to educate women on the basic processes of human reproduction and what is known about the mechanisms of action of birth control methods. For women who believe that life begins at fertilization, this education may be especially important since a birth control method that can exhibit a secondary postfertilization effect may conflict with their ethical or religious beliefs. This information may be necessary for some patients to make a fully informed decision during the process of birth control method selection. Competing interests The author(s) declare that they have no competing interests. Authors' contributions Huong M Dye designed and performed the analysis and wrote the manuscript. Joseph B Stanford conceived the study, supervised the analysis and revised the manuscript. Stephen C Alder, Han S Kim, and Patricia A Murphy provided additional guidance in formulating the analysis and revising the manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements We thank the other investigators who developed the questionnaire or collected the data, including Daniel Jones, Mark Christian, Craig DeLisi, Rafael Mikolajczyk, Kirtly Parker Jones, and Walter Larimore. ==== Refs Moreau C Bouyer J Goulard H Bajos N The remaining barriers to the use of emergency contraception: perception of pregnancy risk by women undergoing induced abortions Contraception 2005 71 202 207 15722071 10.1016/j.contraception.2004.09.004 Hogan DP Sun R Cornwell GT Sexual and fertility behaviors of American females aged 15–19 years: 1985, 1990, and 1995 Am J Public Health 2000 90 1421 1425 10983200 Villarruel AM Jemmott JB 3rdJemmott LS Ronis DL Predictors of sexual intercourse and condom use intentions among Spanish-dominant Latino youth: a test of the planned behavior theory Nurs Res 2004 53 172 181 15167505 10.1097/00006199-200405000-00004 Craig DM Wade KE Allison KR Irving HM Williams JI Hlibka CM Factors predictive of adolescents' intentions to use birth control pills, condoms, and birth control pills in combination with condoms Can J Public Health 2000 91 361 365 11089290 Libbus K Women's beliefs concerning condom acquisition and use Public Health Nurs 1995 12 341 347 7479543 Jemmott JB 3rdJemmott LS Hacker CI Predicting intentions to use condoms among African-American adolescents: the theory of planned behavior as a model of HIV risk-associated behavior Ethn Dis 1992 2 371 380 1490134 Weisman CS Plichta S Nathanson CA Chase GA Ensminger ME Robinson JC Adolescent women's contraceptive decision making J Health Soc Behav 1991 32 130 144 1861049 Spinnato JA Informed consent and the redefining of conception: a decision ill- conceived? J Matern Fetal Med 1998 7 264 268 9848690 10.1002/(SICI)1520-6661(199811/12)7:6<264::AID-MFM2>3.0.CO;2-M Grimes DA Cook RJ Mifepristone (RU 486) – an abortifacient to prevent abortion? N Engl J Med 1992 327 1088 1089 1522845 Moore KL Persaud TVN Before We are born: Essentials of Embryology and Birth Defects 1998 5 Philadelphia, P.A.: W. B. Saunders Company Zogby J American Values. Volume V 2000 Utica, New York: Zogby International Larimore WL Stanford JB Kahlenborn C Does pregnancy begin at fertilization? Fam Med 2004 36 690 691 author reply 691 15531977 Romo LF Berenson AB Wu ZH The role of misconceptions on Latino women's acceptance of emergency contraceptive pills Contraception 2004 69 227 235 14969671 10.1016/j.contraception.2003.10.020 Glanz K Lewis FM Rimer BK Health behavior and health education: theory, research, practice 2002 3 San Francisco: Jossey-Bass Inc Albarracin D Johnson BT Fishbein M Muellerleile PA Theories of reasoned action and planned behavior as models of condom use: a meta-analysis Psychol Bull 2001 127 142 161 11271752 10.1037/0033-2909.127.1.142 Noone J Finding the best fit: a grounded theory of contraceptive decision making in women Nurs Forum 2004 39 13 24 15700482 10.1111/j.1744-6198.2004.tb00010.x Yanda K Smith SV Rosenfield A Reproductive health and human rights Int J Gynaecol Obstet 2003 82 275 283 14499974 10.1016/S0020-7292(03)00226-1 Hessini L Global progress in abortion advocacy and policy: an assessment of the decade since ICPD Reprod Health Matters 2005 13 88 100 16035601 10.1016/S0968-8080(05)25168-6 Dolian G Ludicke F Katchatrian N Campana A Morabia A Contraception and induced abortion in Armenia: a critical need for family planning programs in eastern Europe Am J Public Health 1998 88 803 805 9585750 Kovacs L From abortion to contraception in Europe Eur J Contracept Reprod Health Care 1999 4 229 236 10817093 Grady WR Klepinger DH Nelson-Wally A Contraceptive characteristics: the perceptions and priorities of men and women Fam Plann Perspect 1999 31 168 175 10435215 Kost K Forrest JD Harlap S Comparing the health risks and benefits of contraceptive choices Fam Plann Perspect 1991 23 54 61 2060612 Lethbridge DJ Choosing and using contraception: toward a theory of women's contraceptive self-care Nurs Res 1991 40 276 280 1896325 Wall EM Valued outcomes in the selection of a contraceptive method West J Med 1984 141 335 338 6506671 Falvo DR Effective patient education: a guide to increased compliance 2004 3 Sudbury: Jones and Bartlett Publishers Mazur DJ Shared decision making in the patient-physician relationship 2001 Tampa: American College of Physician Executives Hatcher RA Trussell J Stewart F Stewart GK Kowal D Guest F Cates W JrPolicar MS Contraceptive technology. 18th revised edition 2004 18 New York Croxatto HB Devoto L Durand M Ezcurra E Larrea F Nagle C Ortiz ME Vantman D Vega M von Hertzen H Mechanism of action of hormonal preparations used for emergency contraception: a review of the literature Contraception 2001 63 111 121 11368982 10.1016/S0010-7824(01)00184-6 Larimore WL Stanford JB Postfertilization effects of oral contraceptives and their relationship to informed consent Arch Fam Med 2000 9 126 133 10693729 10.1001/archfami.9.2.126 Spinnato JA Mechanism of action of intrauterine contraceptive devices and its relation to informed consent Am J Obstet Gynecol 1997 176 503 506 9077596 10.1016/S0002-9378(97)70537-4 Kahlenborn C Stanford JB Larimore WL Postfertilization effect of hormonal emergency contraception Ann Pharmacother 2002 36 465 470 11895061 10.1345/aph.1A344 Stanford JB Mikolajczyk RT Mechanisms of action of intrauterine devices: update and estimation of postfertilization effects Am J Obstet Gynecol 2002 187 1699 1708 12501086 10.1067/mob.2002.128091 Mikolajczyk RT Stanford JB False risk attribution results in misleading assessment of the relationship between suppression of ovulation and the effectiveness of the Yuzpe regimen for emergency contraception Contraception 2003 67 333 335 author reply 335–337 12684157 10.1016/S0010-7824(02)00479-1	16313677	PMC1325031	CC BY	2021-01-04 16:30:36	no	BMC Womens Health. 2005 Nov 28; 5:11	utf-8	BMC Womens Health	2,005	10.1186/1472-6874-5-11	oa_comm
==== Front Biomed Eng OnlineBioMedical Engineering OnLine1475-925XBioMed Central London 1475-925X-4-651630955810.1186/1475-925X-4-65ResearchSuppression of AC railway power-line interference in ECG signals recorded by public access defibrillators Dotsinsky Ivan [email protected] Center of Biomedical Engineering, Bulgarian Academy of Sciences, Acad. G. Bonchev str., bl. 105, 1113 Sofia, Bulgaria2005 26 11 2005 4 65 65 3 10 2005 26 11 2005 Copyright © 2005 Dotsinsky; licensee BioMed Central Ltd.2005Dotsinsky; 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 Public access defibrillators (PADs) are now available for more efficient and rapid treatment of out-of-hospital sudden cardiac arrest. PADs are used normally by untrained people on the streets and in sports centers, airports, and other public areas. Therefore, automated detection of ventricular fibrillation, or its exclusion, is of high importance. A special case exists at railway stations, where electric power-line frequency interference is significant. Many countries, especially in Europe, use 16.7 Hz AC power, which introduces high level frequency-varying interference that may compromise fibrillation detection. Method Moving signal averaging is often used for 50/60 Hz interference suppression if its effect on the ECG spectrum has little importance (no morphological analysis is performed). This approach may be also applied to the railway situation, if the interference frequency is continuously detected so as to synchronize the analog-to-digital conversion (ADC) for introducing variable inter-sample intervals. A better solution consists of rated ADC, software frequency measuring, internal irregular re-sampling according to the interference frequency, and a moving average over a constant sample number, followed by regular back re-sampling. Results The proposed method leads to a total railway interference cancellation, together with suppression of inherent noise, while the peak amplitudes of some sharp complexes are reduced. This reduction has negligible effect on accurate fibrillation detection. Conclusion The method is developed in the MATLAB environment and represents a useful tool for real time railway interference suppression. ==== Body Background Public access defibrillators (PADs) are now available for more rapid and efficient treatment of out-of-hospital sudden cardiac arrest [1,2]. They are strongly recommended in The American Heart Association and the International Liaison Committee on Resuscitation Guidelines [3]. PADs are used normally by untrained people on the streets, and in theaters, sports centers, airports, and other public arenas. Therefore, automatic fibrillation detection, or its exclusion, is of high importance. Recently [4-6], special attention has been paid to the susceptibility of PADs to the electromagnetic field generated by the train overhead electric power-lines at railway platforms in countries using 16.7 Hz AC power, such as Switzerland, Germany, Austria, Norway and Sweden [4,5,7]. The interference induced may compromise the fibrillation detection [8,9] and increase the risk of inappropriate shock or of not supplying the life-saving defibrillation. Effective suppression of the railway interference is impeded by its large frequency band. The Official Journal of the European Communities [10] reports possible frequency deviations from 15.69 through 17.36 Hz. No data about the rate of frequency changes are available. Christov and Iliev [6] assume a modulation of 20% around 16.7 Hz per 10 s. There are no papers available on railway interference suppression except for the publication of Christov and Iliev [6] and the recent paper by Jekova and Krasteva [11]. Christov and Iliev [6] apply adaptive filtering based on proposed use of a special antenna within the PAD to feed the reference input of the filter. These authors present good results with simulated experiments of normal ECG signals, tachycardia, and ventricular fibrillation (VF). No signal distortions can be observed. However, the absence of a real embedded antenna is an important disadvantage of the method proposed. Jekova and Krasteva [11] modify some steps of the subtraction procedure [12,13] for railway interference suppression. The subtraction procedure is known to eliminate 50/60 Hz power-line interference without affecting the QRS spectra. It applies a moving average (comb filter) on linear segments (with frequency band near zero) to remove the interference components, which are stored and further subtracted from the ECG signal wherever non-linear segments are encountered. Jekova and Krasteva [11] band-pass filter the contaminated signal as a preliminary step of a sophisticated approach for linear segment detection inside VF, where these segments, if any, are extremely short. These segments are used for interference frequency measurement, necessary to compensate the fractional part of the inter-sample distance, which appears within a non-rated interference period. The obtained results are generally good. However, the QRS spectrum preservation as a necessary condition for accurate fibrillation detection is contestable. The algorithm is only checked in relatively narrow limits of interference frequency variation: from 16.5 through 17.3 Hz per 10 s. Linear segment detection is applied to signals with normal ECG activity, where no difficulties are expected; and in epochs with fibrillation, where compromises with the detection accuracy have no visible effect because of the low frequency spectra of the waves. No example with transition from normal activity to fibrillation is presented. The resulting amplitude errors are calculated by averaging the distortions over 1 s. Method Materials The study was carried out in the MATLAB environment. ECG recordings were taken from the AHA database, section 8. They contain normal cardiac rhythm followed by fibrillation. The sampling rate (SR) is 250 Hz. Sinusoidal oscillations within the band of the railway interference were synthesized and added to the ECG signals. The mixed signals were subjected to some filtrations and digital procedures for interference suppression. The results obtained are also valid for the harmonics and, therefore, may be extrapolate for any arbitrary waveform. The interference suppression as well as the differences between input and processed signals were assessed. The algorithm and program developed have a structure that simulates a real-time ongoing procedure. Assessment of some traditional filtrations applied for railway interference suppression Railway power-line interference may be suppressed by an appropriate notch filter, a moving average (comb filtering with linear phase characteristic), or other relatively simple techniques. All such procedures have the common disadvantage of affecting the ECG spectrum, especially to reduce the amplitude mostly of high and steep QRS complexes. However, it is very unlikely that such shape alternation would cause failure of the algorithms for fibrillation detection. The reason is that the variety of ECG signals is immense, including relatively low frequency low amplitude QRS complexes. Consequently, each processed signal may be assumed to be equal or very close to some non-processed signal. This statement will be supported below by the signals illustrating the proposed approach for 16.67 Hz interference suppression. Therefore, only the level of the residual interference is further assessed and taken into consideration as a limitation of fibrillation detection. The first trace of Fig. 1 shows an 8 s epoch of the AHA 8006 signal. It is assumed to be a 'clean' input signal, which is further mixed with interference of ± 1 mV amplitude and variable frequency from 13.3 trough 19.9 Hz (second trace). As can be seen, the residual interference after notch filtration may compromise the fibrillation detection. No difference between input and processed signal is shown here, as obviously it is too large to be compared with the differences obtained by other approaches. Figure 1 Band-stop filtering of contaminated ECG signal. The same input signal is used for estimation of simple moving averaging (Fig. 2). For odd sample number, n = 2m + 1 in one period of the interference, the ongoing middle filtered sample Yi is given by: Figure 2 Moving averaging (comb filtering) of the same contaminated ECG signal. Here Xi+j stands for the surrounding non-filtered samples. The processed signal (second trace) shows improved interference suppression mainly in the middle part of the epoch where the interference frequency coincides with the first zero of the comb filter. The third trace demonstrates partially eliminated interference together with clipped peaks of some sharp complexes. Adaptive and non-adaptive filters [14,15] are not considered because of their unacceptably long transient times appearing every time the ECG signal course changes. Proposed approach The second trace of Fig. 2 suggests that good suppression can be obtained if the first comb filter zero follows the interference frequency change. As the number of averaged values is integer, the inter-sample intervals have to be modified by adaptive SR of the analog-to-digital conversion (ADC). This approach has been developed as a part of the subtraction procedure [12,13] and implemented by hardware measuring of the interference frequency. However, this technique is not suitable for battery-supplied devices and computer-aided systems. Recently, Dotsinsky and Stoyanov published an alternative software method [16]. The contaminated signal is digitally band-pass filtered at -3 dB from 49 through 51 Hz. The amplitudes of two adjacent samples on a positive-going slope of the interference signal, located below and above the zero line, are measured. Then, the crossing point, CP, of the interference with the zero line is determined by interpolation using the two homogenous triangles defined by the left and right interference samples IL and IR (Fig. 3). The location of CP on the inter-sample distance dS is used to calculate the ongoing fluctuation of the interference period of repetition. Figure 3 Software measuring of the interference frequency. Although the adaptation of the inter-sample distances does not exceed the limit of 1% for 50/60 Hz interference, another solution proposed to eliminate the adaptive ADC, is preferred [17]. The rated ADC is applied without any adjustment. A moving averaging over a constant number of samples, n, is carried out. This average defines total suppression of interference with rated frequency. The ongoing frequency is software measured, and based on the result, an internal re-sampling is carried out. Linear interpolation is used to create new sample values with irregular inter-sample distances in contrast to the digitized input ECG signal but regularly spaced towards the rated interference frequency. Therefore, the interference is suppressed regardless of its original frequency. Then, the 'clean' ECG signal obtained is subjected to back re-sampling, restoring the rated (regular) inter-sample distances. Linear interpolation is used once more. This method is modified for railway interference cancellation to overcome its inherent large and fast frequency variations. Another difficulty derives from the almost total overlap between the spectra of the QRS complexes and the interference, which reduces the accuracy of the linear interpretation. The contaminated signal is processed by a first order Butterworth band-pass filter. When SR = 250 Hz, the recursive equation used is: Yn = 1.7168Yn-1 - 0.8529Yn-2 + 0.0735(Xn - Xn-2), (2) which defines the lower cut-off at 12.5 Hz and the upper cut-off at 18.8 Hz. Yn and Xn are the ongoing filtered and non-filtered samples, respectively. The other variables represent previous filtered and non-filtered samples. Furthermore, n is chosen equal to 15 so as to set the first zero of the comb filter at 16.67 Hz. The four steps of the two-way re-sampling are illustrated in Fig. 4. For a better observation, the signal is limited inside 2.5 and 3.8 s. Figure 4 Four steps of the two-way re-sampling. Both linear interpolations seem to be simple procedures. However, since the program written in the MATLAB language simulates a real-time ongoing process, each cycle begins with ongoing sample of the input signal, and the sequence of non-alternative forward and back interpolations requires a more sophisticated coordination of these steps. The possible cases are presented in Fig. 5. A pointer Pr is assigned to the regular ADC positions marked by 'X'. Two other pointers, Pf and Ps, control the locations 'O' and '' for first and second re-sampling, respectively. All pointers are incremented at the beginning of the cycles. Figure 5 Possible cases of interpolation. The upper part of Fig. 5 presents four possible cases of the first re-sampling C1-C4. C1 is the simplest case of interpolation. Pr indicates the position t3. The sample to be calculated, 'O', is inside the two top ADC samples of Pr located at t2 and t3. Its value depends on the two top ADC samples amplitudes and their distance from 'O'. If the last 'O' address of Pf (case C2) is on the left of the interval defined by two consecutive top 'X' addresses of Pr (case C2, t2and t3), this pointer is decremented (the two 'X' addresses are shifted at t1 and t2) before the interpolation is accomplished in the next cycle (when all pointers will be incremented). Furthermore, C3 and C4 show the last Pf address on the right of the interval t2-t3. Pr is incremented before the interpolation in the current cycle, if possible (case C3 with subsisting ADC sample at t4). Otherwise, in case of C4, Pf is decremented without interpolation to hold the address inside the interval in the next cycle. The second re-sampling, shown in the lower part of Fig. 5, follows the same rules of the first re-sampling (cases C1-C3), except for the last case, C4. Here the procedure is unrealistic because the program begins after a certain delay of Pf and Ps is reached in respect to Pr. In contrast to the software measurement of 50/60 Hz interference [16,17], the influence of wide and high QRS complexes in case of 16.67 Hz interference is an order of magnitude higher. This influence is reduced by clipping the calculated inter-sample intervals beyond a level as defined by a percentage of the weighted sum of the two last intervals. Thus the measured values are restricted to be within expected possible fluctuations, which are better related to the current interference frequency. A constant clipping level will introduce erroneously different allowed rates of frequency changes for the ranges around 13 and 19 Hz. The efficiency of the approach can be seen in Fig. 6, where the black trace represents the preliminary calculated course of the synthesized interference while the red trace is for the measured frequency. Figure 6 Comparison between measured and preliminary calculated course of the synthesized interference. In summary, the proposed approach consists of rated ADC, software frequency measuring, internal irregular re-sampling according to the interference frequency, and moving averaging over a constant number of samples followed by regular back re-sampling. Results The features of the developed algorithm may be assessed in the following figures, where no residual noise in the processed signal can be seen. The second trace of Fig. 7 shows a total railway interference cancellation together with suppression of inherent signal noise, e.g., muscle disturbances inherent to the 'clean' input signal. The two last traces prove the independence of the method with respect to the direction of the interference frequency change. Figure 7 Result of applying the proposed method on the ECG signal of Figs. 1 and 2. Two other AHA signals, chosen to demonstrate the transition from normal heart activity to fibrillation, can be observed in Figs. 8 and 9. The second signal includes high and steep QRS complexes, and has superimposed interference with lower amplitude, which is the worst case of using software frequency measurement of interference [17]. The shown differences between input and processed sharp complexes are to some extent due to the slight shifting within the processed signals. It is a consequence of the irregular transition time (group delay), introduced by the band-pass filtering of signal parts with relatively rapid amplitude change. This effect may be better seen in Fig. 10, where the first trace represents the difference between input and processed signals of Fig. 7, while the second trace points out the delay introduced between them. Figure 8 Processing of signal, which presents transition from normal heart activity to fibrillation. Figure 9 High and steep QRS complexes superimposed by interference with lower amplitude, which is the worst case of using the software frequency measurement. Figure 10 Effect of the slight shifting of the processed signals because of the high dynamics of the amplitude change that leads to irregular group delay after the band-pass filtering. The signals presented in Fig. 11 and Fig. 12 are chosen to test the method with VF, which has a higher frequency spectrum and smaller amplitude. Figure 11 Epoch extracted from the AHA 8007 signal with higher frequency spectrum and smaller amplitude. Figure 12 Epoch extracted from the AHA 8008 signal with higher frequency spectrum and smaller amplitude. Let us compare the two first traces in Fig. 7. If the non-filtered signal has a given probability for correct VF detection, than this probability must be the same for the filtered signal (second trace) despite the smoothed peak immediately after the 3rd s. In the present study, this statement may be only visually supported by knowledge and experience in the analysis of ECG signals. The alternative approach consists of testing a representative set of fibrillation detection algorithms on various ECG signals superimposed by 16.67 Hz interference. However, such a study is out of the aim of this paper. Besides, it is very time-consuming because of the enormous volume of experiments needed for a good statistic. The same considerations may be supported looking at Fig. 9. The filtered third trace has preserved sufficient parameter differences between QRS complexes and fibrillation waves for a correct detection of transition to fibrillation. These signals, as well others tested but not shown here, confirm the conviction that the moving averaging does not compromise the detection of fibrillation. Discussion and conclusion Our algorithm developed in the MATLAB environment is a useful tool for real-time railway interference suppression. The time required for off-line analysis of an 8 s epoch with the program, structured to simulate a real-time going process, is less than 4 s. The shape deviations for high and steep QRS complexes are assumed to be negligible for accurate fibrillation detection. These deviations are due to the moving averaging and suppressed noise inherent to the 'clean' signals taken from AHA database. The real differences between 'clean' and processed contaminated signals are smaller than those shown in the Figures, where the non-identical group delay of the band-pass filtering results in shifting the signals relative to each other. It may be possible that the real differences can be further reduced if 4th or 8th point Lagrange interpolation is used. This possibility is tested with 50/60 Hz interference [17], where it is found to have insignificant increase on the accuracy, perhaps because of the relatively narrow range of interference frequency fluctuations. Acknowledgements This study has been supported by Schiller AG. ==== Refs Cummins RO Ornato JP Thies WH Pepe PE Improving survival from sudden cardiac arrest: the 'chain of survival' concept: a statement for health professionals from the Advanced Cardiac Live Support Subcommittee and the Emergency Cardiac Care Committee. American Heart Association Circulation 1991 83 1832 1847 2022039 Cobb LA Eliastam M Kerber RE Melker R Moss AJ Newell L Paraskos JA Weaver WD Weil M Weisfeldt ML Report of the Anerican Heart Association Task Force on the Future of Cardiopulmonary Resuscitation Circulation 1992 85 2346 2355 1591856 The American Heart Association in collaboration with the International Liaison Committee on Resuscitation (ILCOR) Guidelines 2000 for cardiopulmonary resuscitation and emergency cardiac care - an international consensus on science. Part 4: automated external defibrillator Resuscitation 2000 462 73 91 Schlimp CJ Breiteneder M Lederer W Safety aspects for public access defibrillation using automated external defibrillators near high-voltage power lines Acta Anaesthesiol Scand 2004 48 595 600 15101855 10.1111/j.1399-6576.2004.00370.x Kanz KG Kay MV Biberthaler P Russ W Wessel S Lackner CK Mutschler W Susceptibility of automated external defibrillators to train overhead lines and metro third rails Resuscitation 2004 62 189 198 15294405 10.1016/j.resuscitation.2004.02.018 Christov II Iliev GL Suppression of 16.7 Hz interference generated by the power supply of the railway systems Biomed Eng OnLine 2005 4 16 15766390 10.1186/1475-925X-4-16 Angerer M Cardiac pacemaker and railroad net systems Herzschrittmachertherapie und Elektrophysiologie 2004 15 40 46 10.1007/s00399-004-0410-4 Einav S Weissman C Karl J Lotan C Matot I Future shock: automatic external defibrillators Curr Opin Anaesthesiol 2005 18 175 180 16534335 10.1097/01.aco.0000162837.79215.a7 Schlimp CJ Breiteneder M Lederer W Influence of electromagnetic interference on the human electrocardiogram in automated external defibrillators Resuscitation 2005 66 105 106 15993736 10.1016/j.resuscitation.2005.01.015 Commission decision concerning the technical specification for interoperability relating to the energy subsystem of the trans-European high-speed rail system Official journal of the European communities 2002 45 280 369 Jekova I Krasteva V Subtraction of 16.67 Hz railroad net interference from the electrocardiogram: application for automatic external defibrillators Physiol Meas 2005 26 987 1003 16311447 10.1088/0967-3334/26/6/009 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 Levkov C Mihov G Ivanov R Daskalov I Christov I Dotsinsky I. 'Removal of power-line interference from the ECG: a review of the subtraction procedure BioMed Eng OnLine 2005 4 50 16117827 10.1186/1475-925X-4-50 Thakor NV Zhu YS Applications of adaptive filtering to ECG analysis: noise cancellation and arrhythmia detection IEEE Trans Biomed Eng 1991 38 785 793 1937512 10.1109/10.83591 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 10.1109/10.477707 Dotsinsky I Stoyanov T Power-line interference cancellation in ECG signals Biomed Instr Techn 2005 39 155 162 Dotsinsky I Removal of frequency fluctuating power-line interference from ECG Proceedings of the 3rd European Medical & Biological Engineering Conference: 20–25 November Prague 11	16309558	PMC1325032	CC BY	2021-01-04 16:25:30	no	Biomed Eng Online. 2005 Nov 26; 4:65	utf-8	Biomed Eng Online	2,005	10.1186/1475-925X-4-65	oa_comm
==== Front BMC GastroenterolBMC Gastroenterology1471-230XBioMed Central London 1471-230X-5-391631647410.1186/1471-230X-5-39Research ArticleOutcome following emergency surgery for refractory severe ulcerative colitis in a tertiary care centre in India Pal Sujoy [email protected] Peush [email protected] Girish K [email protected] Subrat K [email protected] Tushar K [email protected] Department of GI Surgery & Liver Transplantation, All India Institute of Medical Sciences, Ansari Nagar, New Delhi-110029, India2 Department of Gastroenterology & Human Nutrition Unit, All India Institute of Medical Sciences, Ansari Nagar, New Delhi-110029, India2005 30 11 2005 5 39 39 11 6 2005 30 11 2005 Copyright © 2005 Pal 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 Steroid-based intensive medical therapy for severe ulcerative colitis is successful in 60–70% of such patients. Patients with complications or those refractory to medical therapy require emergency colectomy for salvage. Little is known about the impact of timing of surgical intervention and surgical outcomes of such patients undergoing emergency surgery in India where the diagnosis is often delayed or missed in patients who are poor, malnourished and non-compliant to medical treatment. Methods The clinical records of all patients undergoing emergency surgery for severe ulcerative colitis or its complication in the Department of GI surgery AIIMS, New Delhi, India, between January 1985 and December 2003 were retrieved and data pertaining to demographic features, duration of intensive medical therapy, presence of complications, time from admission to emergency surgery, surgical procedure, in-hospital morbidity and mortality and follow up status extracted. Results A total of 72 patients underwent emergency surgery (Subtotal colectomy: 60; ileostomy alone under local anaesthesia: 12). Poor nutritional status was seen in 61% of the patients. Twenty-one patients (29%) underwent emergency surgery for complications of severe ulcerative colitis such as colonic perforation (spontaneous 6, iatrogenic 4), massive lower gastrointestinal haemorrhage (5), toxic megacolon (4) and large bowel obstruction (2). The remaining patients (n = 51) underwent emergency surgery following failed intensive therapy; 17 underwent surgery ≤5 days (Group I) and 34 were operated >5 days (Group II) after initiation of intensive therapy. In this group all the post-operative deaths (n = 8) occurred in those who were operated after 5 days. The difference in mortality in these two groups (i.e. surgical intervention ≤ or >5 days) was statistically significant {0/17 (Group I) vs 8/34 (Group II); p = 0.03}. Overall, 12 patients died (in-hospital mortality: 16.7%). The mortality was higher (10/43; 23.3%) in our early experience (i.e. 1985–1995) when compared to our subsequent experience (2/29; 6.9%) (1996–2003). A total of 48 patients (including 3 awaiting a restorative procedure) are alive on follow up (66.7%; 3 patients lost to follow up). A restorative procedure could be successfully completed in 81% of the survivors of the emergency procedure. Conclusion To optimize the outcome, a combined team of physicians and surgeons should be involved in the management of patients with severe ulcerative colitis with focus on nutritional support, correction of metabolic derangements, close clinical monitoring and timely assessment for the need for emergency surgery. This retrospective analysis shows that improved results can be achieved with experience and by following a policy of early surgical intervention within 5 days, especially in patients who have failed intensive medical therapy. ==== Body Background It is known that about 20% of patients with ulcerative colitis, suffer from severe acute relapse and its associated complications at some point in the course of their illness [1,2]. Overall, 5–10% of patients with ulcerative colitis present with a severe first attack. The relapse-free interval in a given patient remains unpredictable. However, long-term prospective cohort studies have shown that virtually all patients will relapse and develop acute exacerbation at some point in the course of their illness [1,2]. Intensive medical treatment (Oxford regimen) is successful in 60%–70% of patients suffering from severe ulcerative colitis, but the rest require emergency surgery either for a complication or because they fail to respond to medical therapy [3-6]. Among those who respond to intensive medical therapy, a colectomy will be required in up to 50% within a year and in at least 75% at around 5 years of follow up [6,7]. Emergency surgery in anaemic, nutritionally depleted, immunosuppressed and toxic patients has a high morbidity and mortality [8,9]. Little is known about the outcome of patients with acute severe ulcerative colitis undergoing emergency surgery in India. We hypothesized that Indian patients with ulcerative colitis may behave differently because they are poor (have difficulty in purchasing drugs) and often illiterate (do not understand the importance of continuing medication and follow up) and the diagnosis is often delayed because of the high incidence of infective diarrhoea and dysentery. We therefore analyzed our experience of emergency surgery in patients with ulcerative colitis focusing on two main issues, i.e. the timing of surgery and the immediate as well as long term outcome following emergency surgery. Methods All patients undergoing emergency surgery for severe ulcerative colitis or its complication in the Department of GI Surgery, All India Institute of Medical Sciences, New Delhi, between January 1985 and December 2003 had a pre-designed proforma filled. The details recorded included demographic features of the patients, duration of intensive medical therapy during the index admission, presence of complications (toxic megacolon, perforation, lower gastrointestinal haemorrhage, intestinal obstruction and colorectal cancer), time from initiation of intensive therapy to emergency operation, indication for emergency surgery, type of surgical procedure, postoperative complications, in-hospital mortality, delayed deaths during follow up, outcome of any further operative procedures and current status. The collected data was retrospectively analyzed. A severe episode of ulcerative colitis was defined according to Truelove and Witt's criteria [3] as: >6 bloody stools/day, fever ≥ 38°C, tachycardia > 100/min, anemia and/or an erythrocyte sedimentation rate > 30 mm in 1st hour.[3] All patients diagnosed to have severe ulcerative colitis were admitted in the Gastroenterology department of our institution and were managed jointly, from the time of the admission, by a team comprising of physicians and surgeons. Nutritional status at admission was judged by a combination of subjective clinical assessment, body weight record (labeled underweight when <10th percentile of their expected weight as per sex and height) and serum albumin levels <3.0 g/dl was considered low. Intensive medical therapy included parenteral steroids (100 mg hydrocortisone intravenously 6 hourly), nil by mouth, nasogastric aspiration, intravenous crystalloids and electrolytes, parenteral nutritional support, blood transfusions and broad spectrum parenteral antibiotics (third-generation cephalosporins, aminoglycosides and metronidazole). The indications for emergency surgery included lack of response to or deterioration while on intensive medical therapy and presence of complications at admission. Patients who continued to have high stool frequency (>6 motions/day), persistent hematochezia, deteriorated clinically, developed steroid toxicity and/or complications and/or remained intolerant to oral feeding were judged to have failed to respond to intensive medical therapy. In our practice (before this study), duration of intensive therapy did not exclusively influence the decision to operate or label a patient refractory to intensive therapy. Operative procedures A subtotal colectomy with Hartmann's pouch (STC) was done in most patients. This included removal of the diseased ascending, transverse, descending and the sigmoid colon with closure of the sigmiodorectal stump at or above the level of sacral promontory without any pelvic mobilization of the rectum. When the rectal stump was unhealthy because of severe rectal disease or when there was ongoing rectal bleeding the open end was brought out as a mucus fistula through the anterior abdominal wall. Some patients underwent an initial loop ileostomy under local anaesthesia, as they were considered too ill to withstand a colectomy under general anaesthesia. Subsequently, in the same hospital admission or at an early readmission these patients underwent a subtotal colectomy. All patients were counseled regarding the long-term benefits/disadvantages of a restorative proctectomy with an ileal pouch (J reservoir) anal anastomosis (IPAA) [10,11]. Depending on their motivation and acceptance a restorative procedure with a diverting loop ileostomy was then done 3–6 months after the emergency colectomy. Finally, 6 weeks later, the ileostomy was closed after obtaining a pouchogram. Follow up protocol Patients were followed up at 3-monthly intervals in the first year, 6-monthly intervals in the second year and at yearly intervals thereafter. Patients who defaulted were sent postal reminders and questionnaires to ascertain their current status. At each visit the patients underwent a general physical examination, their weight was noted, laparotomy wound and ileostomy site were inspected and patients who had had the ileostomy closed were interviewed regarding their bowel frequency and continence. Periodic complete blood counts and liver function tests were also done. The primary outcome measure was operative mortality. The factors which could have influenced the patient outcome such as duration of disease, long term use of steroids or azathioprine, presence of fever, co morbid illnesses, preoperative hemoglobin, total leucocyte count and serum albumin levels were also recorded. We also analyzed the surgical outcome in relation to duration of intensive medical management in the index admission. For this purpose we divided the patients who were considered failures of intensive medical therapy into 2 groups: those who underwent surgery ≤ and >5 days after the initiation of intensive medical therapy. Statistical methods The information was entered in a MS Access format and analyzed using the SPSS version 11.5 software. For comparison, non-parametric tests such as Pearson's chi square test and for continuous variables student's t test were used. A p value less than 0.05 was considered significant. A univariate analysis was done to determine any association of operative mortality with duration of intensive medical therapy, age, preoperative nutritional status duration of disease, long term use of steroids, presence of fever, co morbid illnesses, preoperative hemoglobin, total leucocyte count and serum albumin levels. To identify a cut-off point with regard to duration of intensive medical therapy a receiver-operator characteristic (ROC) curve was plotted and the point on the curve giving the highest sensitivity for the primary outcome measure(postoperative death) was used for the subsequent analysis. Results Between 1985–2003, 154 (88 males) patients underwent operations for ulcerative colitis out of whom 72 (46.5%) underwent emergency surgery. The mean age was 34.8 ± 12.8 years (range: 14–72 years; median: 32 years). The nutritional status of 61% of these patients was poor and the mean serum albumin level was 2.6 g/dl (range: 1.3–4.8 g/dl). Failure of intensive medical therapy was the most common indication for emergency surgery (n = 51; 71%) patients. In these patients the interval between initiation of intensive medical therapy at our hospital and emergency surgery varied between 3–35 days with a mean of 9 days (median: 6 days). The remaining 21 (29%) patients had presented with complications such as colonic perforation (spontaneous 6, iatrogenic 4), massive lower gastrointestinal haemorrhage (5), toxic megacolon (4) and large bowel obstruction (2). A majority of these patients underwent urgent/emergent surgery even before a trial of intensive drug therapy could be initiated. Surgery The most common emergency surgical procedure was a sub total colectomy with a Hartmann's pouch or mucus fistula (n = 60; 83.3%). A diverting loop/divided ileostomy alone (under local anaesthesia) was the initial procedure in 12 patients. Ileostomy was more frequently used as the initial procedure in the later period of our experience (i.e., 1996–2003; 10/29 patients vs. 2/43; p = 0.003). Postoperative morbidity The overall morbidity included major wound sepsis with or without wound dehiscence (27; 37.5%), fever (20; 27.8%), adhesive intestinal obstruction (8.3%), pelvic sepsis (6.6%), ileostomy-related complications (6.6%), blow out of Hartmann's pouch (6.6%) and continued rectal bleeding (3.3%). The median duration of postoperative hospital stay was 10 days. Postoperative mortality Overall, 12 (STC: 10/60; Ileostomy: 2/12) patients died postoperatively in the same admission giving an in-hospital mortality of 16.7%. Eight patients died in the sub group operated following failure of intensive medical therapy (8/51; 15.7%) whereas 4 died following surgery for complications (4/21; 19%). Notably, all patients with perforations alone (n = 10) survived the emergency surgery. The causes of death included septicaemia in 6, severe chest infection with respiratory failure in 2, disseminated intravascular coagulation in 1, diabetic ketoacidosis in 1, metabolic encephalopathy in 1 and upper gastrointestinal bleeding (stress bleeding) in 1. By using chi square tests it was found that the risk of death increased in those patients who had postoperative fever (p = 0.01), ileostomy-related complications (p = 0.02) or evidence of peritonitis at surgery (p = 0.018). Also, the patients who died had lower preoperative albumin levels (mean albumin: 2.25 ± 0.5 g/dl vs 2.73 ± 0.8 g/dl; p = 0.08). We also found that the mortality was higher (10/43; 23.3%) in our early experience (i.e. 1985–1995) when compared to our subsequent experience (2/29; 6.9%) (1996–2003). Timing of surgery Twenty-one patients had emergency surgery for complications and in 51 patients the indication was failure of intensive medical therapy. In the 21 patients undergoing surgery for complications the occurrence of the complication determined the timing of the surgery. In this group, often surgery was done shortly after admission (i.e. in patients with perforation, severe bleeding or obstruction). On analyzing the data of the patients who underwent emergency surgery for failed intensive therapy (n = 51), statistical analysis with regard to mortality as the primary outcome measure showed that the only variable that influenced the outcome significantly was the timing of surgery from the initiation of intensive medical therapy. A cut-off time period of 5 days was obtained after plotting a ROC curve with death as the primary outcome measure. In this group 17 patients underwent surgery ≤ 5 days and the rest (n = 34) were operated >5 days after initiation of intensive therapy. The operative outcome of each group was further analyzed. All the post-operative deaths (n = 8) occurred in those who were operated after 5 days (Table 1). The difference in mortality in the two groups based on the timing of surgery (i.e. ≤ or >5 days) was statistically significant {0/17 (Group I) vs 8/34 (Group II); p = 0.03}. These two groups were otherwise comparable in terms of the nutritional status, age, sex, and associated co-morbidities. Table 2 summarizes the salient clinical, demographic and laboratory data pertaining to these two groups of patients who failed intensive medical therapy. The groups did not show any significant difference in any of the variables compared. Long term outcome Two patients died at home (6 weeks; 8 weeks) after being discharged alive from hospital of which one patient who had severe associated comorbidity in the form of valvular heart disease with infective endocarditis, died of a suspected cerebral thromboembolic phenomenon and another patient died of disseminated colorectal cancer. The ultimate outcome of each patient presenting to us for emergency surgery is summarized in Figure 1 and Table 3. All surviving patients have been followed up for a mean period of 39 months (range: 3–216 months). Of the 58 (12 postoperative and 2 late deaths) patients who survived the emergency procedure, 3 patients were lost to follow up, 2 refused a second stage IPAA and 2 were not offered because of extensive small bowel tuberculosis in one and low rectal cancer in another. Hence, 51 patients were available for a 2nd stage restorative proctectomy. In these, IPAA was done in 46 patients, one patient is alive and well following an ileorectal anastomosis (IRA), one other patient is alive and well following a total proctocolectomy (TPC) with Brooke's ileostomy (a 65 year old patient who had poor anal sphincter tone opted for a TPC) and 3 are awaiting a restorative surgery. One patient died following the pouch procedure and there were 2 late deaths on long term follow up. A total of 48 patients (including 3 awaiting a restorative procedure) are alive on follow up, from the initial cohort of 72 patients (66.7%; status of 3 patients lost to follow up not known). The restorative procedure (IPAA, IRA, TPC) could be successfully completed in 81% of the survivors of the emergency procedure (47/58 with 1 postoperative death). Of these 45 patients are alive and doing well. Discussion Before 1950, the expected mortality of a patient with acute severe ulcerative colitis was as high as 40–50% [1-3]. Such a patient was a physician's nightmare, as the surgeons never came in the picture. Truelove and Witts introduced steroid therapy in the 1950s [3,4]. With this, the mortality decreased to 8–10%, but rates as high as 18% have been reported [4-6]. The last milestone in therapy came in the 1960's after Brooke and Sampson proposed emergency total colectomy for salvage of patients following failure of medical therapy [12]. The timing of the emergency surgery is of crucial importance for these patients, as procrastination leads to worsening of the patient's general condition and nutritional status. To begin with, nearly two-thirds of the patients in this series were poorly nourished and hypoalbuminemic before surgery. It has been shown previously that the surgical outcome worsens with delay beyond 5–7 days of intensive medical therapy [8,13]. Our data supports this view, as the operative mortality in the group of patients who failed intensive medical therapy and were operated late (i.e., beyond 5 days) was high when compared to the group operated within 5 days (8/34 vs. 0/17; p = 0.03). In fact, all the post-operative deaths (n = 8) occurred in patients who were operated on after 5 days. This difference in mortality in the two groups based on timing of surgery is statistically significant. A univariate analysis of this group could not identify any other significant risk factor for mortality (Table 2). In 46 patients (i.e., 64% of the total patient population n = 72), multiple factors were responsible for a delay >5 days in surgical intervention. Some patients had shown an initial response to intensive medical therapy but had a relapse or developed a complication in the same hospitalization; some had associated comorbid conditions in the form of heart disease, uncontrolled diabetes or tuberculosis while some were referred late from other hospitals. In patients who present with complications such as perforation or massive colonic bleeding, emergency surgery should be performed as soon as possible. In patients with toxic megacolon a short period of conservative management (24–48 hours) may be beneficial [2,14]. In patients with large bowel obstruction usually 48–72 hours of preoperative preparation is necessary to optimize the surgical outcome. Similar to other reports in the literature, patients with a combination of toxic megacolon, perforation and fecal peritonitis had the poorest outcomes [2,8,14]. On analyzing the cause of death (pertaining to the entire study group) we found that 8 patients died of septicaemia, the remaining 4 had severe metabolic disturbances with encephalopathy, coagulopathy and respiratory failure from which they could not be salvaged. A combination of severe metabolic derangement along with pre existing sepsis (fecal peritonitis) or coagulopathy was lethal. In only one of the patients who died due to the metabolic sequelae (diabetic ketoacidosis) of pre-existing insulin-dependent diabetes mellitus, death could be attributed to comorbidity. As mentioned earlier, the overall postoperative mortality was higher (10/43; 23.3%) in our early experience (i.e. 1985–1995) when compared to our subsequent experience (2/29; 6.9%). We can only hypothesize that this improvement in operative outcome was in part due to the institution of an aggressive medical management protocol (following 1996) with early parenteral nutrition and emphasis on correction of metabolic derangements preoperatively. Another factor that could have improved the results was the conscious decision to use a decompressing ileostomy alone, as the first stage procedure, in very sick and moribund patients (Post-1996:10/29; 34.5% vs Pre-1996: 2/43; 4.7%; p = 0.003). Although the data is limited, we believe that this approach allows us to buy time and improve the patients overall condition so that the patient is able to withstand colectomy at a later date. Our current mortality figures are comparable to most other western series [8,9,15]. Following sub total colectomy (STC), the commonest morbidity was related to wound sepsis and postoperative fever. Patients with peritonitis at operation, those who developed postoperative fever or had ileostomy related complications were more likely to die postoperatively. In our experience there were 4 (6.6%) instances with stump 'blow outs' which caused pelvic sepsis and abdominal wound dehiscence and led to a prolonged hospital stay. Bleeding from the Hartmann's pouch did occur in a few patients (3.3%), but was managed with a combination of mesalamine and steroid enemas, local adrenaline saline lavages or by tapering the steroids slowly. A Hartmann's procedure or a mucus fistula of the rectal stump has been recommended by other workers also [16]. In the emergency setting, it is believed that a subtotal colectomy is the best option as it is technically simple, avoids dissection in the pelvis, minimizes blood loss and allows a subsequent restorative pouch procedure [8,9,15,17]. It also allows a definitive histological diagnosis to be obtained. In our study, all patients who had STC and survived (n = 60), were weaned off steroids after a month. Fifty-eight patients (2 late deaths) recovered fully and gained weight. Out of the 56 patients finally considered eligible for a second stage IPAA procedure, 47 underwent a second stage procedure (IPAA: 46; IRA: 1) with 1 postoperative death following an IPAA (Figure 1 and Table 3). The details of the patients' management have been listed in the flow chart (Figure 1). The second stage could be successfully performed in 81% (47/58) of the cases who survived the first procedure. Notably, by the time the patients undergo the 2nd stage procedure, the postoperative results obtained are similar to what can be achieved by an elective restorative proctocolectomy done in a patient with long-standing disease. Our second stage operative mortality was low at 2.1% (1/47) and comparable to the mortality rate in the elective setting (1/69; 1.5%; unpublished data). From the initial cohort of 72 patients, 48 patients are alive on follow up. Hence, two-third of the patients have been successfully salvaged and rehabilitated by staged surgery for severe ulcerative colitis. It is worth emphasizing here that the long term results of salvage and rehabilitation reflect the survival rate following the initial emergency surgery, as is apparent from our experience. This attrition following the first stage surgery is often unavoidable in 10–15% of cases because of late deaths, incidental detection of malignancy/Crohn's or indeterminate colitis [17], refusal for further surgery, associated intestinal/miliary tuberculosis and other co-morbid illnesses. Based on this experience we feel that in order to achieve a good postoperative outcome a combined team of physicians and surgeons should be involved in the management of patients with severe ulcerative colitis right from the day of admission. Aggressive nutritional therapy and correction of metabolic derangements combined with close clinical monitoring should be instituted along with the intensive steroid regimen [18-20]. Need for surgery should be assessed in advance [7,21] and early surgery is recommended in the event of failed medical treatment. In a majority of such cases a sub total colectomy is the procedure of choice and gives the best results. A loop ileostomy alone can be used in severely ill and moribund patients to buy time for a subsequent colectomy. These measures, when strictly implemented, would help to achieve a low and acceptable mortality rate. In India, emergency surgery for salvaging ulcerative colitis will continue to play an important role as few medical centers have the experience and resources to treat such patients with optimal skill and judgment, and nearly a third of the patients are refractory to intensive medical therapy. In consonance with the experience of other western centers [8,9,13], our retrospective analysis has reinforced that improved results can be achieved with experience and by following a policy of early surgical intervention within 5 days, especially in patients who have failed intensive medical therapy. Since immediate operative outcome is linked to the timing of the surgery, emphasis on anticipating failures of intensive medical therapy early is crucial for optimal decision-making and successful rehabilitation of these patients. Conflicts of interest The author(s) declare that they have no competing interests. Authors' contributions Sujoy Pal: Planning, data collection, study design and analysis, surgical management of patients, drafting and revising the manuscript Peush Sahni, Girish K. Pande: Planning, study design, surgical management of patients included, drafting and revising the manuscript SK Acharya: Planning, study design, medical management of patients, drafting and revising the manuscript TK Chattopadhyay: Planning, study design, surgical management of patients, drafting and revising the manuscript Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements The authors acknowledge the help provided by Dr Rajbir Singh, Scientist, Deptt. Of Biostatistics, AIIMS, for the statistical analysis of data. Figures and Tables Figure 1 Flow chart showing the procedures done and outcome (short and long term). IPAA: ileal J Pouch anal anastomosis; IRA: ileorectal anastomosis; TPC: total proctocolectomy with permanent Brooke's ileostomy; FU: follow up; LFU: lost to follow up. * One patient had rectal cancer receiving adjuvant therapy and IPAA was not contemplated; the other patient developed intestinal tuberculosis and recurrent small bowel obstruction with multiple reoperations, hence pouch was not advocated. # One patient died of massive cerebral thromboembolism secondary to valvular heart disease with infective endocarditis; one additional patient died of disseminated colorectal cancer. Table 1 Operative mortality with regard to indication and timing* Indication n Died Mortality (%) >5 d Refractory severe UC 51 8 15 8 Severe LGIH 5 2 40 0 Toxic megacolon 4 2 50 1 * No mortality in perforation and obstruction group UC: ulcerative colitis; LGIH: Lower gastrointestinal haemorrhage Table 2 Clinical, biochemical and demographic data of the patients who failed intensive medical therapy (n = 51) according to the 2 subgroups (Group I: those operated ≤5 days and Group II >5 days after initiation of intensive steroid therapy). Variable Group I* (n = 17) Mean (± SEM) Group II* (n = 34) Mean (± SEM) Age (years) 33 (3.8) 35 (2) Disease duration (months) 48 (15.2) 32 (6.7) Stool frequency/day 10.2 (0.9) 11.9 (0.9) Pulse rate (/minute) 92 (5.3) 98 (2.5) Poor nutritional status# 9 (56%) 22 (64%) Hemoglobin (g/dl) 9.9 (0.6) 9.4 (0.4) ESR (mm) 41 (7.1) 44.8 (4.3) Urea (mg/dl) 22.2 (1.9) 26.3 (4.8) Serum albumin (g/dl) 2.8 (0.14) 2.5 (0.13) * On statistical comparison none of these variables were found to be significantly different # Number of patients classified as having poor nutritional status SEM - Standard error of means Table 3 The ultimate fate of each patient undergoing emergency sub total colectomy (STC) Total patients surviving STC 58 Lost to follow up 3 Refused pouch 2 Pouch not offered 2 Patients eligible for 2nd stage 51 Total IPAA done 46 IRA done 1 TPC 1 Awaiting 3 ==== Refs Edwards FC Truelove SC The course and prognosis of ulcerative colitis Gut 1963 4 299 315 Jewell DP Feldman M, Friedman LS, Sleisenger M Ulcerative colitis Chapter in: Sleisenger and Fordtran's Gastrointestinal and Liver disease 2002 7 Saunders-Elsevier Science, Philadelphia 2039 67 Truelove SC Witts LJ Cortisone in ulcerative colitis – final report on a therapeutic trial Br Med J 1955 2 1041 1048 Truelove SC Jewell DP Intensive intravenous regimen for severe attacks of ulcerative colitis Lancet 1974 2 1067 70 10.1016/S0140-6736(74)90552-2 Jarnerot G Rolny P Sandberg-Gertzen H Intensive intravenous treatment of ulcerative colitis Gastroenterology 1985 85 1005 13 Kornbluth A Marion JF Salomon P Janowitz HD How effective is current medical therapy for severe ulcerative and Crohn's colitis? An analytic review of selected trials J Clin Gastroenterol 1995 20 280 4 7665814 Travis SP Farrant JM Ricketts C Nolan DJ Mortensen NM Kettlewell MG Jewell DP Predicting outcome in severe ulcerative colitis Gut 1996 38 905 10 8984031 Goligher JC Hoffman DC de Dombal FT Surgical treatment of severe attacks of ulcerative colitis, with special reference to the advantages of early operation Br Med J 1970 4 703 6 5491253 Hawley PR Emergency surgery for ulcerative colitis World J Surg 1989 12 169 173 10.1007/BF01658049 Grotz RL Pemberton JH The ileal pouch operation for ulcerative colitis Surg Clin North Am 1993 73 909 32 8378832 Farouk R Pemberton JH Surgical options in ulcerative colitis Surg Clin North Am 1997 77 85 94 9092119 10.1016/S0039-6109(05)70534-X Brooke BN Sampson PA An indication for surgery in acute ulcerative colitis Lancet 1964 17 1272 3 14219132 10.1016/S0140-6736(64)92740-0 Veidenheimer MC When should we operate on the patient with ulcerative colitis? Surg Clin North Am 1970 50 743 8 5446561 Grant CS Dozois RR Toxic megacolon: ultimate fate of patients after successful medical management Am J Surg 1984 147 106 10 6691535 10.1016/0002-9610(84)90042-4 Becker JM Surgical therapy for ulcerative colitis and Crohn's disease Gastroenterol Clin North Am 1999 28 371 90 10372273 10.1016/S0889-8553(05)70061-3 Motson RW Manche AR Modified Hartmann procedure for acute ulcerative colitis Surg Gynecol Obstet 1985 160 462 3 3992450 Katz JR Medical and surgical management of severe colitis Sem Gastrointest Dis 2000 11 18 32 Rombeau JL Barot LR Williamson CE Mullen JL Preoperative total parenteral nutrition and surgical outcome in patients with inflammatory bowel disease Am J Surg 1982 143 139 43 6797311 10.1016/0002-9610(82)90144-1 Truelove SC Willoughby CP Lee EG Kettlewell MG Further experience in the treatment of severe attacks of ulcerative colitis Lancet 1978 2 1086 8 82099 10.1016/S0140-6736(78)91816-0 Hyde GM Jewell DP Review article: the management of severe ulcerative colitis Aliment Pharmacol Ther 1997 11 419 24 9218065 10.1046/j.1365-2036.1997.00187.x Lennard-Jones JE Ritchie JK Hilder W Spicer CC Assessment of severity in colitis: a preliminary study Gut 1975 16 579 84 1183857	16316474	PMC1325033	CC BY	2021-01-04 16:03:28	no	BMC Gastroenterol. 2005 Nov 30; 5:39	utf-8	BMC Gastroenterol	2,005	10.1186/1471-230X-5-39	oa_comm
==== Front BMC CancerBMC Cancer1471-2407BioMed Central London 1471-2407-5-1521632114610.1186/1471-2407-5-152Research ArticleKinetic analysis of dynamic 18F-fluoromisonidazole PET correlates with radiation treatment outcome in head-and-neck cancer Thorwarth Daniela [email protected] Susanne-Martina [email protected] Jutta [email protected] Frank [email protected] Markus [email protected] Section for Biomedical Physics, Clinic for Radiation Oncology, University Hospital Tübingen, Germany2 Department of Nuclear Medicine, Radiological Clinic, University Hospital Tübingen, Germany3 Department of Radiation Therapy, Clinic for Radiation Oncology, University Hospital Tübingen, Germany2005 1 12 2005 5 152 152 11 7 2005 1 12 2005 Copyright © 2005 Thorwarth et al; licensee BioMed Central Ltd.2005Thorwarth 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 Hypoxia compromises local control in patients with head-and-neck cancer (HNC). In order to determine the value of [18F]-fluoromisonidazole (Fmiso) with regard to tumor hypoxia, a patient study with dynamic Fmiso PET was performed. For a better understanding of tracer uptake and distribution, a kinetic model was developed to analyze dynamic Fmiso PET data. Methods For 15 HNC patients, dynamic Fmiso PET examinations were performed prior to radiotherapy (RT) treatment. The data was analyzed using a two compartment model, which allows the determination of characteristic hypoxia and perfusion values. For different parameters, such as patient age, tumor size and standardized uptake value, the correlation to treatment outcome was tested using the Wilcoxon-Mann-Whitney U-test. Statistical tests were also performed for hypoxia and perfusion parameters determined by the kinetic model and for two different metrics based on these parameters. Results The kinetic Fmiso analysis extracts local hypoxia and perfusion characteristics of a tumor tissue. These parameters are independent quantities. In this study, different types of characteristic hypoxia-perfusion patterns in tumors could be identified. The clinical verification of the results, obtained on the basis of the kinetic analysis, showed a high correlation of hypoxia-perfusion patterns and RT treatment outcome (p = 0.001) for this initial patient group. Conclusion The presented study established, that Fmiso PET scans may benefit from dynamic acquisition and analysis by a kinetic model. The pattern of distribution of perfusion and hypoxia in the tissue is correlated to local control in HNC. ==== Body Background Local control remains a great challenge in head-and-neck cancer (HNC) treatment. Even with an optimal combination of radio- and chemotherapy, local recurrences are observed in up to 50% of the treated patients [1,2]. Up to now, no reliable parameter could be established that would account for this high rate of local failures. Tumor hypoxia has been known to be associated with poor radiation response for several decades. Recent publications suggested that hypoxia in tumors had a direct influence on treatment success [3,4] by a variety of mechanisms [5,6]. A prognostic impact of tumor hypoxia for therapy outcome in head and neck cancer (HNC) has been shown by different investigators [7-9]. Hypoxia has also been related to lower survival probability and higher risk of recurrence in patients with cervix cancer [4,10]. In these studies, hypoxia was assessed invasively by polarographic Eppendorf electrodes. Positron emission tomography (PET) with appropriate radiotracers enables non-invasive assessment of the presence and distribution of hypoxia. The radiotracers in frequent use are 18F-fluoromisonidazole (Fmiso) [11-13] and chemically similar markers such as 18F-fluoroazomycin (Faza) [14] or, with a different binding mechanism, 60Cu-ATSM [15]. Some investigations report an unclear correlation between Eppendorf measurements and standardized uptake values (SUV) determined on the basis of Fmiso PET [16]; even though a tumor-to-blood ratio of 1.4 was defined as diagnostic of hypoxia [11]. Thus, the predictive value of Fmiso SUV even several hours after tracer injection remains unclear. Based on their chemical structure, nitroimidazoles are trapped inside hypoxic cells. This feature makes these agents ideal markers for hypoxia in in-vitro cell systems [17]. However, transforming this into larger scale biological systems is problematic and the interpretation of Fmiso PET images remains unclear. An advantage of PET compared to Eppendorf measurements is the ability to display spatial distributions, which is necessary for the integration of hypoxia information into adaptive treatments such as hypoxia dose painting [18-20]. For immunohistochemical investigations, the marker pimonidazole is well established [21-23] to stain hypoxic tumor cells. As the functional binding mechanisms of pimonidazole and Fmiso are similar, Fmiso should be specific to hypoxia to a similar degree. However, the immunohistochemical staining patterns are very complex and reveal a highly heterogeneous distribution of perfused blood vessels and hypoxic patches, sometimes interspersed with necrotic islands, all occurring on a microscopic scale. This may hint as to why Fmiso tracer uptake alone is not a reliable diagnostic quantity, and indicates the requirement of an analysis of dynamic Fmiso PET which takes into account the structural complexity of hypoxic tumor tissues. The study described here was designed to develop a kinetic model in order to understand the spatial and temporal distribution of Fmiso in the tumor tissue. Since the predictive character of Fmiso SUV remains unclear in literature [13,16], the time course of tracer accumulation in the tumor was investigated. This analysis delivers patient specific values for perfusion, kinetic constants and the concentration of tracer retaining cells. Furthermore, the relation between these parameters and radiation therapy (RT) treatment outcome for HNC was investigated in a group of 15 HNC patients who were examined with dynamic Fmiso PET prior to treatment with primary radiotherapy. Methods Patients After informed consent, sixteen patients (mean age: 57.2 years old, range: 46 – 69; 14 male, 2 female) with advanced stage head and neck cancer (HNC) were examined between November 2001 and March 2004. The Fmiso examinations were performed prior to radiation therapy (RT) treatment. All patients were treated with primary RT to 70 Gy. Three of these patients were treated with Intensity Modulated Radiotherapy (IMRT) in 35 fractions, 5 fractions a week with a daily dose of 2 Gy. The other 13 patients received conventional RT, 5 fractions with 2 Gy per week until 30 Gy. This first phase was followed by a hyperfractionation composed of a dose of 1.4 Gy applicated twice per day until the end of treatment. In addition, concomitant chemotherapy was prescribed for 14 patients. Seven patients received 5-Fluorouracil/Mitomycin chemotherapy, whereas for six patients Cisplatin/Mitomycin was prescribed; one patient had Paclitaxel/Cisplatin chemotherapy. Whenever possible (n = 12), an additional [18F]-fluorodeoxyglucose (FDG) PET was taken a few days (1 – 3) before or after the Fmiso PET scan. For each patient, additional computed tomography (CT) image data was available. These CT scans, on which delineation of target volumes and organs at risk was performed, were used for RT treatment planning. After the end of therapy, patients were reviewed regularly every three months with clinical examination, flexible endoscopy and computed tomography (CT) when recurrent disease was suspected. Routine CT scans were also acquired six weeks and one year after therapy was finished. Failure was defined as CT proven tumor progression. Data acquisition The Fmiso PET examinations were performed on a whole-body scanner (Advance, GE Medical Systems, Milwaukee, US) after automatic bolus injection of 400 MBq Fmiso. PET data acquisition was started at the time of tracer injection. During the first 15 (9 patients) to 60 min (7 patients), a dynamic image acquisition of 31 (40) frames was performed. Additional static emission scans were taken 2 h and 4 hour post injection (p.i.). Concerning the FDG PET acquisition, a static emission scan was taken 1 h after injection of approximately 400 MBq FDG. For the delineation of the tumor volume relevant in the context of this study, the FDG PET image data was used. The tumor volume was defined as the volume including all voxels with at least 40% of the maximum intensity. This delineation technique was combined with a 12 mm margin (3 PET voxels). The tumor volume variable V used in the current study refers to the described FDG PET volume. It is determined as V = n·v, where n is the number of tumor voxels. v represents the volume of a single voxel, in our case v = (0.42·0.425) cm3 = 0.068 cm3. In order to match the FDG-defined tumor volume onto the three different Fmiso data sets (dynamic, 2 and 4 h p.i.), an automatic coregistration [24] was performed, which achieved a matching accuracy of ≤ 2 mm. The resulting transformation matrices were used to determine a time-activity curve (TAC) for each tumor voxel. Compartment model The voxel-by-voxel TACs were analyzed using a pharmaco-kinetic model which is described in detail elsewhere [25]. Briefly, the kinetic model consists of two compartments, one corresponding to the irreversible binding of the tracer in hypoxic cells, the other representing freely diffusive Fmiso. This two-compartment system is combined with an input function which is individually determined by a reference tissue approach for lack of a blood signal in the field of view of the scanner (see [25] for details). The voxel-by-voxel analysis of the Fmiso TACs was done by fitting the five-parameter analytical model function for the tracer concentration in the tissue compartments to the measured Fmiso curves. This approach yields for each tumor voxel one characteristic value for tracer retention and perfusion. The perfusion value, i.e. the density of perfused blood vessels in the respective tumor voxel, is mainly guided by the shape of the TAC during the first few minutes after injection. In contrast, the amount of tracer retention potential (TRP) in the voxel is related to the properties of the curve several hours after tracer injection. TRP takes into account the number of viable hypoxic cells as well as their grade of hypoxia. In other words, TRP is a measure of the concentration of specifically bound tracer in the considered area. Fluctuations in perfusion states are taken into account by construction of the model [25]. In order to visualize TRP and perfusion characteristics of the whole tumor simultaneously, a scatter plot is introduced (see appendix and fig. 1). In this plot, the TRP in a voxel is plotted along the x-axis, while the contribution to the signal from perfused blood vessels is plotted along the y-axis. The variety of scatter patterns in the patient group leads to the hypothesis that TRP and perfusion are independent and spatially variable parameters of a tumor tissue (see figure 2). Figure 1 Left: Scatter plot for one patient based on tracer retention and perfusion parameters resulting from a kinetic Fmiso analysis. Schematically shown are typical regions for characteristic perfusion-hypoxia patterns: (1) High perfusion without hypoxia, (2) well perfused and simultaneously hypoxic, and (3) severe hypoxia, low vessel density. Right: Corresponding types of characteristic Fmiso Time-Activity Curves. Figure 2 Scatter plots of all 15 patients with increasing M-value. Data analysis and statistics Tumor control was defined on the basis of computed tomography (CT) scans as complete and persistent regression of the primary tumor and failure was defined as local recurrence of the tumor in the irradiated fields. Follow up time was determined from the end of RT treatment until the day of the last CT. Different variables that might influence treatment outcome were compared using the Wilcoxon-Mann-Whitney (Wilcoxon signed rank) U-test between patient groups showing no local relapse and failure. In all cases, a two-sided significance level of 0.05 was used. Correlation of different variables with was assessed using a Pearson correlation coefficient. The impact on treatment outcome was checked for different classes of variables: tumor volume and patient age, SUV related factors and variables derived from the kinetic analysis. The SUV related factors were the maximum standardized uptake value (SUVmax) and the fractional hypoxic volume (FHV) 4 h after Fmiso injection. FHV is defined as the fraction of tumor volume presenting a tumor-to-blood ratio larger than 1.4. Both variables SUVmax and FHV have been correlated with tumor hypoxia in earlier studies [11,13]. Finally, a number of parameters derived from the compartmental analysis were checked for a statistically significant influence on therapy outcome. These parameters were the mean value of TRP, the mean value of perfusion, and two metrics involving both TRP and perfusion parameter values. A first metric was defined intuitively as the volume integral of the TRP-to-perfusion ratio (HPR). A second metric, which was derived from a model of tumor dose-response and reoxygenation, is the malignancy value M as described in more detail in the appendix. Results Kinetic Fmiso data The voxel-by-voxel Fmiso TACs showed a variety of different tracer uptake patterns. Perfusion and hypoxia status of the tissue area can be differentiated by means of the Fmiso TAC shape. The former is determined by the part of the TAC corresponding to time points only a few minutes p.i, whereas the latter is linked to the shape of the curve several hours after tracer injection. The TACs observed for the group of 16 patients showed mainly three different types of perfusion-hypoxia patterns which correspond to (1) tissue areas with a high vessel density, (2) well perfused but also hypoxic, and (3) severely hypoxic tumor areas (see fig. 1). Model The presented compartment model allows us to derive patient specific perfusion-hypoxia patterns. The model is able to describe the different observed types of Fmiso time curves. Characteristic TACs are associated to distinct areas in the scatter plot (figure 1), which indicates high stability of the model. The patterns for the whole group of patients are displayed in scatter plots in figure 2 (appendix). The ultimate purpose of the kinetic model is to subtract the background of unbound tracer from the signal intensity. Patients Characteristics of the group of 15 patients are summarized in table 1. For the examined patient group, the follow up time was in the range of 2 – 21 months (median: 12.8 months). Patients were 46 to 68 years old (median: 59 years). FDG-tumor volumes ranged from 32.4 to 287.6 cm3 with a median volume of 114.5 cm3. Overall, 7 of the 15 patients had local recurrences. All observed failures occurred in the first 8 months after the end of therapy. Table 1 patient characteristics. Tumor characteristics of the examined patients. Patients Pat. nr. primary tumor site age sex TNM stage tumor volume V [cm3] failure site 1 oropharynx 60 m T4 N2b MO 258.1 T/N† 2 oropharynx 51 m T4 N2c MO 126.2 T 3 larynx 66 m T4 N2c MO 153.7 T/N† 4 FOM‡ 46 m T3 N2b MO 59.0 T 5 BOT§ 49 m T4 N2c MO 287.6 T/N† 6 oropharynx 48 m T2 N2c MO 114.7 T/N† 7 FOM‡ 68 m T4 Nl M0 213.7 T/N† 8 oropharynx 65 m T3 N2c MO 74.3 - 9 hypopharynx/BOT§ 51 m T4 N2c MO 100.9 - 10 oropharynx 59 m T2 N3 MO 172.4 - 11 oropharynx 50 f T4 N2c MO 44.0 - 12 larynx-/hypopharynx 60 m T4 NO MO 32.4 - 13 oro-/hypopharynx 49 m T3 N2c MO 80.9 - 14 oropharynx 60 f T4 N2b MO 52.4 - 15 unknown 68 m T4 Nl M0 125.5 - T: tumor; †N: node. ‡FOM: floor of mouth; §BOT: base of tongue. Image analysis of the Fmiso PET scans taken 4 h p.i. revealed maximum SUVs in the tumor volume between 1.36 and 4.02. The median SUVmax was 2.25. The FHV ranged from 0 to 72.5% with a mean of 19.7%. Due to the chosen tumor volume definition strategy, which implies the addition of a margin, the determined FHV can never reach 100%. Examination of the scatter plots showed very different patterns of hypoxia and perfusion. All possible combinations of hypoxia and perfusion parameters were observed: well perfused tumors which were not at all hypoxic, tumors showing at the same time a quite high vascular density and hypoxic subareas, and finally also tumors that were badly perfused and severely hypoxic. These two variables represent physiological tumor characteristics that are not correlated (r = -0.096). As a first result, it has to be stated that hypoxia occurs independently from the degree of perfusion in tumor tissues. The Wilcoxon-Mann-Whitney U-test with respect to the age of the patients showed no difference (P = 0.3) between the subgroups with and without relapse. In contrast, there was a significant difference in tumor volume between the two subgroups (P = 0.014). This corroborates the findings of earlier studies that correlated tumor size with treatment outcome [26]. Also, SUVmax, determined 4 h after injection separated patients according to failure and progression free survival (PFS). The significance for SUVmax was only weak P = 0.041, whereas the U-test for the FHV showed no significance at all (P = 0.13). Regarding the variables derived from the kinetic analysis, mean tumor perfusion and HPR discriminated between the group without recurrence and the failure group (P = 0.05 and 0.008, respectively). The mean TRP value showed no significance (P = 0.18). Finally, the malignancy value M was highly significant, with P = 0.0013 (table 2). The prognostic value of this model based metric M is higher than the value of tumor size or SUVmax after 4 h. Table 2 results of statistical analysis. Results of univariate analysis of prognostic factors. Results U-Test variables P-Value age [years] NS* (0.30) tumor volume V [cm3] 0.014 SUVmax† MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciGacaGaaeqabaqabeGadaaakeaacqqGtbWucqqGvbqvcqqGwbGvdaqhaaWcbaWexLMBbXgBcf2CPn2qVrwzqf2zLnharyGvLjhzH5wyaGabciaa=1gacaWFHbGaa8hEaaqaaiabcccigcaaaaa@3F10@ 0.041 FHV‡ [%] NS* (0.13) mean TRP* NS* (0.18) mean perfusion§ 0.05 HPR§ 0.008 malignancy value M§ 0.001 *NS: not significant (p > 0.05); †SUVmax: maximum SUV; ‡FHV: fractional hypoxic volume. §derived from kinetic model Discussion Recent publications revealed contradictory results concerning the correlation of static Fmiso PET data and tumor hypoxia [11-13,16]. As the irregular architecture of tumors complicates Fmiso uptake, a kinetic model was developed in order to analyze spatial and temporal distribution of the tracer in head-and-neck tumors. The presented model enables to differentiate between tumor perfusion and hypoxia. This feature of the model constitutes the link between Fmiso distribution and retention and the structural architecture of the tumor tissue. The results of this study showed, that SUVmax alone even at late time points has limited predictive value. These findings are in line with results of other investigators [16] who found that SUV 2 h p.i. and Eppendorf did not correlate well. A limiting factor for the retention of Fmiso in the tumor is that binding of the tracer can only take place in viable hypoxic cells which may be few if the tumor is largely necrotic. In other words, a low level of the Fmiso TAC several hours after tracer injection is not necessarily due to non-hypoxic tissue. This might also be caused by largely necrotic tumor areas which contain only a very low number of strongly hypoxic cells. In this case, the low intensity of the PET signal would lead to an underestimation of the extent of hypoxia by the SUV-method. A kinetic analysis subtracts the non-specific background signal and hence enables to determine the local TRP of the tumor. Still, the classical hypoxic tumor core may only give a weak signal due to the low density of tracer retaining cells. Hence, a second parameter is needed to give a more complete picture of the abnormalities of the tissue architecture. The analysis of the parameters derived from the kinetic model demonstrated, that TRP and perfusion values alone do not predict treatment outcome. Additionally, hypoxia occurred independent of degree of perfusion, since no correlation was found between the two variables. Recent immunohistochemical investigations of simultaneous pimonidazole and blood vessel staining of tissue sections [21-23] revealed the co-existence of hypoxic areas and perfused blood vessels. These results were corroborated in our study. Taking both parameters together proved to be reliable predictors for treatment outcome. The malignancy metric M, which involves these two physiological characteristics of the tissue, was found to be the strongest prognostic factor. Most essential for the design of new adaptive treatment strategies is the time until reoxygenation takes place after the beginning of RT. The malignancy metric M involves an estimate of this characteristic time. The worst physiological setting in a tumor seems to be the combination of low perfusion and severe hypoxia, as reoxygenation then appears to be very slow. In contrast, a high degree of perfusion co-existing with hypoxic areas may favor fast reoxygenation. Hence, this setting might be associated with an intermediate level of risk. This interpretation can be supported by follow-up scans during RT, which will be reported in a future publication. Fmiso uptake kinetics are quite slow due to long diffusion distances and for lack of active transport mechanisms. PET scans several hours after injection of the radiotracer are therefore essential. Nevertheless, dynamic scans at short times p.i. cannot be abandoned, as they are needed to determine the degree of perfusion of the tumor. There is no possibility in Fmiso PET to distinguish between acute and chronic hypoxia [27]. On one hand, this is due to a quite large size of the image voxels (≈ (4 mm)3). On the other hand, the slow kinetics of tracer retention do not allow a distinction of fast re-perfusion. Since both effects are a consequence of the deficient vasculature, they may co-exist anyway. The results of this study demonstrate that Fmiso PET has prognostic value for therapy outcome, but only when perfusion and retention are both taken into consideration. A higher predictive value was associated to the malignancy value M derived form kinetic analysis than to tumor volume or SUV based variables. Hence, Fmiso PET might in the future be used to individually select patients for an adapted radiotherapy treatment as e.g. dose painting [18-20]. Furthermore, variables derived from a kinetic analysis [25] may serve to determine individual dose escalation factors in order to overcome hypoxia related treatment resistance. Conclusion The interpretation of Fmiso PET examinations with respect to hypoxia benefits greatly from a kinetic analysis. The presented kinetic analysis determines hypoxia and perfusion parameters, which were shown to be able to stratify patient groups according to RT treatment outcome. The results of this explorating, hypothesis generating study require validation in a larger group of patients. Appendix Scatter plots By virtue of the kinetic model analysis of the time-activity curves of tracer uptake, it is possible to eliminate the non-specific background activity in the signal. The model has five fit parameters, which are determined for each voxel of the tumor volume. Two parameters are of special interest: the relative contribution of the perfused blood vessels, short WP, which dominates the signal during the first few minutes after injection. Further, the tracer retention potential R, which is a combination of the concentration of retaining cells and the kinetic constants of the reaction. In a scatter plot, the values of R are plotted along the x-axis, and the values of WP are plotted along the y-axis for all voxels of the tumor volume. The scatter plots of all 15 patients in figure 2 show clearly that both values are independent variables and vary widely in a population. TCP model The resistance of a hypoxic tumor to RT is governed, among others, by two factors: the initial magnitude of the hypoxic subpopulation of clonogenic cells, and the rate with which these cells are reoxygenated. We assume that the former is related to R, while the latter is related to WP. The rationale for the second assumption is, that in areas where the blood vessel density is high, hypoxic cells have a greater chance to become oxic quickly. Conversely, if the vasculature is severely deficient, reoxygenation is delayed. The common Poisson model of tumor control probability (TCP) states: −ln⁡TCP (D) = ∑in exp⁡(−α0D) (1) MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciGacaGaaeqabaqabeGadaaakeaacqGHsislcyGGSbaBcqGGUbGBcqqGubavcqqGdbWqcqqGqbaucqqGGaaicqqGOaakcqWGebarcqqGPaqkcqqGGaaicqqG9aqpcqqGGaaidaaeqbqaaiabd6gaUjaaykW7cyGGLbqzcqGG4baEcqGGWbaCaSqaaiabdMgaPbqab0GaeyyeIuoakiabcIcaOiabgkHiTiabeg7aHnaaBaaaleaacqaIWaamaeqaaOGaemiraqKaeiykaKIaaCzcaiaaxMaadaqadaqaaiabigdaXaGaayjkaiaawMcaaaaa@4EB7@ where the sum runs over all voxels i of the tumor and n is the number of cells per voxel. D is the total dose and α0 the radiation sensitivity of a non-hypoxic cell. We modify this to include hypoxic subpopulations to read −ln⁡ TCP(D) = ∑i∫dαh nhi(αh) exp⁡(−(α0−αh)D) (2) MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=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@6505@ = n exp⁡(−α0D) ∑i∫dαh hi(αh) exp⁡(αhD). (3) MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciGacaGaaeqabaqabeGadaaakeaacqGH9aqpcqqGGaaicqWGUbGBcaaMc8UagiyzauMaeiiEaGNaeiiCaaNaeiikaGIaeyOeI0IaeqySde2aaSbaaSqaaiabicdaWaqabaGccqWGebarcqGGPaqkcaaMc8+aaabuaeaadaWdbaqaaiabdsgaKjabeg7aHnaaBaaaleaacqWGObaAaeqaaOGaaGPaVlabdIgaOnaaBaaaleaacqWGPbqAaeqaaaqabeqaniabgUIiYdGccqGGOaakcqaHXoqydaWgaaWcbaGaemiAaGgabeaakiabcMcaPiaaykW7cyGGLbqzcqGG4baEcqGGWbaCaSqaaiabdMgaPbqab0GaeyyeIuoakiabcIcaOiabeg7aHnaaBaaaleaacqWGObaAaeqaaOGaemiraqKaeiykaKIaeiOla4IaaCzcaiaaxMaadaqadaqaaiabiodaZaGaayjkaiaawMcaaaaa@613E@ Here, h(αh) is the frequency with which an average reduction of the cell sensitivity by αh occurs over the course of the treatment. The integral is the malignancy M. We define a phenomenological relation between the kinetic model parameters and the malignancy by: M = 1 + exp(bR/(WP + c)), (4) where b and c are fit parameters. Finally, we obtain −ln⁡ TCP(D)=a∑i[1+exp⁡(bRi/(WP,i+c))] (5) MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciGacaGaaeqabaqabeGadaaakeaacqGHsislcyGGSbaBcqGGUbGBcaaMc8UaeeivaqLaee4qamKaeeiuaaLaeeikaGIaemiraqKaeiykaKIaeyypa0Jaemyyae2aaabuaeaacqGGBbWwcqaIXaqmcqGHRaWkcyGGLbqzcqGG4baEcqGGWbaCaSqaaiabdMgaPbqab0GaeyyeIuoakiabcIcaOiabdkgaIjabdkfasnaaBaaaleaacqWGPbqAaeqaaOGaei4la8IaeiikaGIaem4vaC1aaSbaaSqaaiabdcfaqjabcYcaSiabdMgaPbqabaGccqGHRaWkcqWGJbWycqGGPaqkcqGGPaqkcqGGDbqxcaWLjaGaaCzcamaabmaabaGaeGynaudacaGLOaGaayzkaaaaaa@598C@ with a = n exp(-α0D) as an additional fit parameter. This sum can be computed as a sum over the points of a scatter plot. The parameters a, b and c were determined by a maximum log-likelihood fit of TCP to the group of 15 patients in this study. Their values obtained as a = 5.7·10-5, b = 198.6 and c = 0.565. The goodness of fit was estimated by evaluation of the deviance Δ. The deviance is defined as twice the difference between the current and the full log-likelihood Δ = -2(Lc - Lf), which is supposed to follow a χ2 distribution. In our case, the deviance confirmed an acceptable fit (Δ = 2.75, p > 0.05). Figure 3 shows the the fitted TCP curve as a function of the malignancy value M together with the grouped data points obtained from outcome data for the group of 15 patients in this study. Figure 3 TCP curve fitted to the outcome data of 15 patients irradiated with a total dose of 70 Gy. Competing interests The author(s) declare that they have no competing interests. Authors' contributions DT developed the kinetic model, the parameter analysis strategy, and the TCP model, performed all data analysis, and drafted the manuscript. SE carried out all PET examinations and acquired the data. FP was involved in the design of the study and its coordination. JS participated in the design of the study and performed the statistical analysis. MA conceived the study, developed the kinetic model and the TCP model, has made substantial contributions for data interpretation and was participated in drafting 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 This project has been financially supported by the German Research Foundation (DFG), grant no. AL 877/1. The authors would like to thank Prof. H-J Machulla and his team (Radiopharmacy Section, University Hospital Tübingen) for excellent [18F]-Fmiso production. We also thank Dr. C Belka for careful manuscript revision. ==== Refs Brizel DM Albers ME Fisher SR Scher RL Richtsmeier WJ Hars V George SL Huang AT Prosnitz LR Hyperfractionated irradiation with or without concurrent chemotherapy for locally advanced head and neck cancer N Engl J Med 1998 338 1798 1804 9632446 10.1056/NEJM199806183382503 Budach V Stuschke M Budach W Baumann M Geismar D Grabenbauer G Lammert I Jahnke K Stueben G Herrmann T Bamberg M Wust P Hinkelbein W Wernecke K Hyperfractionated accelerated chemoradiation with concurrent fluorouracil-mitomycin is more effective than dose-escalated hyperfractionated accelerated radiation therapy alone in locally advanced head and neck cancer: final results of the radiotherapy cooperative clinical trials group of the German Cancer Society 95-06 Prospective Randomized Trial J Clin Oncol 2005 23 1125 1135 15718308 10.1200/JCO.2005.07.010 Coleman CN Mitchell JB Camphausen K Tumor Hypoxia: Chicken, Egg, or a Piece of the Farm? (editorial) J Clin Oncol 2002 20 610 615 11821437 Fyles A Milosevic M Hedley D Pintilie M Levin W Manchul L Hill RP Tumor Hypoxia has Independent Predictor Impact Only in Patients With Node-Negative Cervix Cancer J Clin Oncol 2002 20 680 687 11821448 10.1200/JCO.20.3.680 Weinmann M Belka C Plasswilm L Tumour hypoxia: impact on biology, prognosis and treatment of solid malignant tumours Onkologie 2004 27 83 90 15007254 10.1159/000075611 Weinmann M Jendrossek V Guner D Goecke B Belka C Cyclic exposure to hypoxia and reoxygenation selects for tumor cells with defects in mitochondiral apoptotic pathways FASEB 2004 18 1906 1908 Nordsmark M Overgaard M Overgaard J Pretreatment oxygenation predicts radiation response in advanced squamous cell carcinoma of the head and neck Radiother Oncol 1996 41 31 39 8961365 Nordsmark M Overgaard J A confirmatory prognostic study on oxygenation status and loco-regional control in advanced head and neck squamous cell carcinoma treated by radiation therapy Radiother Oncol 2000 57 39 43 11033187 10.1016/S0167-8140(00)00223-1 Brizel DM Sibley GS Prosnitz LR Scher RL Dewhirst MW Tumor hypoxia adversely affects the prognosis of carcinoma of the head and neck Int J Radiat Oncol Biol Phys 1997 38 285 289 9226314 10.1016/S0360-3016(97)00101-6 Höckel M Knoop C Schlenger K Vorndran B Knapstein PG Vaupel P Intratumoural pO2 predicts survival in advanced cancer of the uterine cervix Radiother Oncol 1993 26 35 50 Rasey JS Koh WJ Evans ML Peterson LM Lewellen TK Graham MM Krohn KA Quantifying regional hypoxia in human tumours with positron emission tomography of [18F] fluoromisonidazole: A pretherapy study of 37 patients Int J Radiat Oncol Biol Phys 1996 36 417 428 8892467 10.1016/S0360-3016(96)00325-2 Koh WJ Rasey JS Evans ML Grierson JR Lewellen TK Graham MM Krohn KA Griffin TW Imaging of hypoxia in human tumours with [F-18]fluoromisonidazole Int J Radiat Oncol Biol Phys 1992 22 199 212 1727119 Eschmann SM Paulsen F Reimold M Dittmann H Welz S Reischl G Machulla HJ Bares R Prognostic Impact of Hypoxia Imaging with 18F-Misonidazole PET in Non-Small Cell Lung Cancer and Head and Neck Cancer Before Radiotherapy J Nucl Med 2005 46 253 260 15695784 Piert M Machulla HJ Picchio M Reischl G Ziegler S Kumar P Wester HJ Beck R McEwan AJB Wiebe LI Schwaiger M Hypoxia-Speciflc Tumor Imaging with 18F-Fluoroazomycin Arabinoside J Nucl Med 2005 46 106 113 15632040 Dehdashti D Grigsby PW Aintun MA Lewis JS Siegel BA Welch M Assessing tumour hypoxia in cervical cancer by positron emission tomography with 60Cu-ATSM: Relationship to therapeutic response – a preliminary report Int J Radiat Oncol Biol Phys 2003 55 1233 1238 12654432 10.1016/S0360-3016(02)04477-2 Bentzen L Keiding S Nordsmark M Falborg L Hansen SB Keller J Nielsen OS Overgaard J Tumour oxygenation assessed by 18F-fluoromisonidazole PET and polarographic needle electrodes in human soft tissue tumours Radiother Oncol 2003 67 339 344 12865184 10.1016/S0167-8140(03)00081-1 Rasey JS Nelson NJ Chin L Evans ML Grunbaum Z Characteristics of the binding of labeled fluoromisonidazole in cells in vitro Radiat Res 1990 122 301 308 2356284 Ling HummiJCC Larson S Amols H Fuks Z Leibel S Koutcher JA Towards multidimensional radiotherapy (MD-CRT): biological imaging and biological conformality Int J Radiat Oncol Biol Phys 2000 47 551 560 10837935 10.1016/S0360-3016(00)00467-3 Chao KS Bosch WR Mutic S Lewis JS Dehdashti F Mintun MA Dempsey JF Perez CA Purdy JA Welch MJ A novel approach to overcome hypoxic tumour resistance: Cu-ATSM-guided intensity-modulated radiation therapy Int J Radiat Oncol Biol Phys 2001 49 1171 1182 11240261 10.1016/S0360-3016(00)01433-4 Alber FM Paulsen Eschmann SM Machulla HJ On biologically conformal boost dose optimization Phys Med Biol 2003 48 N31 N35 12587912 10.1088/0031-9155/48/2/404 Ljungkvist ASE Bussink J Rijken PFJW Kaanders JH van der Kogel AJ Denekamp J Vascular architecture, hypoxia, and proliferation in first-generation xenografts of human head-and-neck squamous cell carcinomas Int J Radiat Oncol Biol Phys 2002 54 215 228 12182995 10.1016/S0360-3016(02)02938-3 Janssen HL Haustermans KMG Sprong D Blommestijn G Hofland I Hoebers FJ Blijweert E Raleigh JA Semenza GL Varia MA Balm AJ van Velthuysen MF Delaere P Sciot R Begg AC HIF-1A, pimonidazole, and iododeoxyuridine to estimate hypoxia and perfusion in human head-and-neck tumours Int J Radiat Oncol Biol Phys 2002 54 1537 1549 12459383 10.1016/S0360-3016(02)03935-4 Janssen HL Hoebers FJ Sprong D Goethals L Williams KJ Stratford IJ Haustermans HM Balm AJ Begg AC Differentiation-associated staining with anti-pimonidazole antibodies in head and neck tumours Radiother Oncol 2004 70 91 97 15036858 10.1016/j.radonc.2003.09.012 Viola P Wells WM Alignment by Maximization of Mutual Information Int J Comp Vision 1997 24 137 154 10.1023/A:1007958904918 Thorwarth D Eschmann SM Paulsen F Alber M A kinetic model for dynamic [18F]-Fmiso PET data to analys Phys Med Biol 2005 50 2209 2224 15876662 10.1088/0031-9155/50/10/002 Plataniotis GA Theofanopoulou ME Kalogera-Fountzila A Haritanti A Ciuleanou E Ghilezan N Zamboglou N Dimitriadis A Sofroniadis I Fountzilas G Prognostic impact of tumor volumetry in patients with locally advanced head-and-neck carcinoma (non-nasopharyngeal) treated by radiotherapy alone or combined radiochemotherapy in a randomized trial Int J Radiat Oncol Biol Phys 2004 59 1018 1026 15234035 10.1016/j.ijrobp.2004.01.021 Denekamp J Dasu A Inducible Repair and the Two Forms of Tumour Hypoxia – Time fir a Paradigm Shift Acta Oncol 1999 38 903 918 10606420 10.1080/028418699432590	16321146	PMC1325034	CC BY	2021-01-04 16:39:11	no	BMC Cancer. 2005 Dec 1; 5:152	utf-8	BMC Cancer	2,005	10.1186/1471-2407-5-152	oa_comm
==== Front Nutr JNutrition Journal1475-2891BioMed Central London 1475-2891-4-371635172210.1186/1475-2891-4-37ResearchNutrient estimation from an FFQ developed for a black Zimbabwean population Merchant Anwar T [email protected] Mahshid [email protected] Jephat [email protected] Getrude [email protected] Salim [email protected] Population health Research Institute, McMaster University, Hamilton ON, Canada2 Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton ON, Canada3 Department of Medicine, McMaster University, Hamilton ON, Canada4 Department of Physiology, University of Zimbabwe College of Health Sciences, Harare, Zimbabwe2005 13 12 2005 4 37 37 22 9 2005 13 12 2005 Copyright © 2005 Merchant et al; licensee BioMed Central Ltd.2005Merchant 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 little information in the literature on methods of food composition database development to calculate nutrient intake from food frequency questionnaire (FFQ) data. The aim of this study is to describe the development of an FFQ and a food composition table to calculate nutrient intake in a Black Zimbabwean population. Methods Trained interviewers collected 24-hour dietary recalls (24 hr DR) from high and low income families in urban and rural Zimbabwe. Based on these data and input from local experts we developed an FFQ, containing a list of frequently consumed foods, standard portion sizes, and categories of consumption frequency. We created a food composition table of the foods found in the FFQ so that we could compute nutrient intake. We used the USDA nutrient database as the main resource because it is relatively complete, updated, and easily accessible. To choose the food item in the USDA nutrient database that most closely matched the nutrient content of the local food we referred to a local food composition table. Results Almost all the participants ate sadza (maize porridge) at least 5 times a week, and about half had matemba (fish) and caterpillar more than once a month. Nutrient estimates obtained from the FFQ data by using the USDA and Zimbabwean food composition tables were similar for total energy intake intra class correlation (ICC) = 0.99, and carbohydrate (ICC = 0.99), but different for vitamin A (ICC = 0.53), and total folate (ICC = 0.68). Conclusion We have described a standardized process of FFQ and food composition database development for a Black Zimbabwean population. ==== Body Introduction Diet is central in the development of obesity and chronic diseases and is changing rapidly in low and middle income countries [1]. Some of the constraints in studying diet and its correlates in poor countries is that diet varies considerably from place to place and among different socio-economic classes, its assessment methods have evolved relatively recently (1980s and 1990s), and have not been developed for those populations [2]. Moreover, the technical expertise and financial resources to develop these instruments may be lacking. We are conducting a longitudinal investigation in about 120000 people in 14 countries at different stages of development, one of which is Zimbabwe, measuring diet with an FFQ. The FFQ is commonly used for nutritional assessment in large epidemiologic studies, because it measures long-term diet, is quick and comparatively inexpensive to administer, and provides quantitative information on nutrients and foods [2]. To calculate nutrient intake from an FFQ we need to have data from a food composition database that lists the nutrient content of the foods contained in the FFQ. This is critical because the nutrient content of the same food can vary substantially [3]. However, the food composition database that we found for Zimbabwe was 24 years old and was not updated regularly [4]. Moreover, many nutrients were not estimated at all. Therefore we could not use the Zimbabwean food composition database to estimate nutrient intake from the FFQ. We thus created a food composition database for Zimbabwe using the Zimbabwe and USDA nutrient databases. In this paper we briefly describe the development of a semi-quantitative FFQ, and creation of food composition table from USDA database for Black Zimbabwean population. Subjects and methods Study population The urban participants lived in Harare, either in a relatively affluent area, or in an urban low-income area; the rural participants lived in Chidamoyo, 345 km from Harare. All participants were = 35 years of age, and free of any reported co-morbid conditions. The study was approved by the Institutional Review Boards of the participating institutions and the relevant authorities. Development of FFQ There were three major steps in FFQ development. First, we prepared a list of commonly eaten foods in Zimbabwe, including estimates of the usual portion sizes and frequencies of intake. Second, based on this list we developed a long FFQ. Third, we tested the long FFQ in the field and shortened it. We are using this version of the FFQ in our study and are validating by comparing it with multiple 24-hour dietary recalls. Development of the food list We used three approaches to create a food list: we conducted a 24-hour dietary recall in November 2003 among 200 persons in urban and rural areas. Second, we added foods to this list that were commonly eaten in that Zimbabwean population at other times of the year and on special occasions. Third, we further expanded this list by adding foods that were nutrient rich but were not captured in the previous steps (caterpillar for example) from a previously prepared food composition table for Zimbabwe and the observations of the local Zimbabwean nutritionists. We then formatted the foods on the expanded list into questions. We organized the foods items based on their similarities in nutrient content into 6 major food groups as has been done elsewhere [5]. The food groups were: 1. Bread, cereal and starches (different types of bread, sadza, rice, pasta, porridge and bakery products), 2. Meat and eggs(including insects and sea food), 3. Dairy products, 4. Vegetables (fresh and cooked vegetables excluding potatoes), and fruits (fresh and canned), 5. Beverages (fruit juice, alcoholic and non-alcoholic drinks, coffee and tea), 6. Sweets and baked goods (nuts, candy and cake). Frequency of intake To estimate average frequency of intake of most foods in the previous year we used closed-ended responses consisting of 9 categories: Never or less than once/month, 1–3/m, 1/week, 2–4/wk, 5–6/wk, 1/day, 2–3/d, 4–5/d, >6/d [5]. Portion size To obtain food portion size, we physically examined all reported portions and took the most commonly consumed portion as the unit of measurement. Seasonality Unlike in most western countries fruit and vegetable availability and cost vary greatly by season in Zimbabwe. We asked how often on average fruits and vegetables with seasonal variation were consumed in season. Our colleagues in Zimbabwe prepared the list of seasonal fruits and vegetables and estimated the length of season for each item. This allowed us to take seasonality into account to calculate the average daily intakes of such foods. Testing of pilot FFQ We pilot tested the long FFQ by administering it to 100 participants in July 2004, and excluded infrequently eaten foods. Food composition database The food composition database that we found for Zimbabwe was 24 years old and was not updated regularly. Moreover, the reported nutrients in the food composition table included those recommended by Southgate in 1974 [4]. Therefore we could not use the Zimbabwean food composition database to estimate nutrient intake from the FFQ. We therefore created a food composition database for Zimbabwe using the Zimbabwe and USDA nutrient databases. However, the USDA nutrient database has many different entries for the same food with different nutrient compositions. If we arbitratily chose any one of those foods its nutrient composition might have been different from that of the same food used in Zimbabwe. We therefore used the Zimbabwen food composition database to help us choose the same food item from the USDA nutrient databse that was most similar in nutrient composition to the one used in Zimbabwe. For example, among 13 apples in USDA nutrient database, Apples, raw, with skin code 9003 was the most similar to the apple in Zimbabwe food composition table (Table 1). We compared the energy, macronutrient and some micronutrient content of each food item reported in both food composition tables and chose the food from the USDA nutrient database that was closest in nutrient content to the food described in the Zimbabwean food composition table (Table 2). Table 1 Comparison of nutrient composition of an apple described in the local Zimbabwe food composition table with different apples in the USDA nutrient database NDB No Shrt_Desc Water/g Energ/Kcal Protein/g Lipid/g CHO/g Ca/mg Fe/mg P/mg K/mg Na/mg Vit_C/mg Vit B1/mg Vit B2/mg Niacin/mg Vit B6/mg Folate/mg Vit_A RAE 09003 Apple, raw w Skin 85.56 52 0.26 0.17 13.81 6 0.12 11 107 1 4.6 0.017 0.026 0.091 0.041 3 3 09004 Apple raw wo, skin 86.67 48 0.27 0.13 12.76 5 0.07 11 90 0 4 0.019 0.028 0.091 0.037 0 2 09005 Apple raw wo, skin Ckd bld 85.47 53 0.26 0.36 13.64 5 0.19 8 88 1 0.2 0.016 0.012 0.095 0.044 1 2 09006 Apple raw wo, skin / ckd microwave 84.63 56 0.28 0.42 14.41 5 0.17 8 93 1 0.3 0.017 0.011 0.061 0.046 1 2 09007 Apple cnd, swtnd, sliced, dr nd, unhtd 82.36 67 0.18 0.49 16.7 4 0.23 5 68 3 0.4 0.009 0.01 0.073 0.044 0 3 09008 Apple cnd, swtnd Sliced, drnd, htd 82.28 67 0.18 0.43 16.84 4 0.24 6 70 3 0.2 0.009 0.01 0.081 0.044 0 3 09009 Apple, Dehyd (lo moist) sulfured unckd 3 346 1.32 0.58 93.53 19 2 55 640 124 2.2 0.046 0.13 0.68 0.28 1 4 09010 Apple, Dehyd (lo moist) sulfured stwd 79.36 74 0.28 0.12 19.91 4 0.43 12 136 26 0.6 0.008 0.029 0.14 0.054 0 1 09011 Apple, dried sulfured, unckd 31.76 243 0.93 0.32 65.89 14 1.4 38 450 87 3.9 0 0.159 0.927 0.125 0 0 09012 Apple, dried sulfured, stwd, wo/sugar 84.13 57 0.22 0.07 15.32 3 0.33 9 105 20 1 0.006 0.019 0.129 0.05 0 1 09013 Apple, dried sulfured, stwd, w/sugar 78.76 83 0.2 0.07 20.73 3 0.31 8 98 19 0.9 0.006 0.018 0.121 0.047 0 1 09014 Apple, frz, unswtnd, unhtd 86.85 48 0.28 0.32 12.31 4 0.18 8 77 3 0.1 0.013 0.011 0.042 0.034 1 2 09015 Apple, frz, unswtnd, htd 87.16 47 0.29 0.33 12 5 0.19 8 76 3 0.4 0.014 0.011 0.043 0.032 1 1 Apple_Zimbabwe 84.6 56.2 0.4 0.6 13.1 6.5 0.6 10.8 106 1.5 5.4 0.03 0.03 0.15 0.03 3.9 13.39 Table 2 Comparison of nutrient content of some foods calculated from Zimbabwe and USDA food composition tables Foods Code Energy kcal Protein g Fat g CHO g Ca mg P mg Fe mg K mg Na mg Vit† A RE Vit C mg B1 mg B2 mg B3 mg B6 mg B12 mg Folate mg Corn flour Zim_FCT* 356.2 4.5 2.1 84.0 7.0 45.7 1.6 30.5 26.0 0 0 0.01 0 0.10 tr** 0 tr Corn, white 20314 365 9.42 4.74 74.2 7 210 2.71 287 35 0 0 3.627 0.622 0 Corn flour, whole-grain, white 20316 361 6.93 3.86 76.85 7 272 2.38 315 5 0 0 1.900 0.370 0 25 Cornmeal, whole-grain, white 20320 362 8.12 3.59 76.89 6 241 3.45 287 35 0 0 3.632 0.304 0 25 Shakata (Mobola Plum) this must be a mistake 167.5 1.2 0 0.5 41.9 2.2 158.3 55.7 0.5 Plums, raw 09279 46 0.70 0.28 11.42 6 16 0.17 157 0 17 9.5 0.417 0.029 0 5 Plums, canned, purple, water pack, solids and liquids 09281 41 0.39 0.01 11.03 7 13 0.16 126 1 46 2.7 0.370 0.027 0 3 Plums, canned, purple, light syrup pack, solids and liquids 09283 63 0.37 0.10 16.28 9 13 0.86 93 20 12 0.4 0.297 0.027 0 3 Guavas this also 66 1 0.5 14.6 16.6 26.0 0.9 290 4.0 50.0 221.4 0.05 0.04 1.1 0.0 Guavas, common, raw 09139 68 2.55 0.95 14.32 18 40 0.26 417 2 31 228.3 1.084 0.110 0 49 Chicken raw Zim_FCT 195.1 19 12.9 0 12 179 1.3 227 68.3 105 0.9 0.8 0.15 7.7 0.3 0.4 5 Chicken raw 5006 215 18.6 15.1 0 11 147 0.9 189 70 41 1.6 0.06 0.12 6.8 0.35 0.3 6 Chicken raw, meat and skin stewing 5123 258 17.6 20.3 1.04 10 172 1.04 204 71 52 0 0.11 0.17 6.26 0.33 0.3 6 Chicken roast, skin and meat Zim_FCT 216 26.7 8.2 0 12 214 1.5 308 82.4 45 0 0.7 0.23 5.6 Chicken roast, skin and meat 5111 216 17.1 15.9 0 10 166 1.01 196 68 38 0 0.06 0.12 6.57 0.32 0.3 6 Chicken roasted, skin and meat 5055 239 28.2 13.2 0 24 165 1.39 237 96 28 0 0.07 0.22 7.069 0.34 0.3 7 Chicken, cooked, roasted 5069 216 27 11.2 0 12 175 1.33 229 90 30 0 0.07 0.22 5.994 0.34 0.3 8 Spinach (Mova) Zim_FCT 15.9 1.2 0.2 2.4 33.1 30.3 0.9 228.3 8.9 419.4 11.10 0.07 0.08 0.35 0.06 0 32.80 Malabar, Spinach, cooked 11986 23 2.98 0.78 2.71 124 36 1.48 256 55 58 5.9 0.787 0.086 0 114 New Zealand spinach, cooked, boiled, drained, without salt 11277 12 1.30 0.17 2.20 48 22 0.66 102 107 181 16.0 0.390 0.237 0 8 * Zim_FCT = Zimbabwe food composition table ** Tr = trace † Vit = vitamin For mixed dishes we entered local recipes into an Excel spreadsheet and applied yield and retention factors taken form USDA Handbook No.102 (food yields, summarized by different stages of preparation) [3] to obtain the nutrient composition for given portions of Zimbabwean mixed dishes, thus accounting for the method of food preparation. We then estimated nutrient content for the most commonly used portion of that mixed dish. For those food items, which were not available in the USDA nutrient database such as caterpillar, we chose a food item from the same food category (sushi caterpillar) [6] and imputed those nutrient values in our database for caterpillar. Finally, for some other foods which were not available in USDA we found the common name, scientific name of local food items and matched them with the same scientific name in English and then chose similar food item from USDA. For example, taro is Colacasia antiquorum and is from the yam family. Therefore, we imputed yam nutrient content for taro. Recipe analysis To calculate the nutrient content of mixed dish gathered local recipes as described elsewhere [7]. There were 112 recipes in the recipe database and we calculated the nutrient content of those recipes based on the chosen food items from USDA, as described in the previous section. We did not choose enriched foods items to estimate nutrient content because these are not generally used in Zimbabwe. In nutrient estimation we took into account the loss of minerals, vitamins and moisture in food preparation, such as boiling, frying or steaming. From this information we were able to calculate the nutrient content of 100 g or a portion of food. Data management and statistical analysis We entered information from the FFQ and basic demographic characteristics of the participants into SPSS. We calculated the frequency of intake of a portion of each food on the FFQ. We administered the FFQ to 100 people and estimated the mean and standard deviation (SD) of nutrient intake from 30 commonly eaten foods using the USDA and Zimbabwean food composition tables. To compare the estimates we calculated intraclass correlations. Results Of the 200 persons who responded to the 24 hr Dr and FFQ slightly more than half were women (Table 3). The mean age was nearly 51 years and ranged from 34–93 years. Many people (40%) did not respond to the question on income. Among the respondents more than half reported incomes in the very low category. There were few illiterate or never married persons in this sample. The mean age for those who did not report their income was 54 ± 15.4 years old and they were mostly illiterate or had only primary school education (16.1% illiterate, 54% primary and 30% secondary level education). Table 3 Characteristics of subjects who participated in 24 hr dietary recall and pilot testing Long-FFQ* FFQ 24 hr DR Women N = 55 N = 43 Age (years) 50.6 (12.1) 52.3 (14.4) BMI (kg/m2) 26.3 (4.8) 23.6 (5.0) Men (N = 43) N = 45 Age (years) 51.4 (16.1) 51 (13.6) BMI (kg/m2) 23.7 (6.0) 21.8 (3.6) Income (ZW $)** Very-Low 22 (52.4%) 63 (88.7%) Low 15 (35.7%) 7 (9.9%) Middle 5 (11.9%) 1 (1.4%) No response 56 24 Marital status Never married 3 (3%) 4 (4.5%) Married 72 (72.2%) 48 (53.9%) Widow, divorced or separated 25 (24.2%) 37 (41.6%) Education Illiterate 10 (10.1%) 20 (22.5%) Primary or secondary school 87 (87.8%) 61 (68.6%) University, college 2 (2%) 5 (5.6%) Profession Professionals 1 (1.3%) 4 (5.6%) Technicians and associated worker 17 (22.1%) 3 (4.2%) Elementary occupation 22 (28.6%) 40 (56.3%) Homemaker 37 (48.1%) 3 (4.2%) * The numbers in each of the categories in this table are different because the number of reponses we received varied ** Exchange rate US $ 1 = 5300 Zimbabwe dollar Very low < 3 × 105 ZW $/month Low 3 × 105 to 1 × 106 ZW $/month Middle 1 × 106to 2.5 × 106 ZW $/month High >2.5 × 106 ZW $/month The staple food in this population was maize. It was cooked in different ways such as sadza, samp, or mealie porridge (Table 4). Lacto, a commercially produced fermented milk product, was a frequently consumed dairy food. Certain types of caterpillars were eaten, either dried or fried. Matemba (small fish) were preserved and eaten in small quantities with other food like condiments. In addition people ate many fruits and vegetables unique in that area (Table 4). A copy of the FFQ is in Appendix 1. Table 4 Description of some foods unique to Zimbabwe Food Description Bread, cereal and starches Sadza Stiff porridge (prepared from meal of maize, millet, sorghum or rice) and contain a small amount of fibre. The meal of maize is the staple food in Zimbabwe Samp (Mashakada) Boiled maize grain (previously dried), sometimes its degerminated broken grain, pounded and boiled Mealie Porridge Ground corn, boiled not too thick, if more is added becomes sadza, other foods can be added like peanut butter (sometimes margarine), sugar (amounts depends on quantity), milk, some people even add bread Pumpkin Porridge African bread Mash prepared from cattle melon or pumpkin ingredients: corn flour, egg, sugar and water Fruits Baobab Adansonia digitata Paw Paw Carica Papaya, Papaya Naartjie Citrus aurantium, Tangerine Shakata Parinari Curatellifolia, Mobola Plum Vegetables Gourd (dende) Lagenaria siceraria, Containe of groundnut butter Taro Colacasia antiquorum, Yam Mowa Decumbent perennial weed; Amaranthus thunbergii& graecizans (leaves cooked as spinach) Meat Caterpillar Edible caterpillar (dark, with white fur, found on musasa trees. Eaten fried or sun-dried Matemba, Kapenta Mouse Small fish caught mostly from lake Kariba Wild mice (not rats or house mice), these are trapped in the wild or fields. Eaten cooked with soup or dried. Milk and dairy product Lacto Commercially produced fermented milk and taste sour * Name of food in Shona In Table 3 we have presented the reported consumption of selected foods from the FFQ. Almost all the participants reported eating sadza at least 5 times a week, but there was more variation in reported intakes of other foods in that group, for instance porridge (Table 5). About half the participants ate matemba and caterpillar more than once a month, and 44% reported they ate mice at least monthly. The main source of dairy was lacto. Most people ate mango in season and banana the year round. People either never drank coffee or consumed it very frequently (Table 5). Table 5 Reported consumption of selected foods in the past year in Zimbabwe from FFQ Foods Frequency of consumption Never, <1/mo 1–3 / mo 1/wk 2–4 / wk ≥ 5 / wk Bread, cereal and starches Sadza (maize) 0 0 0 4 (2%) 97 (98%) White bread 40 (40.8%) 16 (16.3%) 4 (4.1%) 7 (7.1%) 31 (31.6%) Starch roots and tubers Sweet potato 3 (3%) 9 (9.1%) 9 (9.1%) 32 (32.3%) 44 (44.5%) Meat and eggs Matemba 46 (46.5%) 24 (24.2%) 16 (16.2%) 11 (11.1%) 2 (2%) Caterpillar 52 (52.5%) 28 (28.3%) 5 (5.1%) 12 (12.1%) 2 (2%) Mice 66 (66.7%) 12 (12.1%) 11 (11.1%) 3 (3%) 7 (7.1%) Dairy products Lacto 36 (36.4%) 38 (38.4%) 9 (9.1%) 15 (15.2%) 1 (1%) Whole milk 79 (79.8%) 9 (9.1%) 2 (2%) 6 (6.1%) 3 (3%) Fruits in season Mango 2 (2%) 3 (3%) 12 (12.1%) 19 (19.2%) 63 (63.8%) Guava 9 (9.1%) 9 (9.1%) 10 (10.1%) 24 (24.2%) 47 (47.5%) Fruits out of season Banana 33 (33.3%) 25 (25.3%) 7 (7.1%) 10 (10%) 24 (24.2%) Paw Paw 77 (77.8%) 16 (16.2%) 3 (3%) 1 (1%) 2 (2%) Type of oil most commonly used for cooking (N,%) Vegetable oil 98 (99%) We obtained very similar estimates by the USDA and Zimbabwean food composition tables for energy intake from 30 selected foods (1492 kcal versus 1401 kcal, ICC = 0.99), and carbohydrate (300 g versus 289 g, ICC = 0.99) (Table 6). There were differences in estimates obtained by the USDA and Zimbabwean food composition tables respectively for the micronutrients especially vitamin A (4362 versus 1240 retinol equivalents based on μg, ICC = 0.53), and total folate (99 versus 309 μg/d, ICC = 0.68). A copy of the Zimbabwean food composition table is in Appendix 2. Table 6 Comparison of nutrient intake of 100 participants by USDA and Zimbabwe food composition table Nutrients USDA Mean (SD) Zim FCT* Mean (SD) Intraclass correlation Energy (kcal) 1492 (661) 1401 (597) 0.99 Protein (g) 33 (15) 46 (25) 0.94 Fat (g) 20 (12) 26 (15) 0.97 Carbohydrate (g) 300 (131) 289 (133) 0.99 Calcium (mg) 443 (222) 360 (226) 0.98 Phosphorus (mg) 560 (250) 987 (367) 0.94 Potassium (mg) 2746 (1830) 3955 (2177) 0.98 Sodium (mg) 1160 (401) 1561 (644) 0.78 Vitamin A, (RE)** 4362 (2978) 1240 (789) 0.53 Vitamin C (mg) 567 (462) 506 (523) 0.99 Total folate (μ-g) 99 (66) 309 (183) 0.68 * FCT = Food composition table ** RE = Retinol equivalents Discussion In this paper we have described the development of an FFQ and a food composition database for the Black Zimbabwean population. We have also provided a copy of the FFQ and food composition table for the benefit of other researchers in the field. Our sample included a wide cross-section of Zimbabwean society – men, women, rich and poor, urban and rural. We did this to get a list of nutrient rich foods that were eaten by a wide cross-section of people in the community that would help in categorizing people by nutrient intake. There was a wide spread of intakes of porridge and white bread and are examples of discriminating foods (Table 4). We tried to phrase the questions in the FFQ so that the respondent could picture the food we were asking about in his or her mind. For example the portion size of meat in Zimbabwe is considerably smaller than that in western countries. Matemba (fish) is eaten in very small quantities almost like a condiment. We took these factors into account when designing the questions and analyzing nutrient content. The questions were short, used average portion sizes, and categories of intake so that the interviewer simply checked a category to estimate frequency of intake, and did not have to estimate portion size. As a result its administration (by a trained interviewer) took on average 10 minutes. The instrument has face validity. For example, sadza is the staple food in Zimbabwe and almost all the participants reported that they ate it at least 5 times a week. We are conducting a formal validation of this instrument by comparing nutrient and food intakes assessed from the FFQ to multiple 24-hour recalls. By using the USDA food composition database as our major source of information we ensured that nutrient estimates were available for most foods, and the assays to estimate nutrients were current [3]. By using the Zimbabwean food composition database as a guide we were able to select the food item from the USDA food composition database that was closest in nutrient content to the corresponding Zimbabwean food. The choice was probably reasonable as the nutrient estimates from the two methods were very similar for most nutrients. The discrepancies, particularly for vitamin A and folate, were probably because certain foods were not analyzed for these nutrients and hence there were missing values in the Zimbabwean table, or old methods of nutrient analysis were used. The methods to estimate vitamin A and folate from foods have changed in the USDA nutrient database [3], but those for macronutrients have not. Instead of the backward selection procedure to develop a food list, we collected a 24-hour dietary data, supplemented with input from local nutritionists. Such an approach has also been used for the development of FFQs [2]. Moreover the variation of food intake in low income countries is low [2] so it is unlikely that we missed out any important food in the FFQ. The participants in this study were persons over 35 years of age. The FFQ could likely be used among younger adults as well because it probably captures most of the foods eaten by this group as well but this caveat should be borne in mind when extending its use. Another limitation was that we did not have nutrient contents of certain foods, for example, caterpillar. We overcame this by using the closest category (sushi caterpillar) that was in the ESHA food composition table to that food. These assumptions would inevitably be the source of some errors in nutrient estimation. The discrepancies in nutrient estimates obtained from the same FFQ data but different food composition tables underscores the need to have standard methods, not only to develop the FFQ, but also food composition tables. While there is much literature on methods for FFQ development there is much less information on food composition database development. We have described a standardized process of FFQ and food composition database development for a Black Zimbabwean population. List of abbreviations FFQ Food Frequency Questionnaire 24 hr DR 24 hour Dietary Recall USDA United State Department of Agriculture SPSS Statistical Packages for Social Sciences ESHA The Food Processor and Genesis SD Standard deviation Authors' contributions ATM Participated in design of study, performed statistical analysis, drafted the manuscript MD Participated in design of study, coordinated and performed statistical analysis, helped to draft the manuscript JC Coordinated study in Zimbabwe, helped to draft the manuscript GT Facilitated data collection in Zimbabwe SY Participated in securing funding, helped to draft the manuscript All authors read and approved the final manuscript. Figure 1 The process of choosing Food items from USDA nutrient database. (FCT* : Food Composition Table) ESHA: Computer program to estimate nutrient content of foods Supplementary Material Additional File 1 pdf file containing the FFQ for Zimbabwe Click here for file Additional File 2 Excel file containing the food composition database for Zimbabwe Click here for file Acknowledgements We are grateful to the participants of the study for their cooperation, the Population Health Research Institute for providing the funds, and Ms Pam Mackie for secretarial support. ==== Refs Popkin BM Urbanization, lifestyle changes and the nutrition transition World Development 1999 27 1905 1916 10.1016/S0305-750X(99)00094-7 Willett W Willett W Food Frequency Methods Nutritional epidemiology 1998 5 Second New York, Oxford University Press 74 100 USDA National Nutrient Database for Standard Reference Release 17 2004 Chitsiku IC Nutritive Value of Foods in Zimbabwe 2000 University of Zimbabwe Rimm EB Giovannucci EL Stampfer MJ Colditz GA Litin LB Willett WC Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals Am J Epidemiol 1992 135 1114 1126 1632423 ESHA: The food Processor 2004 Dehghan M Al Hamad N Yusufali A Nusrath F Yusuf S Merchant AT Development of a semi-quantitative food frequency questionnaire for use in United Arab Emirates and Kuwait based on local foods Nutr J 2005 4 18 15921524 10.1186/1475-2891-4-18	16351722	PMC1325035	CC BY	2021-01-04 16:23:37	no	Nutr J. 2005 Dec 13; 4:37	utf-8	Nutr J	2,005	10.1186/1475-2891-4-37	oa_comm
==== Front Microb Cell FactMicrobial Cell Factories1475-2859BioMed Central London 1475-2859-4-331633664710.1186/1475-2859-4-33ReviewRecombinant production of bacterial toxins and their derivatives in the methylotrophic yeast Pichia pastoris Gurkan Cemal [email protected] David J [email protected] Department of Biochemistry, University of Cambridge, 80 Tennis Court Road, Cambridge CB2 1GA, UK2 Department of Cell Biology, The Scripps Research Institute, 10550 North Torrey Pines Road, Mail Drop: MB6, La Jolla, CA 92037, USA2005 7 12 2005 4 33 33 18 10 2005 7 12 2005 Copyright © 2005 Gurkan and Ellar; licensee BioMed Central Ltd.2005Gurkan and Ellar; 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 methylotrophic yeast Pichia pastoris is a popular heterologous expression host for the recombinant production of a variety of prokaryotic and eukaryotic proteins. The rapid emergence of P. pastoris as a robust heterologous expression host was facilitated by the ease with which it can be manipulated and propagated, which is comparable to that of Escherichia coli and Saccharomyces cerevisiae. P. pastoris offers further advantages such as the tightly-regulated alcohol oxidase promoter that is particularly suitable for heterologous expression of foreign genes. While recombinant production of bacterial toxins and their derivatives is highly desirable, attempts at their heterologous expression using the traditional E. coli expression system can be problematic due to the formation of inclusion bodies that often severely limit the final yields of biologically active products. However, recent literature now suggests that P. pastoris may be an attractive alternative host for the heterologous production of bacterial toxins, such as those from the genera Bacillus, Clostridium, and Corynebacterium, as well as their more complex derivatives. Here, we review the recombinant production of bacterial toxins and their derivatives in P. pastoris with special emphasis on their potential clinical applications. Considering that de novo design and construction of synthetic toxin genes have often been necessary to achieve optimal heterologous expression in P. pastoris, we also present general guidelines to this end based on our experience with the P. pastoris expression of the Bacillus thuringiensis Cyt2Aa1 toxin. ==== Body Review With the advent of modern molecular biology, recombinant expression is now routinely used for the production of proteins of sufficient purity and quantity for their functional characterization and/or use in downstream applications. For example, heterologous expression systems have facilitated the development of recombinant vaccines against the bacterial toxins that are the causative agents of human diseases such as tetanus, botulism and cholera [1-4]. Concurrently, biosynthesis of novel proteins is feasible by engineering of recombinant DNA constructs that comprise of unrelated genes, which are also often from very diverse organisms. For instance, immunotoxins are therapeutic agents that are typically composed of DNA encoding a tumour-specific antibody fragment fused to a gene coding for a highly potent bacterial toxin or its subunits [5]. Despite their crucial roles in vaccine development, therapeutic applications, control of crop pests and disease vectors, as well as in basic research and functional characterization, heterologous expression of bacterial genes and their novel recombinant fusions may still pose unique challenges. For instance, bacterial toxins often have deleterious effects on the host cell physiology that may limit the final yields or may even exclude the use of certain recombinant expression systems altogether. Furthermore, bacterial genes may be unsuitable for heterologous expression in certain recombinant expression hosts due to the inherent features of the prokaryotic DNA sequences such as differences in codon usage and/or high A+T-content that may contain cryptic eukaryotic polyadenylation signals. Finally, if the bacterial toxins or their derivatives are destined for clinical use, more stringent recombinant production methods are necessary to ensure utmost purity, hence in some cases further limiting the choice of heterologous expression hosts. In this manuscript, we review the use of the Pichia pastoris (P. pastoris) expression system for the recombinant production of bacterial toxins and their derivatives, with special emphasis on their potential clinical applications. P. pastoris as a recombinant expression host As a methylotrophic yeast, P. pastoris can use methanol as its sole carbon and energy source in the absence of a repressing carbon source [6,7]. The first step in the metabolism of methanol is its oxidation to formaldehyde by the enzyme alcohol oxidase (AOX) using molecular oxygen. In addition to formaldehyde, this reaction also generates hydrogen peroxide. To avoid hydrogen peroxide toxicity, methanol metabolism takes place within a specialised organelle called the peroxisome that sequesters the toxic by-products away from the rest of the cell. Since AOX has a poor affinity for oxygen, P. pastoris compensates by generating large amounts of the enzyme, which can accumulate to comprise up to 30% of total cell protein (tcp) during induction with methanol [8]. There are now a variety of vectors available that are mostly based on the powerful AOX1 promoter for the regulated overproduction of intracellular and secreted proteins in P. pastoris [9-11]. In contrast to the prokaryotic recombinant expression systems such as those based on Escherichia coli (E. coli), P. pastoris possesses eukaryotic features such as a secretory pathyway based on compartmentalized endomembranes, which is better equipped for post-translational modifications. Consequently, P. pastoris allows efficient secretory expression of complex recombinant proteins with correct intra- and inter-molecular disulphide bonds that do not require additional in vitro unfolding and refolding strategies. Furthermore, secreted expression in P. pastoris is a particularly attractive option because while it only secretes low-levels of endogenous proteins, it is capable of high-level secretion of the heterologously expressed proteins. P. pastoris can also be grown on simple, chemically-defined media, therefore secretion of the heterologous protein often becomes an effective purification step itself. Other key features that contributed to the rapid emergence of P. pastoris as a robust recombinant expression host include: (1) the speed, ease and cost-effectiveness with which it can be manipulated and propagated compared to the other eukaryotic expression systems [12], (2) possession of tightly-regulated promoters, such as that of the alcohol oxidase 1 gene (AOX1), which is uniquely suited for the controlled expression of foreign genes [13,14], (3) synthesis of N-linked glycosylation moieties that resemble the mammalian high-mannose type [15], and (4) a strong preference for aerobic growth, a key physiological trait that greatly facilitates culturing at high cell densities relative to the fermentative yeast, Saccharomyces cerevisiae (S. cerevisiae). Indeed, P. pastoris can be grown up to 130 g·l-1 dry cell weight on simple defined media [6]. Generally an immediate improvement in the percentage yield of heterologous protein expression is also observed on going from shake-flask cultures to bioreactor cultures [6]. Heterologous expression of bacterial toxins and their derivatives in P. pastoris As discussed in the previous section, P. pastoris is a popular recombinant expression host for a wide variety of prokaryotic and eukaryotic proteins [6,7]. Here we present a recent literature survey of the bacterial toxins and/or their derivatives that have been successfully produced in P. pastoris (Table 1). Table 1 Bacterial toxins and their derivatives successfully expressed in P. pastoris. The bacterial toxin and the species it is originating from are given, along with brief notes on the specifics of the reported recombinant expression strategies. Bacterial toxin (species) Remarks (expression culture type) [reference] Final Yield§ TeNT(HC) (Clostridium tetani) intracellular expression† of a synthetic‡ gene encoding the tetanus toxin fragment C (B) [18] 12 g·l-1 culture* BoNTA(HC) (Clostridium botulinum) intracellular expression† of a synthetic gene‡ encoding the heavy fragment C of the botulinum neurotoxin serotype A [BoNTA(HC)] (B) [19, 22-25] 770 mg·l-1 culture BoNTB(HC) (Clostridium botulinum) intracellular expression† of a synthetic gene‡ encoding the heavy fragment C of the botulinum neurotoxin serotype B [BoNTB(HC)] (B) [1, 20, 24] 390 mg·kg-1 cells BoNTC1(HC) (Clostridium botulinum) intracellular expression† of a synthetic gene‡ encoding the heavy fragment C of the botulinum neurotoxin serotype C1 [BoNTC1(HC)] (B) [25] 200–500 mg·kg-1 cells BoNTE(HC) (Clostridium botulinum) intracellular expression of a synthetic gene‡ encoding the heavy fragment C of the botulinum neurotoxin serotype E [BoNTE(HC)] (B) [25] 200–500 mg·kg-1 cells BoNTF(HC) (Clostridium botulinum) intracellular expression of a synthetic gene‡ encoding the heavy fragment C of the botulinum neurotoxin serotype F [BoNTF(HC)] (B) [21, 26] 240 mg·kg-1 cells DT (Corynebacterium diphtheriae) secreted expression of a synthetic gene‡ encoding the truncated diphtheria toxin (DT) fused to a bivalent antibody fragment (B) [30-33] 120 mg·l-1 culture* BSP1 and BSP2 (Bacillus sphaericus) intracellular co-expression† of synthetic genes‡ encoding the mosquitocidal B. sphaericus polypeptides 1 and 2 (BSP1 and 2) (SF) [44] <30% tcp* Cry2 (Bacillus thuringiensis) intracellular expression of Cry2 using the native bacterial DNA sequence (SF) [43] N.D. Cyt2Aa1 (Bacillus thuringiensis) intracellular expression of a synthetic gene‡ encoding Cyt2Aa1 (SF) [34] ~1 mg·l-1 culture* Cyt2Aa1 (Bacillus thuringiensis) synthetic gene‡ encoding Cyt2Aa1 fused to a human scFv; secretory targeting resulted in ER-retention of the recombinant product (SF) [35] 10 mg·l-1 culture Ace (Vibrio cholerae) secreted expression of the accessory cholera enterotoxin (Ace) using the native bacterial DNA sequence (SF) [28] 7 mg·l-1 culture* Cef (Vibrio cholerae) secreted expression of Chinese hamster ovary (CHO) cell-elongating factor (Cef) using the native bacterial DNA sequence (SF) [29] N.D. CTB (Vibrio cholerae) secreted co-expression of the cholera toxin subunit B (CTB) and CTB-viral antigen fusion protein using the native bacterial DNA sequence (SF) [4] N.D. LTB (Escherichia coli) secreted expression of the heat-labile enterotoxin subunit B (LTB) using the native bacterial DNA sequence (SF) [27] 8 mg·l-1 culture LTB (Escherichia coli) intracellular expression of a LTB and a viral antigen fusion protein using the native bacterial DNA sequence (SF) [27] N.D. †Using P. pastoris transformants that are selected for the presence of multiple copies of the chromosomally-integrated heterologous expression cassettes; ‡synthetic gene with optimal P. pastoris codon usage and reduced A+T-content; §only the highest final yields are reported in this table; estimated total expression; (SF): shake-flask culture, (B): bioreactor culture; N.D.: no data available; ER: the endoplasmic reticulum. Experience with the recombinant production of the Clostridium neurotoxin fragments in P. pastoris provides good examples for the typical problems encountered with the heterologous expression of bacterial toxins in this yeast and the subsequent high yields attainable once these problems are properly addressed. Clostridium botulinum is the causative agent of botulism, which is a severe neuroparalytic disease brought about by one of the seven antigenically distinct neurotoxin (BoNT) variants (A, B, C1, D, E, F and G) produced by this bacterium [1-3]. Similarly, Clostridium tetani produces tetanospasmin or the tetanus neurotoxin (TeNT) that causes the spastic paralysis condition associated with the tetanus disease. Both TeNT and the BoNT variants are potent exotoxins that are initially synthesized as a single polypeptide chain that typically undergoes subsequent proteolytic processing into a heterodimer of heavy and light chains bound together by a disulphide bond. In both TeNT and the BoNT variants, the carboxyl-terminal domain of the heavy chain (HC) is non-toxic and associated with binding to specific receptors present on the target nerve cells, and since it is antigenic, it has been exclusively used for vaccine development [1,2]. Currently, a pentavalent botulinum toxoid from natural sources composed of variants A through E and a toxoid of variant F are used to immunize at-risk individuals, such as scientists and health care workers that handle BoNT or armed forces personnel that may be subject to weaponized forms of the bacterial toxin [2,3]. However, this strategy has many shortcomings because: (1) C. botulinum produces only low-levels of the most toxin variants, (2) large-scale production is very costly and dangerous, requiring dedicated facilities in accordance with the current Good Manufacturing Practices, (3) the final products are whole toxins that are only partially homogenous, which may in turn influence immunogenicity or reactivity of the vaccine, and (4) the toxoiding process involves the use of chemicals such as formaldehyde and thimerosal that are still present in the final product formulation, hence rendering it reactogenic [2,3]. Consequently, there is a great demand for the development of a new generation of recombinant vaccines that would alleviate many of the problems associated with the current toxoid formulations. Recombinant tetanus neurotoxin fragment C [TeNT(HC)] was the first bacterial toxin that was successfully expressed in P. pastoris [16]. Earlier attempts at heterologous TeNT(HC) expression in E. coli and S. cerevisiae necessitated the use of synthetic genes due to the unfavorable codon bias of the A+T-rich C. tetani DNA sequence and the presence of cryptic polyadenylation signals that led to premature mRNA transcript termination in yeast [17]. Following a similar approach, Clare et al. used a synthetic gene with altered codon usage that had a substantially reduced A+T-content to achieve recombinant production of TeNT(HC) in P. pastoris with final yields as high as 12 g per liter of bioreactor culture [18]. Recombinant production in P. pastoris of BoNT variants was also very successful when using synthetic genes that were optimized for heterologous expression in this yeast [1,2,19-26]. As in the case of TeNT(HC) [18], heterologous expression of BoNTA(HC), B(HC), and E(HC) in P. pastoris was also attempted by secretory targeting of the recombinant products [1,2]. However, in both cases the recombinant proteins secreted into the culture medium were glycosylated due to the presence of fortuitous N-linked glycosylation sites in the prokaryotic primary amino acid sequences. This glycosylation rendered them immunologically inactive, hence unfit for vaccine development unless a costly in vitro deglycosylation step was carried out [1,2,18]. Accordingly, both TeNT(HC) and the BoNT(HC) variants are now exclusively produced by intracellular heterologous expression in P. pastoris (Table 1). For vaccine development, production in P. pastoris offers additional advantages over E. coli in avoiding the formation of inclusion bodies during heterologous expression and eliminating the potential presence of bacterial endotoxins requisite to achieve Food and Drug Administration licensure [2,3]. P. pastoris also proved very useful in the development of vaccines for the heat-labile enterotoxin (LT) of E. coli and the cholera toxin (CT) of Vibrio cholerae, which both cause diarrhea in humans [4,27]. Both LT and CT have a hetero-hexameric structure consisting of a toxic A subunit and five non-toxic B subunits that function in binding to the target cells. The LT subunit B (LTB) was successfully expressed in P. pastoris using the bacterial gene and efficiently secreted into culture medium in a native-like pentameric form that was biologically active and immunogenic [27]. Fingerut et al. also reported intracellular expression in P. pastoris of a genetic fusion of LTB with a viral antigen to demonstrate the adjuvant activity of recombinant LTB produced in the methylotrophic yeast [27]. Similarly, CT subunit B (CTB) and a genetic fusion of CBT with a viral vaccine antigen were successfully co-expressed in P. pastoris using the native bacterial CBT gene [4]. This allowed efficient co-secretion of the recombinant CBT and CBT fusion proteins into the culture medium in a biologically active hetero-pentameric form, which could then be purified by a single-step affinity-tag based chromatography strategy. Other V. cholerae toxins also successfully expressed in P. pastoris are the accessory cholera toxin (Ace) and the Chinese hamster ovary (CHO) cell-elongating factor (Cef) [28,29]. Despite having a key role in V. cholerae pathogenesis, the accessory cholera toxin (Ace) is produced only at low levels by it natural host, which initially hampered its further characterization [28]. While recombinant production of Ace in E. coli was not feasible due to inherent toxicity effects to the host cells, Trucksis et al. reported subsequent success using the P. pastoris expression system, where secreted enterotoxin could be purified to homogeneity in a biological active form and at levels as high as 7 mg·l-1 culture [28]. Bacterial toxins have further clinical applications, such as in the development of novel therapeutic agents. These include immunotoxins (ITs) comprised of a potent bacterial toxin that is recombinantly fused to a cell-binding ligand such as an antibody fragment specific for tumor cells [5]. Recombinant expression of ITs can be particularly challenging due to the deleterious effects of the toxin moiety on the host cell physiology and/or the presence of multiple disulphide bonds in the antibody fragment moiety that are requisite for its function. However, recent literature suggests that the P. pastoris expression system might be an attractive alternative for recombinant IT production. For example, Woo et al. reported successful fine-tuning of the P. pastoris expression system for the production of a recombinant IT based on a truncated version of the diphtheria toxin (DT) [30-32]. This strategy necessitated the construction of a synthetic gene optimized for P. pastoris expression that encoded the first 390 amino acids of the DT toxin (DT390) previously shown to be the minimum DT truncate suitable for IT production [30]. The multi-domain DT390-based IT could be efficiently secreted by P. pastoris in a biologically active form and at yields as high as 10 mg·l-1 of shake-flask culture [30]. Notably, P. pastoris proved to be a particularly suitable recombinant expression host in this case as it has a higher tolerance for DT toxicity compared to S. cerevisiae and other eukaryotes. While the introduction of a DT resistant mutation into the chromosomal EF-2 locus of P. pastoris did not help to further increase the final yields of biologically active IT, secreted expression levels as high as 120 mg·l-1 culture were eventually achieved using a bioreactor and empirically optimized methanol induction conditions [31-33]. We have recently reported the successful recombinant production in P. pastoris of the Cyt2Aa1 δ-endotoxin from the Bacillus thuringiensis (B. thuringiensis) subspecies kyushuensis, as well as that of a membrane-acting Cyt2Aa1-based IT [34,35]. B. thuringiensis is a ubiquitous aerobic, gram-positive bacterium that is best known for its crystalline δ-endotoxin inclusions produced during sporulation [36]. These δ-endotoxins are pore-forming proteins with very specific larvicidal activities for insects in the order of Lepidoptera, Coleoptera and Diptera. All active δ-endotoxins belong to either the Cry or Cyt family of toxins that share very little amino acid sequence identity but are both initially produced as protoxins that need to be solubilized at the appropriate pH prior to activation by proteolytic processing. Cyt toxins are smaller than the Cry toxins and are further distinguished from the latter by: (1) their highly specific mosquitocidal activity in vivo, (2) their broad cytolytic activity to a variety of invertebrate and vertebrate cells in vitro after solubilisation and activation by proteolytic processing, and (3) their ability to spontaneously insert into membranes containing zwitterionic phospholipids with unsaturated acyl chains [37-39]. This unique combination of features makes Cyt toxins highly suitable for the development of membrane-acting ITs, an alternative idea in the field that was initially explored in our laboratory using chemical conjugation strategies [40,41]. Considering that recombinant production methods would provide more homogenous Cyt-based ITs compared to the chemical conjugation strategies, subsequent attempts in our laboratory were based on the use of the E. coli expression system. However, this strategy led to only limited success due to the invariable formation of inclusion bodies in this prokaryotic expression host, which in turn limited the final yields of biologically active Cyt-based ITs. Consequently, we next attempted the recombinant production of Cyt2Aa1-based ITs in P. pastoris using the native bacterial gene. However, as it has been the case for the majority of other bacterial toxins that are also encoded by A+T-rich genes (Table 1), recombinant production of Cyt2Aa1 and Cyt2Aa1-based ITs in P. pastoris necessitated de novo design and construction of a synthetic toxin gene that was optimized for heterologous expression in this yeast [34,35]. Since de novo design and construction of synthetic genes is often a prerequisite for achieving heterologous expression of bacterial toxins in P. pastoris (Table 1), we present general guidelines to this end in the next section based on our experience with the heterologous Cyt2Aa1 expression in this yeast. In contrast to the intracellular expression of the native bacterial gene in P. pastoris, that of the synthetic gene led to the recombinant production of the Cyt2Aa toxin, albeit severe product toxicity effects were observed [34]. Similar toxicity effects were also observed with the intracellular expression of the Cyt2Aa1-based IT in the same heterologous expression host, which could be largely alleviated by the secretory targeting of the recombinant product. While the Cyt2Aa1-based IT failed to be secreted from the P. pastoris cells, secretory targeting proved beneficial in this case since it sequestered the deleterious recombinant product from the yeast cytosol, where a wide range of organelles would otherwise be prone to Cyt2Aa1-based membrane damage [35]. Instead, the recombinant Cyt2Aa1-based IT accumulated to high-levels in the yeast endoplasmic reticulum, where the high local Ca2+ concentration in this organelle is expected to be inhibitory to the basic Cyt2Aa1 toxin activity [35,42]. Furthermore, secretory targeting allowed proper formation of the disulphide bonds requisite for the function of the cell-binding domain of the recombinant Cyt2Aa1-based IT, which could then be recovered in a biologically active form at 10 mg·l-1 culture by a chaotropic denaturation step that was followed with an on-the-column refolding strategy [35]. While the final yield of biologically active Cyt2Aa1-based IT could be potentially increased through the selection of multi-copy integrants of the recombinant expression cassette and/or large-scale bioreactor cultures (Table 1), we did not find this to be necessary for the purposes of our project, which was the development of an in vitro model system to test the potency of Cyt2Aa1-based ITs. However, it has also not escaped our attention that Cyt2Aa1-expressing P. pastoris cells can have further potential use in the control of disease vectors, as has proved to be the case for the last two examples of P. pastoris heterologous expression that we present below. Recombinant expression in P. pastoris of the B. thuringiensis insecticidal Cry2 toxin has also been described using the native bacterial gene [43]. In addition, high-level (up to 30% tcp) P. pastoris co-expression of two biologically active B. sphaericus mosquitocidal proteins BSP1 and BSP2 was reported using synthetic genes that were optimized for heterologous expression in this yeast [44]. Here, P. pastoris cells expressing the B. sphaericus insecticidal proteins were heat-killed without a significant reduction in the biological activity of the recombinant toxins and then fed to Dipteran larvae, which are filter-feeders that usually find yeast cells palatable [44]. This strategy has a minimal risk of releasing the heterologous toxin gene into environment since it would be integrated into the yeast genome unlike the autonomous plasmids used for heterologous expression in E. coli. Design and de novo synthesis of bacterial genes for optimal expression in P. pastoris There are now various commercial services available that offer total gene synthesis at competitive prices. However, it is also possible to design and construct any given DNA sequence using well established protocols [30,34,44-46]. Here we present as an example, the strategy that we have successfully used for the design and de novo construction of a synthetic gene coding for the B. thuringiensis Cyt2Aa1 toxin that was optimized for expression in P. pastoris [34,35]. As discussed previously, our initial attempts at heterologous Cyt2Aa1 expression in P. pastoris were unsuccessful due to inherent problems with the eukaryotic expression of the bacterial gene. This was attributed to the high A+T-content of the native Cyt2Aa1 gene containing cryptic polyadenylation sites that resulted in premature transcription termination in yeast [17,18]. To achieve optimal heterologous expression in P. pastoris, we designed a synthetic gene based on the primary amino acid sequence of the proteinase K-activated form of the Cyt2Aa1 toxin [34]. To this end, the overall A+T-content of the bacterial gene was systematically reduced by changing its codon usage to that preferred by P. pastoris (Table 2). Our manual selection largely favoured the most-preferred P. pastoris codons, but in certain instances the second-most preferred codons were selected instead to ensure an overall reduction in the A+T-content of the resulting DNA sequence. This strategy resulted in the reduction of the A+T-content from ~70% to 50%, while retaining only 18.5% of the original codon usage. Furthermore, our synthetic gene design also ensured that the initial 50–75 nucleotides of the corresponding mRNA would be free of stable secondary structures, especially in the vicinity of the translation initiation codon [47], and the overall DNA sequence would not contain the restriction enzyme sites that would be used during the subsequent cloning strategies, etc. Rational design of the synthetic gene was facilitated by the use of the Genetics Computer Group (GCG) software package (Wisconsin Package version 10.2-UNIX, Madison, WI) [48], especially the programs MFold, PlotFold and Map. A Kozak consensus translation initiation sequence for yeast was also introduced into synthetic gene to ensure its efficient heterologous expression in P. pastoris [49]. Finally, de novo synthesis of the synthetic Cyt2Aa1 gene was readily achieved by a recursive PCR strategy that used overlapping oligonucleotides representing the partial sequence of the sense and anti-sense strands of the proposed DNA sequence [34,35,45]. Briefly, all oligonucleotides were designed to be between 57–71 nucleotides and to have a similar theoretical melting temperature (52–56°C), as well as a 19–23 bp overlap at their 3'-end. To ensure the specificity of each pairing and the absence of any undesirable secondary structures, all oligonucleotide selections were extensively analysed by GCG FastA and Stemloop programs [48,50]. The mutual extension of the overlapping oligonucleotides produces longer double-stranded products, and ultimately the full-length synthetic gene construct, which is then amplified by the 5'-outermost flanking primers [34]. Table 2 P. pastoris codon preference. This codon preference table was compiled from literature and is based on highly expressed genes in P. pastoris, as well as those in other yeast species such as S. cerevisiae [44, 51-53]. Amino acid 1st preference 2nd preference Amino acid 1st preference 2nd preference Ala (A) GCT GCC Leu (L) TTG CTT/CTG Arg (R) AGA CGT Lys (K) AAG AAA§ Asn (N) AAC AAT Met (M) ATG - Asp (D) GAC GAT Phe (F) TTC TTT§ Cys (C)* TGT TGC Pro (P) CCA CCT Gln (Q)† CAA CAG Ser (S)† TCT TCC Glu (E)‡ GAG GAA Thr (T)† ACT ACC Gly (G) GGT GGA Trp (W) TGG - His (H)* CAC CAT Tyr (Y)* TAC TAT Ile (I)† ATT ATC Val (V) GTT GTC †Amino acids for which there is a minimal bias between the first and second-most preferred codons; ‡rare amino acids constituting a major discrepancy between the P. pastoris and S. cerevisiae codon preferences; *amino acids with a very high bias for the first preference codon. Other general trends observed with yeast codon preferences are as follows: (1) §codons that contain 100% G, C, A or T are best avoided, (2) there is a strong avoidance of side-by-side GC base pairs in codon-anticodon interactions, (3) there are three codons used for translational termination, which are used with the frequency TAA > TAG > TGA, and (4) the S. cerevisiae consensus sequence for translation initiation context is A/Y A A/U A AUG UCU (where Y is a pyrimidine base, C or T), however it has been shown to have only a moderate effect on translation [51, 53, 54]. Conclusion P. pastoris is a robust recombinant expression host that has also seemingly emerged as an alternative heterologous expression host for a variety of bacterial toxins and their derivatives. In particular, secretory targeting is an advantageous strategy for the recombinant production of toxins and/or their derivatives that require proteolytic processing and/or proper disulphide bond-formation for their activity. In this respect, P. pastoris may be better suited than the E. coli- and S. cerevisiae-based expression systems and it may also allow higher yields of biologically active recombinant protein as it can be grown to high cell densities under aerobic conditions. As in the case of B. thuringiensis Cyt2Aa1 toxin that is not secreted by the native host, secretory targeting of the fusion proteins may also help alleviate product toxicity effects on the P. pastoris cells. However, undesirable glycosylation of the secreted bacterial toxins may need to be addressed when using this strategy, such as by: (1) introducing silent mutations to remove cryptic glycosylation sites present in the prokaryotic primary amino acid sequence, (2) although it may be cost-prohibitive for large-scale applications, in vitro enzymatic deglycosylation can be carried out, or alternatively, (3) intracellular expression of the toxin can be attempted. A further potential problem that is often encountered during heterologous expression of the bacterial toxins in P. pastoris centers on differences in the codon bias of the A+T-rich prokaryotic toxin genes that can minimize or even preclude the recombinant production of the full-length proteins. However, there are now many examples in the literature on the successful use of de novo synthesized bacterial genes that are optimized for heterologous expression in this yeast. List of abbreviations used AOX: alcohol oxidase; AOX1: P. pastoris major alcohol oxidase gene; tcp: total cell protein; BoNT and TeNT, botulinum and tetanus neurotoxins, respectively; BoNT(HC) and TeNT(HC), the carboxyl-terminal domain of the heavy chain fragment of the botulinum and tetanus neurotoxins, respectively; LT and CT: the heat-labile E. coli enterotoxin and the V. cholerae toxin, respectively; LTB and CTB: B subunit of the heat-labile E. coli enterotoxin and the V. cholerae toxin, respectively; CHO: Chinese hamster ovary; Cef: cell-elongating factor; IT: immunotoxin; DT: diphtheria toxin; DT390: truncated version of DT corresponding to the first 390 amino acid residues. Acknowledgements C.G. is currently sponsored by a Cystic Fibrosis Post-Doctoral Research Fellowship. We thank Drs Atanas V. Koulov and Paul LaPointe for a critical review of this manuscript. ==== Refs Smith LA Development of recombinant vaccines for botulinum neurotoxin Toxicon 1998 36 1539 1548 9792170 10.1016/S0041-0101(98)00146-9 Byrne MP Smith LA Development of vaccines for prevention of botulism Biochimie 2000 82 955 966 11086225 10.1016/S0300-9084(00)01173-1 Smith LA Jensen MJ Montgomery VA Brown DR Ahmed SA Smith TJ Roads from vaccines to therapies Mov Disord 2004 19 S48 52 15027054 10.1002/mds.20009 Harakuni T Sugawa H Komesu A Tadano M Arakawa T Heteropentameric cholera toxin B subunit chimeric molecules genetically fused to a vaccine antigen induce systemic and mucosal immune responses: a potential new strategy to target recombinant vaccine antigens to mucosal immune systems Infect Immun 2005 73 5654 5665 16113283 10.1128/IAI.73.9.5654-5665.2005 FitzGerald DJ Kreitman R Wilson W Squires D Pastan I Recombinant immunotoxins for treating cancer Int J Med Microbiol 2004 293 577 582 15149034 10.1078/1438-4221-00302 Cereghino JL Cregg JM Heterologous protein expression in the methylotrophic yeast Pichia pastoris FEMS Microbiol Rev 2000 24 45 66 10640598 10.1016/S0168-6445(99)00029-7 Macauley-Patrick S Fazenda ML McNeil B Harvey LM Heterologous protein production using the Pichia pastoris expression system Yeast 2005 22 249 270 15704221 10.1002/yea.1208 Couderc R Baratti J Oxidation of methanol by the yeast Pichia pastoris: purification and properties of alcohol oxidase Agric Biol Chem 1980 44 2279 2289 Sears IB O'Connor J Rossanese OW Glick BS A versatile set of vectors for constitutive and regulated gene expression in Pichia pastoris Yeast 1998 14 783 790 9675822 10.1002/(SICI)1097-0061(19980615)14:8<783::AID-YEA272>3.0.CO;2-Y Higgins DR Busser K Comiskey J Whittier PS Purcell TJ Hoeffler JP Small vectors for expression based on dominant drug resistance with direct multicopy selection Methods Mol Biol 1998 103 41 53 9680632 Gurkan C Symeonides SN Ellar DJ High-level production in Pichia pastoris of an anti-p185HER-2 single-chain antibody fragment using an alternative secretion expression vector Biotechnol Appl Biochem 2004 39 115 122 12962542 Cregg JM Barringer KJ Hessler AY Madden KR Pichia pastoris as a host system for transformations Mol Cell Biol 1985 5 3376 3385 3915774 Cregg JM Madden KR Barringer KJ Thill GP Stillman CA Functional characterization of the two alcohol oxidase genes from the yeast Pichia pastoris Mol Cell Biol 1989 9 1316 1323 2657390 Tschopp JF Brust PF Cregg JM Stillman CA Gingeras TR Expression of the lacZ gene from two methanol-regulated promoters in Pichia pastoris Nucleic Acids Res 1987 15 3859 3876 3108861 Grinna LS Tschopp JF Size distribution and general structural features of N-linked oligosaccharides from the methylotrophic yeast, Pichia pastoris Yeast 1989 5 107 115 2711751 10.1002/yea.320050206 Clare J Sreekrishna K Romanos M Expression of tetanus toxin fragment C Methods Mol Biol 1998 103 193 208 9680641 Romanos MA Makoff AJ Fairweather NF Beesley KM Slater DE Rayment FB Payne MM Clare JJ Expression of tetanus toxin fragment C in yeast: gene synthesis is required to eliminate fortuitous polyadenylation sites in AT-rich DNA Nucleic Acids Res 1991 19 1461 1467 2027754 Clare JJ Rayment FB Ballantine SP Sreekrishna K Romanos MA High-level expression of tetanus toxin fragment C in Pichia pastoris strains containing multiple tandem integrations of the gene Biotechnology (N Y) 1991 9 455 460 1367310 10.1038/nbt0591-455 Byrne MP Smith TJ Montgomery VA Smith LA Purification, potency, and efficacy of the botulinum neurotoxin type A binding domain from Pichia pastoris as a recombinant vaccine candidate Infect Immun 1998 66 4817 4822 9746584 Potter KJ Bevins MA Vassilieva EV Chiruvolu VR Smith T Smith LA Meagher MM Production and purification of the heavy-chain fragment C of botulinum neurotoxin, serotype B, expressed in the methylotrophic yeast Pichia pastoris Protein Expr Purif 1998 13 357 365 9693060 10.1006/prep.1998.0910 Byrne MP Titball RW Holley J Smith LA Fermentation, purification, and efficacy of a recombinant vaccine candidate against botulinum neurotoxin type F from Pichia pastoris Protein Expr Purif 2000 18 327 337 10733887 10.1006/prep.2000.1200 Potter KJ Zhang W Smith LA Meagher MM Production and purification of the heavy chain fragment C of botulinum neurotoxin, serotype A, expressed in the methylotrophic yeast Pichia pastoris Protein Expr Purif 2000 19 393 402 10910730 10.1006/prep.2000.1256 Zhang W Bevins MA Plantz BA Smith LA Meagher MM Modeling Pichia pastoris growth on methanol and optimizing the production of a recombinant protein, the heavy-chain fragment C of botulinum neurotoxin, serotype A Biotechnol Bioeng 2000 70 1 8 10940857 10.1002/1097-0290(20001005)70:1<1::AID-BIT1>3.0.CO;2-Y Weatherly GT Bouvier A Lydiard DD Chapline J Henderson I Schrimsher JL Shepard SR Initial purification of recombinant botulinum neurotoxin fragments for pharmaceutical production using hydrophobic charge induction chromatography J Chromatogr A 2002 952 99 110 12064550 10.1016/S0021-9673(02)00074-2 Zhang W Smith LA Plantz BA Schlegel VL Meagher MM Design of methanol Feed control in Pichia pastoris fermentations based upon a growth model Biotechnol Prog 2002 18 1392 1399 12467476 10.1021/bp025516w Johnson SK Zhang W Smith LA Hywood-Potter KJ Todd Swanson S Schlegel VL Meagher MM Scale-up of the fermentation and purification of the recombinant heavy chain fragment C of botulinum neurotoxin serotype F, expressed in Pichia pastoris Protein Expr Purif 2003 32 1 9 14680933 10.1016/j.pep.2003.07.003 Fingerut E Gutter B Meir R Eliahoo D Pitcovski J Vaccine and adjuvant activity of recombinant subunit B of E. coli enterotoxin produced in yeast Vaccine 2005 23 4685 4696 15951067 10.1016/j.vaccine.2005.03.050 Trucksis M Conn TL Fasano A Kaper JB Production of Vibrio cholerae accessory cholera enterotoxin (Ace) in the yeast Pichia pastoris Infect Immun 1997 65 4984 4988 9393786 McCardell BA Sathyamoorthy V Michalski J Lavu S Kothary M Livezey J Kaper JB Hall R Cloning, expression and characterization of the CHO cell elongating factor (Cef) from Vibrio cholerae O1 Microb Pathog 2002 32 165 172 12079406 10.1006/mpat.2001.0492 Woo JH Liu YY Mathias A Stavrou S Wang Z Thompson J Neville DM Jr Gene optimization is necessary to express a bivalent anti-human anti-T cell immunotoxin in Pichia pastoris Protein Expr Purif 2002 25 270 282 12135560 10.1016/S1046-5928(02)00009-8 Woo JH Liu YY Stavrou S Neville DM Jr Increasing secretion of a bivalent anti-T-cell immunotoxin by Pichia pastoris Appl Environ Microbiol 2004 70 3370 3376 15184133 10.1128/AEM.70.6.3370-3376.2004 Woo JH Liu YY Neville DM Jr Minimization of aggregation of secreted bivalent anti-human T cell immunotoxin in Pichia pastoris bioreactor culture by optimizing culture conditions for protein secretion J Biotechnol 2005 (doi:10.1016/j.jbiotec.2005.1007.1004) Liu YY Woo JH Neville DM Targeted introduction of a diphtheria toxin resistant mutation into the chromosomal EF-2 locus of Pichia pastoris and expression of immunotoxin in the EF-2 mutants Protein Expr Purif 2003 30 262 274 12880776 10.1016/S1046-5928(03)00129-3 Gurkan C Ellar DJ Expression of the Bacillus thuringiensis Cyt2Aa1 toxin in Pichia pastoris using a synthetic gene construct Biotechnol Appl Biochem 2003 38 25 33 12628007 10.1042/BA20030017 Gurkan C Ellar DJ Expression in Pichia pastoris and purification of a membrane-acting immunotoxin based on a synthetic gene coding for the Bacillus thuringiensis Cyt2Aa1 toxin Protein Expr Purif 2003 29 103 116 12729731 10.1016/S1046-5928(03)00012-3 Schnepf E Crickmore N Van Rie J Lereclus D Baum J Feitelson J Zeigler DR Dean DH Bacillus thuringiensis and its pesticidal crystal proteins Microbiol Mol Biol Rev 1998 62 775 806 9729609 Thomas WE Ellar DJ Bacillus thuringiensis var israelensis crystal δ-endotoxin: effects on insect and mammalian cells in vitro and in vivo J Cell Sci 1983 60 181 197 6874728 Thomas WE Ellar DJ Mechanism of action of Bacillus thuringiensis var israelensis insecticidal δ-endotoxin FEBS Lett 1983 154 362 368 6832375 10.1016/0014-5793(83)80183-5 Du J Knowles BH Li J Ellar DJ Biochemical characterization of Bacillus thuringiensis cytolytic toxins in association with a phospholipid bilayer Biochem J 1999 338 185 193 9931315 10.1042/0264-6021:3380185 Drobniewski FA Immunotoxins up to the present day Biosci Rep 1989 9 139 156 2669993 10.1007/BF01115993 al-yahyaee SA Ellar DJ Cell targeting of a pore-forming toxin, CytA δ-endotoxin from Bacillus thuringiensis subspecies israelensis, by conjugating CytA with anti-Thy 1 monoclonal antibodies and insulin Bioconjug Chem 1996 7 451 460 8853459 10.1021/bc960030k Knowles B Blatt M Tester M Horsnell J Carroll J Menestrina G Ellar D A cytolytic delta-endotoxin from Bacillus thuringiensis var. israelensis forms cation-selective channels in planar lipid bilayers FEBS Lett 1989 244 259 262 2465921 10.1016/0014-5793(89)80540-X Ogunjimi AA Chandler JM Gbenle GO Olukoya DK Akinrimisi EO Heterologous expression of cry2 gene from a local strain of Bacillus thuringiensis isolated in Nigeria Biotechnol Appl Biochem 2002 36 241 246 12452809 10.1042/BA20020053 Sreekrishna K Prevatt WD Thill GP Davis GR Koutz P Barr KA Hopkins SA Production of Bacillus entomotoxins in methylotrophic yeast US Patent 1998 5,827,684 Prodromou C Pearl LH Recursive PCR: a novel technique for total gene synthesis Protein Eng 1992 5 827 829 1287665 Withers-Martinez C Carpenter EP Hackett F Ely B Sajid M Grainger M Blackman MJ PCR-based gene synthesis as an efficient approach for expression of the A+T-rich malaria genome Protein Eng 1999 12 1113 1120 10611405 10.1093/protein/12.12.1113 Baim SB Sherman F mRNA structures influencing translation in the yeast Saccharomyces cerevisiae Mol Cell Biol 1988 8 1591 1601 2837649 Womble DD GCG: The Wisconsin Package of sequence analysis programs Methods Mol Biol 2000 132 3 22 10547828 Kozak M Initiation of translation in prokaryotes and eukaryotes Gene 1999 234 187 208 10395892 10.1016/S0378-1119(99)00210-3 Pearson WR Lipman DJ Improved tools for biological sequence comparison Proc Natl Acad Sci USA 1988 85 2444 2448 3162770 Bennetzen JL Hall BD Codon selection in yeast J Biol Chem 1982 257 3026 3031 7037777 Ernst JF Codon usage and gene expression Trends Biotechnol 1988 6 196 199 10.1016/0167-7799(88)90046-7 Ikemura T Correlation between the abundance of yeast transfer RNAs and the occurrence of the respective codons in protein genes. Differences in synonymous codon choice patterns of yeast and Escherichia coli with reference to the abundance of isoaccepting transfer RNAs J Mol Biol 1982 158 573 597 6750137 10.1016/0022-2836(82)90250-9 Cigan AM Donahue TF Sequence and structural features associated with translational initiator regions in yeast – a review Gene 1987 59 1 18 3325335 10.1016/0378-1119(87)90261-7	16336647	PMC1325036	CC BY	2021-01-04 16:24:35	no	Microb Cell Fact. 2005 Dec 7; 4:33	utf-8	Microb Cell Fact	2,005	10.1186/1475-2859-4-33	oa_comm
==== Front CytojournalCytoJournal1742-6413BioMed Central London 1742-6413-2-211633667210.1186/1742-6413-2-21ResearchFine needle aspiration cytology of primary thyroid lymphoma: a report of ten cases Gupta Nalini [email protected] Raje [email protected] Radhika [email protected] Arvind [email protected] Pinaki [email protected] Anil [email protected] SC [email protected] Department of Cytopathology, Postgraduate Institute of Medical Education and Research, Chandigarh, 160012, India2 Department of Endocrinology, Postgraduate Institute of Medical Education and Research, Chandigarh, 160012, India3 Department of Radiotherapy, Postgraduate Institute of Medical Education and Research, Chandigarh, 160012, India2005 9 12 2005 2 21 21 2 8 2005 9 12 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. Primary lymphoma is an uncommon malignancy of the thyroid, comprising of 0.6 to 5 per cent of thyroid cancers in most series. Primary thyroid lymphomas (PTL) occur most commonly in elderly women and are commonly of B- cell origin. These frequently present in clinical stage IE and IIE. We report here ten cases of PTL diagnosed over a period of about 7 years in our institute. Out of these ten cases, nine were diagnosed on fine needle aspiration cytology (FNAC) and one case was misdiagnosed as lymphocytic thyroiditis. This case was diagnosed as Non- Hodgkin's lymphoma on surgical specimen. Five patients are disease free and doing well, while two died of disease and the other two were lost to follow-up. One patient is currently on chemotherapy. The salient clinical, biochemical, radiological features, FNA findings along with diagnostic difficulties are discussed. Primary thyroid lymphomaFine needle aspiration cytologyPrognosis. ==== Body Introduction Primary thyroid lymphoma is a rare disease that continues to produce diagnostic and therapeutic dilemmas. Most thyroid lymphomas are of B-cell origin [1]. There appear to be two distinct clinical and prognostic groups of these rare tumors. The more common subtype, comprising of up to 70% of cases, is a diffuse large B-cell lymphoma [1]. This subtype appears to have the most aggressive clinical course with almost 60% of these tumors diagnosed with disseminated disease. The other subtype is mucosa-associated lymphoid tissue (MALT) lymphomas comprising of approximately 6% to 27% of thyroid lymphomas [1]. These have a relatively indolent course. These occur more commonly in females with female to male ratio of 2–4:1. The majority of these patients have underlying Hashimoto's thyroiditis, which increases the risk for thyroid lymphoma by 50 times [2]. The importance of recognizing primary thyroid lymphoma lies in the fact that this disease is quite curable without the need for extensive surgery if recognized early and treated appropriately. In this report, we discuss FNA findings of ten such patients. Materials and methods Ten cases of primary thyroid non-Hodgkin's lymphoma (PTL) were retrieved from files of department of Cytopathology, Postgraduate Institute of Medical Education and Research, Chandigarh, India over a seven year period (Jan 1998- Oct 2004), during this period about 4500 patients were aspirated for thyroid enlargement. Clinical information including age, sex, presenting symptoms, treatment and subsequent course were recorded. FNAC was performed with 23 G needle and on an average 2–3 passes were taken to obtain adequate material for diagnosis. Histopathology of the surgically resected specimens was available in three cases. Immunocytochemistry (ICC) was performed for leukocyte common antigen (LCA), cytokeratin, CD 20 and CD3 in six cases. Results Of the 10 cases, 5 were female and 5 were males with female to male ratio of 1: 1. The age ranged from 10–61 years with a mean age of 39.6 years. Two patients were less than 15 years of age. Duration of follow-up ranged from 1 month to 4.3 years. A summary of clinical information and cytological findings is shown in Table 1. All patients presented with thyroid enlargement of variable duration (20 days to 2 years). Seven patients presented with a diffuse goiter, two patients had a multinodular goiter and only one patient presented as a solitary thyroid nodule. Other signs and symptoms included stridor, hoarseness, dyspnoea, dysphagia and sense of heaviness in the neck. Four patients had associated cervical lymphadenopathy and three of them had lymphomatous involvement while one had reactive hyperplasia. Table 1 showing clinical presentation, biochemical investigations and follow- up of ten cases of primary thyroid lymphoma. No Age/ Sex Clinical Presentation Biochemistry Anti thyroid antibody Diagnosis Follow- up Prognosis 1 61 F Nodular goiter- 2 months Hypothyroid TMA + NHL – HG 3 yrs Disease free 2 48 F Diffuse goiter- 20 months Hypothyroid TMA + FNA- LT Surgery- NHL Lost to follow- up 3 44 M Diffuse goiter- 24 months Euthyroid TMA + NHL – IG 4.3 years Disease free 4 60 F Diffuse goiter- 1 month, local lymph nodes Euthyroid TMA + NHL – IG 4 years Local nodal recurrence after 1 1/2 years- RT given, Stable. 5 13 M Diffuse goiter- 6 months Euthyroid TMA + NHL – HG 3 years Disease free 6 10 M Diffuse goiter, stridor- 2 months, local lymph nodes Euthyroid TMA -ve NHL – HG 6 months Lost to follow- up, Came after 6 months with SVC syndrome, Died. 7 60 M Solitary nodule- 12 months Hypothyroid TMA + NHL – IG 6 months Disease free 8 20 M Diffuse goiter, Hoarse voice, stridor- 6 months, local lymph nodes Euthyroid TMA -ve NHL – HG 1 year Died of disseminated disease 9 32 F Nodular goiter, local lymph nodes- 5 months Euthyroid TMA -ve NHL – HG 1 month On treatment- chemotherapy 10 48 F Diffuse goiter, stridor- 20 days Euthyroid TMA -ve NHL – HG Lost to follow- up Imaging modalities and bone marrow examination did not reveal other areas of involvement. CT scan done in these cases revealed a lesion primarily involving the thyroid gland (Figure 1). Hematological parameters, serum biochemistry and thyroid function tests were normal at presentation except in three patients, who were hypothyroid. Thyroid microsomal antibody (TMA) titer was significantly increased (1:80) in six patients. Figure 1 CECT scan showing a homogenous isodense soft tissue mass (3.2 × 4.5 cm) in the region of left lobe of thyroid. Cytologically in nine cases, a diagnosis of non- Hodgkin's lymphoma (NHL) was offered on FNAC. These included six cases of high grade NHL and three cases of intermediate grade NHL. The smears in high grade NHL cases were cellular and comprised of monomorphic population of large atypical lymphoid cells. These cells were 2–3 times the size of a mature lymphocyte with opened up chromatin and conspicuous 1–2 nucleoli (Figure 2). In the background many lymphoglandular bodies were seen. One of these six cases had subsequent histological confirmation. Figure 2 Aspiration smear showing monomorphic population of atypical lymphoid cells in a case of high grade non- Hodgkin's lymphoma (MGG × 512). In three cases, FNAC smears showed a dual population of lymphoid cells comprising of an admixture of mature lymphocytes and larger lymphoid cells. These large lymphoid cells were 2–3 times the size of a mature lymphocyte and had opened up chromatin and conspicuous nucleoli (Figure 3). Many monocytoid cells were also seen. A diagnosis of florid lymphocytic thyroiditis was considered less likely, because, in thyroiditis, a reactive population of lymphoid cells of variable sizes admixed with plasma cells and tingible body macrophages is usually present. Hence, in these three cases, combining the clinical presentation and cytological findings, a diagnosis of NHL – intermediate grade was suggested. Figure 3 Aspiration smear showing a typical dual population of lymphoid cells in a case of intermediate grade non- Hodgkin's lymphoma (H&E × 512). In one case, the diagnosis of lymphoma was missed on FNA. The patient presented with a diffuse rather than a nodular swelling and cytology revealed a polymorphic infiltrate consistent with a reactive hyperplasia which was seen to infiltrate the follicular epithelial cell clusters. Hence a diagnosis of lymphocytic thyroiditis was offered. A subtotal thyroidectomy was performed because of associated pressure symptoms. Histopathology revealed a Non- Hodgkin's lymphoma morphologically and immunohistochemically consistent with a marginal zone B cell lymphoma. In six cases, immunocytochemistry (ICC) was performed on the fine needle aspirate and all cases were CD 20 positive and CD3 negative. In all the cases, protocol staging as per the ANN Arbor staging of NHL was performed. Bone marrow examination, whole body 67Gallium scan, CECT abdomen and chest and CSF for malignant cytology was done in all patients before labeling a case as primary thyroid lymphoma. In the present series, 7 cases were of stage I (E) and three cases were of stage II (E). Lactate dehydrogenase (LDH) levels were done in six out of ten patients and the mean LDH level was 628 U/L. Follow- up information All the patients received combination chemotherapy (CHOP regime) with local radiotherapy. Five patients are alive and are free of disease till date, whereas, two patients died of the disease. These two patients had high grade lymphoma and succumbed to disseminated disease. Two were lost to follow-up and one patient is currently on chemotherapy. Discussion Primary thyroid lymphoma (PTL) is defined as a lymphomatous process involving the thyroid gland without contiguous spread or distant metastases from other areas of involvement at diagnosis [3]. PTL constitutes 5% of all thyroid malignancies and occurs in less than 1% of all non- Hodgkin's lymphomas [3]. The majority of patients are middle to old aged women [4]. Studies have shown that PTL typically arises in the setting of autoimmune thyroiditis and it takes, on an average, 20 to 30 years to develop after the onset of lymphocytic thyroiditis [5]. A short history of a rapidly enlarging neck mass often associated with dyspnoea, difficulty in swallowing, or voice change is the hallmark presentation of thyroid lymphoma [6]. Therefore, clinically this may be confused with anaplastic thyroid carcinoma. Hoarseness, respiratory distress, cough and dysphagia were the usual presenting manifestations in our patients. Hypothyroidism at the time of diagnosis is documented in 30–40% of patients due to replacement of thyroid parenchyma by the lymphomatous process or due to underlying Hashimoto's thyroiditis. Thyrotoxicosis is exceedingly rare [7]. Seven of our patients at presentation were euthyroid and three were hypothyroid. Circulating antibodies to thyroid peroxidase are positive in 60–80% of patients suggesting underlying lymphocytic thyroiditis as a predisposing factor. In the present series, six patients had significantly elevated thyroid microsomal antibody titer. Fine needle aspiration has become the procedure of choice for the initial pathological diagnosis of thyroid nodule. However, studies have also shown inconsistent results in the diagnosis of lymphoma of the thyroid. In one series, a correct diagnosis with FNAC was made in 70–80% of patients with thyroid lymphoma [8], but in others, FNA was suggestive but not diagnostic in only 50–60% of patients [9-12]. In the present study, nine out of ten cases (90%) of the cases of PTL were correctly diagnosed by FNA. A primary thyroid non-Hodgkin lymphoma is usually of large cell type [13] and a diagnosis of large cell lymphoma is easy on FNA and features like lack of cellular cohesion and presence of lymphoglandular bodies in the background are features strongly against a diagnosis of anaplastic carcinoma [14]. ICC confirms the lymphoid origin of the cells and their B or T- lineage. By contrast, cytological diagnosis of MALT- lymphomas is difficult, because of heterogeneous appearance of the neoplastic infiltrate [14]. The principal problem with a cytological diagnosis of low- grade lymphoma of the thyroid is its differentiation from HT. The distinguishing features may be the abundance of lymphoid tissue and a high proportion of intermediate centrocyte- like cells in low- grade NHL as compared to HT. False negative results may be due to sampling error also as low- grade B- cell MALT lymphoma originates from HT and the two usually coexist [14]. Due attention to dual population of lymphoid cells, presence of monocytoid cells in FNA smears and extensive follicular epithelial destruction and the clinical setting enabled us to diagnose three cases of non- Hodgkin's lymphoma of intermediate grade. However, we missed the diagnosis of NHL on FNAC in one case reported as lymphocytic thyroiditis, which on subsequent subtotal thyroidectomy, was reported as non- Hodgkin's lymphoma. FNAC cytomorphology in conjunction with flow cytometric (FC) immunophenotyping has become a reliable and accurate method for the diagnosis and classification of many lymphoproliferative disorders. CD4/CD8 T-cell ratio comparisons are made with cytomorphological diagnoses of reactive, atypical, non-Hodgkin lymphoma, and Hodgkin lymphoma cases [15,16]. PTL frequently present in clinical stage IE and IIE. Treatment is similar to other nodal lymphoma. For patients with intermediate or high-grade lymphoma, the best results are obtained from cyclophosphamide, daunorubicin, vincrstine and prednisolone (CHOP) based chemotherapy. Radiation therapy is used most commonly after 3–6 courses of chemotherapy in form of modified mantle irradiation including thyroid, bilateral neck, supraclavicular area and mediastinum [17]. Our patients received CHOP based chemotherapy, two of our patients had relapsed with bone marrow involvement and local nodal recurrence. One died of lymphomatous process and other was controlled with radiotherapy. The poor prognostic factors include age more than 60 years, performance status grater than 1, elevated lactate dehydrogenase (LDH) and β2 microglobulin, extranodal sites more than 1 and Ann -Arbor staged III-IV [18,19]. In summary, we report a retrospective study of ten cases of primary thyroid lymphomas. The diagnosis was established by fine needle aspiration in nine cases and one case was misdiagnosed as lymphocytic thyroiditis, which was diagnosed on surgical specimen. The cytological diagnosis of high grade lymphoma is easy and ICC can confirm suspicious cases. The diagnosis of low grade lymphoma is more difficult but clinical and radiological suspicion and cytomorphological features can help reaching the correct diagnosis in such cases. Abbreviations NHL – Non- Hodgkin's lymphoma HG- High grade IG- Intermediate grade TMA- Thyroid microsomal antibody HT- Hashimoto's thyroiditis LT- Lymphocytic thyroiditis RT- Radiotherapy SVC- Superior vena cava. Acknowledgements Co-editors of CytoJournal Vinod B. Shidham, MD, FRCPath, FIAC and Barbara F. Atkinson, MD thank: the academic editor Sherif Ibrahim MD, PhD, Room 383, 560 First Ave. Department of Pathology, New York University Medical Center, New York, NY 10016 [E-mail:[email protected]] for organizing and completing the peer-review process for this manuscript." ==== Refs Widder S Pasieka JL Primary thyroid lymphoma Curr Treat Options Oncol 2004 5 307 14 15233907 Holm LE Blomgren H Lowhagen T Cancer risks in patients with chronic lymphocytic thyroiditis N Engl J Med 1985 312 601 4 3838363 Ansell SM Grant CS Habermann TM Primary thyroid lymphoma Semin Oncol 1999 26 316 23 10375088 Derringer GA Thompson LDR Frommelt RA Bijwaard KE Heffess CS Abbondanzo SL Malignant lymphoma of the thyroid gland: a clinicopathologic study of 108 cases Am J Surg Path 2000 24 623 39 10800981 10.1097/00000478-200005000-00001 Pedersen RK Pederson NT Primary non- Hodgkin's lymphoma of the thyroid gland: a population based study Histopathology 1996 28 25 32 8838117 10.1046/j.1365-2559.1996.268311.x Singer JA Primary lymphoma of the thyroid Am Surg 1998 64 334 7 9544144 Campagno J Oertel JE Malignant lymphomas and other lymphoproliferative disorders of the thyroid gland. A clinicopathological study of 245 cases Am J Clin Pathol 1980 74 1 11 7395811 Matsuzuka F Miyauchi A Katayama S Clinical aspects of primary thyroid lymphoma: Diagnosis and treatment based on experience of 119 cases Thyroid 1993 3 93 9 8369658 Skarsgard ED Connors JM Robins RE A current analysis of primary lymphoma of the thyroid Arch Surg 1991 126 1199 1204 1929820 Das DK Gupta SK Francis IM Fine needle aspiration cytology diagnosis of non- Hodgkin's lymphoma of thyroid: A report of four cases Diagn Cytopathol 1993 9 639 45 8143536 Klyachkin ML Schwartz RW Cibull M Thyroid lymphoma: Is there a role for surgery? Am Surg 1998 64 234 8 9520813 Matsuda M Sone H Koyama H Fine- needle aspiration cytology of malignant lymphoma of the thyroid Diagn Cytopathol 1987 3 244 9 3665692 Nguyen GK Lee MW Ginsberg J Wragg T Bilodeau D Fine-needle aspiration of the thyroid: an overview Cytojournal 2005 2 12 15987502 10.1186/1742-6413-2-12 Sangalli G Serio G Zampatti C Lomuscio G Colombo L Fine needle aspiration cytology of primary lymphoma of the thyroid: a report of 17 cases Cytopathology 2001 12 257 63 11488875 10.1046/j.1365-2303.2001.00338.x 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? A study of 85 lymph node FNAs with increased CD4/CD8 ratio Cytojournal 2005 2 14 16153296 10.1186/1742-6413-2-14 Beaty MW Geisinger KR Commentary: Hodgkin lymphoma: flow me? Cytojournal 2005 2 13 16150141 10.1186/1742-6413-2-13 Miller TP Dahlberg S Cassady JR Chemotherapy alone compared with chemotherapy plus radiotherapy for localized high-grade non-Hodgkin's lymphoma N Engl J Med 1998 339 21 6 9647875 10.1056/NEJM199807023390104 Swan F JrVelasquez WS Tucker S A new serological staging system for large cell lymphomas based on initial β-2 microglobulin, lactate dehydrogenase levels J Clin Oncol 1989 7 1518 27 2674337 Tupchong L Hughes F Hormer CL Primary lymphoma of the thyroid: Clinical features, prognostic factors and result Radiat oncol Biol Phys 1986 12 1813 21	16336672	PMC1325037	CC BY	2021-01-04 16:24:26	no	Cytojournal. 2005 Dec 9; 2:21	utf-8	Cytojournal	2,005	10.1186/1742-6413-2-21	oa_comm
==== Front BMC Ear Nose Throat DisordBMC Ear, Nose and Throat Disorders1472-6815BioMed Central London 1472-6815-5-121633668910.1186/1472-6815-5-12Study ProtocolCoronary artery bypass grafting and sensorineural hearing loss, a cohort study Ashraf Omer [email protected] Aga Khan University, Stadium Road, Karachi, 74800, Pakistan2005 10 12 2005 5 12 12 5 3 2005 10 12 2005 Copyright © 2005 Ashraf; 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 Sudden sensorineural hearing loss is routinely encountered by the otologist. The etiology is varied and often identifiable. One of the relatively less frequent causes is surgery. Apart from being an established entity with otological surgeries, sensorineural hearing loss has also been known to occur after non-otological procedures under general anesthesia. Commonest amongst these procedures is cardiopulmonary bypass, an association that has long been recognized. However, despite the proposition of diverse hypotheses in the past, the pathophysiology remains unclear. Methods The study is a prospective matched cohort study that will be carried out in Aga Khan University Hospital, Karachi, Pakistan. Participants among exposed would include all those patients who would be undergoing coronary artery bypass surgery in the hospital who fall under the criteria for inclusion. Unexposed group would comprise of patients undergoing a non-bypass procedure of similar duration under the same type of anesthesia who meet the selection criteria. Both these groups will undergo audiometric testing at our hospital on three different occasions during the course of this study. Initially before the procedure to test the baseline hearing capacity; then one week after the procedure to assess any changes in hearing ability following the surgery; and finally a third audiogram at six weeks follow-up to assess further changes in any hearing deficits noted during the second phase of testing. Certain variables including the subjects' demographics and those concerning the procedure itself will be noted and used later for risk factors analysis. A detailed past medical and surgical history will also be obtained. Data analysis would include calculation of relative risk and significance of the results, by running the chi-square test. Other statistical tests like Fisher exact test may then be employed to facilitate data interpretation. Continuous scale may then be employed and multivariate linear regression used. Discussion This study is planned to obtain a better understanding of the correlation between sudden sensorineural hearing loss and cardiopulmonary bypass. Being the first major cohort trial in this line of investigation, the project is designed to identify the existence of any significant relationship between cardiopulmonary bypass and sensorineural hearing deficit. ==== Body Background Hearing is one of the most significant of human senses. For an individual it is perhaps the most crucial link in communication with the outside world. It's a prerequisite for a fulfilling life and career. Any deficit in it is as frightening for the person as a loss of vision. It's a handicap potentially severe enough to deny a person a normal life and livelihood. Sensorineural hearing loss (SNHL) is often encountered by the ENT surgeon. It is established globally as a significant morbidity. Even more traumatizing is the development of a sudden sensorineural hearing loss by a patient, a disease afflicting 1 in every 5000 patients in the general population [1]. Various etiologies are proposed to explain the sudden SNHL in general population and include inflammatory (bacterial, viral and autoimmune), vascular (hemorrhage, microemboli, thrombosis), metabolic (hormonal, chemical, drugs), traumatic (direct, noise, surgical), membrane rupture, functional, or idiopathic. In many of the unidentifiable cases there is still debate between a viral versus vascular etiology. Sensorineural hearing loss can also be explained as a surgical complication of otologic surgeries. However, what has raised the great interest of medical professionals around the world in recent decades is the occasional development of sudden sensorineural hearing loss with non-otologic surgeries carried out under general anesthesia. A considerable majority of these surgeries include cardiopulmonary bypass procedures. The first report of this kind was made by Arenberg (1972) [2] who reported sudden unilateral deafness immediately following cardiopulmonary bypass. Earlier Brownson et al (1971) [3] had tried unsuccessfully to achieve the same in a prospective study with a limited sample. This was followed by a larger retrospective study by Plasse et al [4,5] in which he evaluated 7000 patients and found the incidence of SNHL with aortocoronary bypass surgery to be 1 in 1000 (0.1%). However, these results are highly questionable considering the presence of false-negatives and false-positives associated with retrospective studies. It has been established since that SNHL can be objectively determined only in routine and prospective evaluation. There was controversy in the past in the demarking of criteria for SNHL. This was evident in the study by Shapiro et al [6] who applied overly inclusive criteria and found the hearing loss to be 13.2% in a prospective study involving 68 patients. Doubts were always raised though, regarding the inclusion in his study of the patients who reported a hearing deficit of up to 10 dB, which can be owing to the potential variations and audiometric testing errors due to patient concentration and co-operation during the testing. This was followed by two case reports from Millen et al [7] and a prospective study by Ness et al [8]. A review of the literature indicates several distinct variants. First, it is found that the SNHL in patients undergoing coronary artery bypass grafting (CABG) occurs in an older age group with known preexisting cardiovascular disease [1]. Second, there is a clear male to female preponderance in the incidence of these cases that cannot merely be explained by the greater number of male patients undergoing the procedure [4]. It has also been found that there is little or no recovery from the insult with a partial recovery of hearing occurring in less than 50% of patients [1]. Another point is the absence of vertigo in most of these cases [4,5]. Tinnitus may also be found [6]. The magnitude of the problem has long been recognized considering the incidence rates that are reported up to 10–15% [6], and the fact that even in 1980 over 100,000 bypass operations were carried out in the United States [9]. Various etiologies had thus been proposed to look into the pathophysiology of this phenomenon and thus offer a possible preventive or curative measure. Microembolic phenomena (fat, air or particulate thrombi)[2], perioperative hypotension or perfusion failure [6], hypercoagulable states [10], and ototoxic drug usage [6] are some of the better recognized ones. However, many of the mechanisms involving CABG (e.g. fat emboli, anti-foaming agents) do not apply to those other non-otologic surgeries, which have reported sudden idiopathic SNHL. Since all of the reported cases have occurred under general or spinal anesthesia, the hemodynamic fluctuations during induction and maintenance of anesthesia common to both of these could be a factor [11]. Cochlear membrane leaks and perilymphatic fistulae have been invoked as possible mechanisms for loss of hearing. Both implosive and explosive mechanisms have been implicated [12]. Nitrous oxide is often blamed as it can generate very high pressures in the middle ear [13,14]. However, there have been described cases in which nitrous oxide was not used, and its presence in many cases may only reflect its very widespread anesthetic use. It is also important to exclude a brainstem event causing bilateral hearing loss [15,16], though in these reports other neurological signs like ataxia, dysarthria, gaze palsies and multiple cranial nerve and neurological deficits have been mentioned alongside. Methods Study design This is a prospective matched cohort study. Setting Aga Khan University Hospital, Karachi, Pakistan. Inclusion criteria for exposed group All patients undergoing CABG who give consent and can comply with the testing. They will be taken by consecutive sampling. Inclusion criteria for unexposed group All patients who qualify for the criteria of being age and sex matched with the exposed (on an individual basis) undergoing a non-cardiac and non-otologic surgery under general anaesthesia, of similar time duration, who give consent and comply with the testing. There is no report in literature of association of SNHL with any specific non-otologic, non-cardiac surgery, apart from certain neurosurgical procedures, especially around the brainstem and/or the VII-VIIIth cranial nerve complex. So any surgery apart from the above mentioned, which employs the same type of anaesthesia and is of similar time duration (the two pertinent factors) can be taken as control to generate the sample. We have decided to use the patients undergoing surgeries for head and neck tumors, not having any otologic, brain stem or the VII-VIIIth cranial nerve complex involvement, as members of the unexposed group. The patient age group is likely to be similar and the procedure will last a similar time period. They are being taken on a basis of one unexposed per exposed. Age, within a five-year period, and sex matching will be done, for the selection of unexposed. Exclusion criteria 1. Those who don't give consent. 2. Those who can't comply with the testing. 3. Those who have a previous history of ear surgery. Sample size calculation As regards sample size calculation, there are two widely varying outcomes of hearing loss, in literature, 0.1% and 13.2%. Both studies had their intrinsic flaws, the former being a retrospective study, and the latter using overly inclusive criteria. Taking the highest reported incidence of deafness to be 13.2 % with a 95% level of significance, 80% power, the sample size comes out to be 138 (69 each group). The expected frequency of disease in the unexposed group is taken as equivalent of that in the general population which is 0.2%. However, to further increase the power of the study, the best possible approach (that is expected to generate a manageable sample size) is to take the frequency of disease as 6.55% (between the outcomes of the two above mentioned studies), then the sample size, with a 95% level of significance and 80% power, is 290 (145 per group), which is manageable. This estimate is both manageable and appropriate considering the intrinsic errors in both the quoted studies. Data collection tool Three audiograms will be done for objective assessment of patients and one questionnaire will be filled from interviewing of the patients, for subjective analyses. Certain perioperative data will be noted. The first audiogram will be done pre-operatively to assess the baseline level of a subject's hearing. The second audiogram will be done after the procedure, after one week, prior to the discharge of patients. This will assess the alteration in hearing, if any, of the subject. The third and final audiogram will be done at six weeks follow up. This will aid in providing information about any late deterioration or recovery. The audiogram will assess auditory function of both ears obtained via a complete audiometric evaluation in a soundproof room in the ENT clinic of our hospital, including pure-tone air and bone conduction levels, speech reception thresholds, and speech discrimination score testing. The questionnaire will make note of patient demographics and various questions that comprise an otologic and past medical and surgical history asked to determine if the patient had any prior hearing loss, ear disease, ear surgery, known ototoxic medication usage, tinnitus, vertigo, family history of hearing loss, noise exposure, neurological disease, diabetes, hyperlipidemia, hypertension and hypotension. After the procedure it will be inquired if the patient had the subjective complaint of hearing loss, tinnitus, vertigo, imbalance, any ear pain or discharge. Similar questions will be asked at the six week follow up. Perioperative data collected on each patient will include the type of surgery (CABG simply or CABG with some other surgery like valve replacement), time on perfusion pump, the type of pump used, the type of filter used, aortic cross clamp time, blood loss, fluids, maximum and minimum systolic and diastolic pressures during surgery, pulse, oxygen saturation, any straining, emesis or increased venous pressure from over ventilation, extension of neck for long period during the procedure, postoperative complications (e.g. arrhythmias, prolonged hypotension, hypertension and neurological changes), and medication usage. Also noted will be the components of anesthesia, vitals, blood loss, fluids and medications in the recovery room. Any other medical ailment or surgical procedure during the six-week post-operative period will also be noted in detail. Data analysis Data analysis will be carried out by entering the data into the Microsoft Windows based Statistical Package for the Social Sciences (SPSS- released 10.1, standard version, copyright SPSS; 1989–1999). Anonymity of the subjects will be maintained. Relative risk will be calculated. Significance of the results will be obtained by running the chi-square test and later further interpretations carried out by using the Fisher exact test. Change in hearing acuity may be calculated on a continuous scale (as by calculating the mean hearing change in the two groups, and comparing them between the exposed and the unexposed group). Data analysis will then be carried out, in case of usage of continuous variables, by multivariate linear regression. Discussion What is the exact cause, what then is the pathophysiology and how can we apply better curative and preventive measures to eliminate any further incidence of the sudden SNHL from not just CABG procedures but other non-otologic procedures as well, is a source of some concern to surgeons around the planet and the solution lies in evidence based medicine in this direction. The rationale of this study is not just the need for better understanding of the issue but also to reduce the psychological trauma and morbidity encountered by patients, and to help in increasing patient confidence in life-saving procedures as CABG. Abbreviations Sensorineural hearing loss: SNHL Ear, nose and throat: ENT Coronary artery bypass grafting: CABG Competing interests The author(s) declare that they have no competing interests. Authors' contributions Being the sole author, Omer Ashraf was involved in design of study, drafting and revision of the article and in final approval of the manuscript version to be published. Pre-publication history The pre-publication history for this paper can be accessed here: ==== Refs Byl FM Sudden hearing loss: eight years' experience and suggested prognostic table Laryngoscope 1984 94 647 661 6325838 Arenberg IK Allen GW DeBoer A Sudden deafness immediately following cardiopulmonary bypass J Laryngol Otol 1972 86 73 77 5007954 Brownson RJ Stroud MH Carver WF Extracorporeal cardiopulmonary bypass and hearing Arch Otolaryngol 1971 93 179 182 5100939 Plasse HM Spencer FC Mittleman M Frost JO Unilateral sudden loss of hearing: an unusual complication of cardiac operation J Thorac Cardiovasc Surg 1980 79 822 826 6966351 Plasse HM Mittleman M Frost JO Unilateral sudden hearing loss after open heart surgery: a detailed study of seven cases Laryngoscope 1981 91 101 109 7453456 Shapiro MJ Purn JM Raskin C A study of the effects of cardiopulmonary bypass surgery on auditory function Laryngoscope 1981 91 2046 2052 7321724 Millen SJ Toohill RJ Lehman RH Sudden sensorineural hearing loss: operative complication in non-otologic surgery Laryngoscope 1982 92 613 617 6979666 Ness JA Stankiewicz Kaniff T Piffar R Allegretti J Sensorineural hearing loss associated with aortocoronary bypass surgery: a prospective analysis Laryngoscope 1993 103 589 593 8502091 Report from the national center for health care technology, coronary bypass surgery JAMA 1981 246 1645 1649 6115949 10.1001/jama.246.15.1645 Wright JL Saunders SH Clinical records: sudden deafness following cardiopulmonary bypass surgery J Laryngol Otol 1975 89 757 759 1185060 Cox AJ Sargent EW Sudden sensorineural hearing loss following non-otologic non cardiopulmonary bypass surgery Archives of Otolaryngology-Head and Neck Surgery 1997 123 994 998 9305253 Goodhill V Harris L Brockman SJ Hantz O Sudden deafness and labyrinthic window ruptures Annals of Otology 1973 82 2 12 Thomsen KA Terkildsen K Arnfed I Middle ear pressure variations during anesthesia Archives of Otolaryngology 1965 82 609 611 5846568 Patterson ME Bartlett PC Hearing impairment caused by intratympanic pressure changes during general anesthesia Laryngoscope 1976 86 399 404 1256214 Keane J Locked In Syndrome With Deafness Neurology 1985 35 1395 1396 4022393 Egan CA Davies L Halmagyi GM Bilateral total deafness due to pontine hematoma Journal of Neurology and Neurosurgical Psychiatry 1996 61 628 631	16336689	PMC1325038	CC BY	2021-01-04 16:03:26	no	BMC Ear Nose Throat Disord. 2005 Dec 10; 5:12	utf-8	BMC Ear Nose Throat Disord	2,005	10.1186/1472-6815-5-12	oa_comm
==== Front BMC Emerg MedBMC Emergency Medicine1471-227XBioMed Central London 1471-227X-5-101634333810.1186/1471-227X-5-10Case ReportFlecainide overdose – support using an intra-aortic balloon pump Timperley Jonathan [email protected] Andrew RJ [email protected] Peter D [email protected] Nicholas EJ [email protected] From the Department of Cardiac Rhythm Management, John Radcliffe Hospital, Oxford, OX3 9DU, UK2 From the Department Cardiology of Gloucestershire Royal Hospital, Great Western Road, Gloucester, GL1 3NN, UK2005 12 12 2005 5 10 10 15 7 2005 12 12 2005 Copyright © 2005 Timperley et al; licensee BioMed Central Ltd.2005Timperley 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 Flecainide is an antiarrhythmic agent which is being used increasingly for the management of super-ventricular arrhythmias. Overdose with flecainide is frequently fatal with mortality reported as high as 22% due to arrhythmias, myocardial depression and conduction defects leading to electro-mechanical dissociation and asytole. Supportive measures are often required during the case and previously have included inotropes, extracorporeal membrane oxygenation and cardiopulmonary bypass. Case presentation A 47 year old lady presented to the emergency department with a four hour history of severe central chest pain. Her ECG showed atrial fibrillation and broad QRS complexes with a sine wave appearance. She had a past history of paroxysmal atrial fibrillation and significant psychiatric history. Following thrombolysis for a presumed myocardial infarction she developed cardiogenic shock with severely impaired left ventricular function. An intra-aortic balloon pump was inserted and coronary angiography demonstrated normal coronary arteries. With inotropic support she improved over 48 hours, with both her QRS duration and left ventricular function returning to normal. Biochemical testing following her discharge demonstrated significantly elevated levels of flecainide. Conclusion The use of an intra-aortic balloon pump is a useful supportive measure during the acute phase of flecainide overdose associated with severe myocardial depression. ==== Body Background Flecainide is an increasingly used class 1C antiarrhythmic drug used for the management of both supra-ventricular and ventricular arrhythmias. It causes rate-dependent slowing of the rapid sodium channel slowing phase 0 of depolarisation [1] and in high doses inhibits the slow calcium channel. Flecainide also slows conduction in all cardiac fibres, increasing conduction times in the atria, ventricles, atrio-ventricular node and His-Purkinje system and can cause myocardial depression. Case presentation A 47 year old lady presented to the emergency department with a four hour history of severe central chest pain radiating to the neck. She had a significant psychiatric history with multiple episodes of self harm and had previously been hospitalised with behavioural problems. The patient had a history of multiple admissions with cardiac sounding chest pain, all with normal ECG and cardiac troponin levels. An outpatient stress nuclear perfusion scan was normal. The patient also had a history of arterial hypertension and paroxysmal atrial fibrillation and was managed on a combination of amitriptyline, losartan, amlodipine and flecainide. During this admission she repeatedly denied taking an overdose. On admission to the emergency room the patient was unresponsive with a GCS of 6. She rapidly regained consciousness but remained slightly drowsy. Pupil response was normal. Her pulse was 70 bpm and blood pressure 60/40 mmHg. The 12-lead ECG (figure 1) revealed atrial fibrillation with a broad QRS complex (320 ms) with a sine wave appearance. An ECG had been performed 4 weeks previously and was normal. She had normal serum potassium (4.1 mmol/l). The presentation with chest pain led to a presumptive diagnosis of an acute myocardial infarction and she was treated with thrombolysis with intravenous alteplase. Ten minutes later there was a significant change in her ECG (figure 2). Inotropic support was instituted with dobutamine and subsequently with adrenaline to maintain her blood pressure. She remained hypotensive with poor urine output and was transferred to the regional cardiothoracic unit for further management. Blood was taken for paracetamol, salicylate levels which were normal. Flecanide levels were requested. Blood gases showed pH 7.37, pCO2 6.1kPa, pO2 11.7, bicarbonate 25.3 mmol/l Figure 1 12-lead ECG on admission. Figure 2 12-lead ECG after thrombolysis. On arrival at our institution, transthoracic echocardiography revealed global moderate impairment of left ventricular function. Emergency coronary angiography was performed demonstrating normal epicardial coronary arteries and left ventriculography confirmed the echocardiography findings. An intra-aortic balloon pump was inserted with a rapid improvement in haemodynamics. Over the following 48 hours the patient was weaned from the inotropes and the intra-aortic balloon pump was removed. During this period her QRS duration gradually returned to normal (figure 3) with left axis deviation and first degree heart block (which was present 12 months previously). Repeat echocardiography demonstrated normal left ventricular function. Figure 3 12-lead ECG after treatment. Following discharge her flecanide levels were returned as 2340 μg/l (normal therapeutic range 200–700 μg/l) at 24 hours post admission. Discussion The mortality of overdose with class 1C agents has been reported as high as 22% compared to less than 1% for all drug overdoses [2]. In such cases fatal outcome is often related to the progression of conductions disturbance to electromechanical dissociation and asytole. Although less common, ventricular arrhythmias may occur and are frequently unresponsive to electrical cardioversion [3]. Other features may occur due to impaired tissue perfusion including hypoxia, metabolic acidosis, coma and convulsions. Hypotension can occur within minutes of a significant flecainide overdose, and subsequent reduction in hepatic and renal flow leading to prolongation of the duration of toxicity. The proarrhythmic effects of flecanide may be related to flecainide promoting re-entry in ventricular tissue [4]. Worsening of existing ventricular arrhythmias or the onset of new ones can occur in up to 30% of patients [5]. Flecainide depresses cardiac contractility especially in patients with underlying impairment of function. The management of an overdose includes supportive measures and specific pharmacological agents. The benefit of activated charcoal is uncertain but this may be considered within the first hour of ingestion [8]. Hemodynamic support may be required including the fluid replacement, inotropes and an intra-aortic balloon pump as in this case. Both extracorporeal membrane oxygenation and cardiopulmonary bypass have been used in cases of severe overdose [6,7] but the use of intra-aortic balloon pump has not previously been described. Sodium bicarbonate is recommended for a metabolic acidosis that persists despite correction of hypoxia and adequate fluid resuscitation [8]. Following administration of intravenous sodium bicarbonate QRS narrowing has been reported and also termination of ventricular tachycardia. [9]. Ventricular arrhythmias may be difficult to cardiovert electrically and both lidocaine and amiodarone have been used successfully in such cases [3,10]. Cardiac pacing may be required but ventricular capture may be poor [3]. Conclusion Flecainide overdose is frequently fatal and supportive measures including the use of intra-aortic balloon pumping may be required for severe myocardial depression that may occur during the acute phase. Competing interests The author(s) declare that they have no competing interests. Authors' contributions JT, AM and PB drafted the manuscript. NW made final alterations to manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: ==== Refs Wang Z Fermini B Nattel S Mechanism of flecainide's rate-dependent actions on action potential duration in canine atrial tissue J Pharmacol Exp Ther 1993 267 575 581 8246130 Koppel C Oberdisse U Heinemeyer G Clinical course and outcome in class IC antiarrhythmic overdose J Toxicol Clin Toxicol 1990 28 433 444 2176700 Bauman JL Gallastegui J Tanenbaum SR Hariman RJ Flecainide-induced sustained ventricular tachycardia successfully treated with lidocaine Chest 1987 92 573 575 3113836 Krishnan SC Antzelevitch C Flecainide-induced arrhythmia in canine ventricular epicardium. Phase 2 reentry? Circulation 1993 87 562 572 8425300 Miller JM Zipes DP Braunwald E, Zipes DP, Libby P Management of the patient with cardiac arrhythmias Heart Disease 2001 WB Saunders Company 700 774 Corkeron MA van Heerden PV Newman SM Dusci L Extracorporeal circulatory support in near-fatal flecainide overdose Anaesth Intensive Care 1999 27 405 408 10470398 Auzinger GM Scheinkestel CD Successful extracorporeal life support in a case of severe flecainide intoxication Crit Care Med 2001 29 887 890 11373489 10.1097/00003246-200104000-00041 Flecainide in Toxbase Goldman MJ Mowry JB Kirk MA Sodium bicarbonate to correct widened QRS in a case of flecainide overdose J Emerg Med 1997 15 183 186 9144059 10.1016/S0736-4679(96)00345-9 Siegers A Board PN Amiodarone used in successful resuscitation after near-fatal flecainide overdose Resuscitation 2002 53 105 8 11947987 10.1016/S0300-9572(01)00503-2	16343338	PMC1325039	CC BY	2021-01-04 16:31:03	no	BMC Emerg Med. 2005 Dec 12; 5:10	utf-8	BMC Emerg Med	2,005	10.1186/1471-227X-5-10	oa_comm
==== Front Nutr JNutrition Journal1475-2891BioMed Central London 1475-2891-4-341630067710.1186/1475-2891-4-34ResearchEvaluating changeability to improve fruit and vegetable intake among school aged children Nanney Marilyn S [email protected] Debra [email protected] Michael [email protected] Kimberly [email protected] Ross C [email protected] Department of Health Promotion & Education, University of Utah, Salt Lake City, UT2 Department of Community Health, School of Public Health, Saint Louis University, St. Louis, MO2005 21 11 2005 4 34 34 7 9 2005 21 11 2005 Copyright © 2005 Nanney et al; licensee BioMed Central Ltd.2005Nanney 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 purposes of this paper are two fold. First, to describe an approach used to identify fruits and vegetables to target for a child focused dietary change intervention. Second, to evaluate the concept of fruit and vegetable changeability and feasibility of applying it in a community setting. Methods Steps for identifying changeable fruits and vegetables include (1) identifying a dietary database (2) defining geographic and (3) personal demographics that characterize the food environment and (4) determining which fruits and vegetables are likely to improve during an intervention. The validity of these methods are evaluated for credibility using data collected from quasi-experimental, controlled design among 7–9 year old children (n = 304) participating in a tutoring or mentoring program in St. Louis, MO. Using a 28-item food frequency questionnaire, parents were asked to recall for their child how often foods were eaten the past 7 days. This questionnaire was repeated eight months later (response rate 84%). T-test analyses are used to determine mean serving differences from baseline to post test. Results The mean serving differences from baseline to post test were significant for moderately eaten fruits (p < .001), however, not for vegetables (p = .312). Among the intervention group, significantly more children ate grapes (p < .001), peaches (p = .022), cantaloupe (p < .001), and spinach (p = .044) at post testing – all identified as changeable with information tailored to participants. Conclusion Data driven, food focused interventions directed at a priority population are feasible and practical. An empirical evaluation of the assumptions associated with these methods supports this novel approach. However, results may indicate that these methods may be more relevant to fruits than vegetables. This process can be applied to diverse populations for many dietary outcomes. Intervention strategies that target only those changeable fruits and vegetables are innovative and warrant further study. ==== Body Background Children in the United States follow eating patterns that do not meet national recommendations for fruits and vegetables. [1,2] To date, dietary interventions have been only modestly successful in increasing fruit and vegetable (FV) intake among children [3] with effect sizes ranging from .2 servings (Gimme 5) [4] to .99 servings (High 5 Project). [5] One means of improving the impact of dietary interventions is to assure the intervention is relevant to a child by targeting foods that are available and accessible in their environment. [6] Program planning models suggest that directing program resources toward those factors that are most changeable helps to ensure program efficiency and effectiveness. [7,8] For dietary interventions, the concept of changeability can be defined as identifying those FV that have the greatest likelihood for increased consumption and targeting them for an intervention. Changeability is moderated by variations in food consumption patterns across racial and ethnic, [9-12] gender, [9,10,12-15] age, [11,13,15] marital status, [11] and regional [11,13,16] differences. Changeability is also influenced by the discrepancies in the availability of food in the community and home environments, particularly in priority populations. [17-19] From an individual perspective, the food experiences to which young children are exposed are critical to the early development of food acceptance patterns and choices. This exposure to FV influences familiarity, preferences, and intake. For purposes of an intervention, FV that are preferred and eaten by a child are less changeable because consumption goals are met. Instead, moderate intake of FV suggests foods that are familiar, accessible, and changeable. Systematic approaches derived from program planning constructs, are needed to determine program relevance and changeability in order to successfully impact health behaviors. [8] The purpose of this paper is to describe a systematic approach used for identifying changeable FV that were targeted for intervention in Partners of all Ages Reading About Diet and Exercise (PARADE), a school based mentoring program designed to improve FV intake of children ages 7 to 9 years. Children enrolled in the mentoring program received one to one tutoring on a weekly basis during regular school hours. PARADE was incorporated within the routine curriculum of the mentoring program. PARADE was based on social cognitive theory and included (1) eight lesson plans with computer-generated storybooks tailored to the dietary patterns of each 7–9 year old child and (2) eight mailed parent newsletters introducing each book and offering tips on how to role model healthy eating at home. PARADE was evaluated using a quasi-experimental study design. Participants in the evaluation include children and their parents. Approval for this study was obtained from the Saint Louis University Institutional Review Board. Methods Steps for identifying changeable FV Limited time and resources increased the importance of identifying appropriate foods for change in this priority population. The following steps were used to identify and tailor information on FV as part of PARADE lesson plans and storybooks. Step 1: Identify a comprehensive nutrition database Dietary measurement is complex. Accuracy can be maximized with methods that help individuals correctly recall the foods and amounts eaten on any given day that represents usual intake. Furthermore, a comprehensive list of foods and nutrient values is essential to calculating the dietary outcomes of interest. [20] The Continuing Survey of Food Intakes by Individuals (CSFII) database represents data from the priority population and is methodologically and psychometrically sound. Also popularly known as the "What We Eat in America Survey," the CSFII is a national food consumption survey conducted by the Agricultural Research Service of the United States Department of Agriculture. [21] The 10th edition of the CSFII (1994–1996) provides nationally representative data by over sampling for low-income individuals and young children. Individuals are asked to provide food intakes on two nonconsecutive days using a multiple pass 24-hour dietary recall administered in the home by trained interviewers. Sample sizes include 12,700 adults of all ages and 11,800 children birth-19 years. Response rates include 1-day 80% and 2-day 76%. Food consumption data are available on CD-ROM from the USDA. More information, including ordering the CD's, can be found at . Step 2: Define the sociodemographic characteristics of the priority population Variations in dietary patterns are influenced by individual characteristics such as race, [9-12] gender, [9,10,12-15] and age. [11,13,15] Thus, it is critical to assess these factors with regard to determining changeability of fruits and vegetables. For the CSFII database, food consumption differences can be viewed by five race categories; white, black, Pacific Asian, Native American, and Other. Furthermore, the data can be segmented by gender and age. The CSFII database reports age in months for children less than one year of age, and in years for those over 1 year of age. For Project PARADE, the CSFII data was analyzed for children 7–9 years of age (9% of CSFII sample), male (49%) and female (51%), and African American (18%). 892 children matched these criteria. Step 3: Define the geographic characteristics of the priority population Dietary intake varies by geographic region and urban versus rural influences. The CSFII divides the United States into five geographic regions that are further defined by urbanization type. Urbanization types include city, outside the city, and rural areas. PARADE participants lived in ten urban and suburban counties in a large Midwest city. Therefore, the CSFII data was analyzed for those living in a metropolitan city and outside the city in the Midwest. Results identified nearly three-quarters of the sample living in a metropolitan area (central city, 21% and suburban, 53%). These geographic criteria resulted in a final sample of dietary intake data for 164 children, mean age 7.93 (SD = .82). Steps 1–3 identified 2423 food entries by the CSFII. This included 670 unique foods including 49 fruits and 79 vegetables consumed by the sample. Step 4: Identify changeable FV for the priority population Individual characteristics and environmental exposures influence food familiarity and determine frequency of consumption. We next selected those foods meeting criteria for 'moderate consumption'. CSFII food consumption data is organized by numerical food codes and food amounts. Data can be viewed in a number of ways; specific food, food form (canned, frozen, raw), how often a food was eaten, and the amount eaten (gram weight). Individual food items were rank ordered by the percentage of respondents who reported consuming them over the observation period. Foods whose rank fell into the 25th to 75th percentile were defined as "moderately consumed". Five fruits and eight vegetables were identified as moderately consumed and targeted for an intervention among these school-aged children. Foods in the top quartile (>75%) or most frequently consumed (i.e., oranges, carrots) and FV at the lower range (<25%) of consumption (i.e., grapefruit, sweet potatoes) were not included in the intervention. Table 1 summarizes the steps that define the key characteristics to consider when determining relevant foods for a dietary intervention. Table 1 Steps to identify relevant foods to target in a dietary intervention Steps to Determine Foods for a Dietary Intervention Key Characteristics that Effect Dietary Intake Define Sociodemographic Variables Marital status • Married, Divorced, Separated, Single, Widowed Race • Caucasian, African American, Pacific Asian, Native American, and Other Gender • Male, Female Age • 0–12 Months, Years Define Geographic Variables Regional • North, South, East, West, Midwest Urbanization • City, Suburban, Rural Identify Changeable Foods to Target Consumption Frequency • Rank % between 25 and 75 Results Using national frequency data to estimate local FV consumption An important assumption being made is that accurate conclusions can be drawn from food consumption frequencies collected from a national sample because they mirror local consumption patterns. Table 2 compares FV consumption frequency identified from the CSFII with baseline PARADE data (intervention and delayed intervention groups) and indicates agreement between CSFII rank % and PARADE rank % consumptions. For example, according to the national data, those FV eaten most frequently (oranges, juice, apples, potatoes, lettuce) correspond with the PARADE baseline data as indicated by the 76–100 rank percents. Similar agreement occurs for those fruits and vegetables eaten moderately (25–75 rank percent) and least often (<25 rank percent). However, a few discrepancies are noted. Although not identified as a moderately eaten fruit from the national data set, kiwi falls within the moderately eaten range at PARADE baseline (13% versus 31%) and therefore should have been targeted as changeable in the intervention. Vegetable consumption was less congruent than fruit consumption. Two vegetables (corn, carrots) were eaten more often in the national sample than reported by PARADE children at baseline. Conversely, PARADE children ate green beans and cabbage slaw more often. Overall, the authors conclude that using national dietary data to identify percent rank cut points defining consumption frequency can be useful when developing community interventions. These methods may be more accurate in predicting fruit consumption than vegetable consumption. Table 2 Fruits and vegetables identified as changeable and targeted for intervention Fruit CSFII Weighted % Consumed1 (n = 502) CSFII Rank % PARADE Baseline % Consumed2 (n = 304) PARADE Baseline Rank % Oranges/juice 24.3% 100% 79.6% 100% Other juice 17.3% 94% 67.8% 94% Apples 13.4% 88% 49.7% 88% Bananas 9.0% 81% 49.0% 81% CF Fruit Cocktail 4.5% 75% 32.9% 56% CF Grapes 4.2% 69% 46.4% 75% CF Peaches 2.8% 50% 28.9% 25% CF Cantaloupe 2.4% 44% 17.4% 31% CF Strawberries 2.1% 31% 43.7% 63% CF Pineapple 1.8% 25% 20.1% 38% Kiwi .84% 13% 18.2% 31% Nectarines .48% 6% 15.2% 19% Grapefruit .12% 0% 9.2% 6% Vegetables CSFII Weighted % Consumed1 CSFII Rank % PARADE Baseline % Consumed2 PARADE Baseline Rank % Potatoes 48.4% 100% 95.1% 100% Lettuce 12.1% 89% 62.2% 79% Corn 9.6% 84% 52.6% 63% Carrots 8.7% 79% 54.6% 68% CV Green beans 7.62% 74% 72.0% 89% CV Mixed Vegetables 6.3% 68% 46.7% 53% CV Tomatoes 5.2% 63% 49.0% 58% CV Cabbage slaw 2.8% 47% 13.2% 11% CV Beans 4.3% 53% 25.3% 26% CV Green Peas 2.7% 42% 38.2% 37% CV Broccoli 2.4% 32% 55.3% 74% CV Spinach/greens 1.2% 26% 52.6% 63% Bell Pepper 1.1% 16% 12.1% 5% Sweet potatoes .6% 11% 14.2% 16% Cauliflower .5% 5% 18.2% 21% CF = Changeable Fruits CV = Changeable Vegetables 1 CSFII weighted percent consumed was calculated from the percentage of respondents who reported consuming each food item at least once during the period under observation (two days). These data were weighted to represent urban and sub-urban Midwesterners using weights provided by the CSFII study. 2 PARADE baseline percent consumed was calculated from the percentage of respondents who reported consuming each food item during the past week. After identifying FV for intervention, we incorporated this information into 8 tailored storybooks. Professionally trained interviewers then administered a telephone survey to parents of children enrolled in participating tutoring programs. Using a 28-item FV food frequency questionnaire, parents were asked to recall for their child how often select foods were eaten within the past 7 days. This questionnaire was repeated eight months later. Response choices included "none, 1 time, 2 times, 3–4 times, 5–6 times, 7 or more times in the past week". The questionnaire demonstrated acceptable internal consistency (Chronbach's alpha = .744). Table 3 presents the PARADE changeability results. Six PARADE fruits (PF) and eight vegetables (PV) identified as moderately consumed are hypothesized to be the most likely to improve from baseline to post intervention. Differences in the percent of children who consumed the "same (unchanged) or more" of each PF and PV from baseline to posttest indicate that those targeted as changeable in the intervention were more likely to be eaten. Table 3 Fruits and vegetables targeted for Project PARADE improved at post intervention Fruit Pre-Post Change p-value Average Pre % consumed Average Post % consumed Oranges/Juice Unchanged .464 Other Juice Unchanged .138 62% 61% Apples Unchanged .139 Bananas Unchanged .414 PF Fruit Cocktail Unchanged .387 PF Grapes More <.001 PF Peaches More .022 32% 38% PF Cantaloupe More <.001 PF Strawberries Unchanged .355 PF Pineapple Unchanged .426 Grapefruit Unchanged .149 14% 7% Vegetable Pre-Post Change p-value Average Pre % consumed Average Post % consumed Potatoes More .025 Lettuce Unchanged .118 66% 71% Corn More .044 Carrots Unchanged .268 PV Green beans Unchanged .456 PV Mixed Vegetables Unchanged .062 PV Tomatoes Unchanged .230 PV Cabbage slaw Unchanged .500 44% 45% PV Beans Unchanged .202 PV Green Peas Unchanged .440 PV Broccoli Unchanged .441 PV Spinach/greens More .044 Sweet potatoes Unchanged .500 15% 14% Other vegetables Unchanged .470 PF = PARADE Fruits PV = PARADE Vegetables Among the intervention group, significantly more children ate grapes (p < .001), peaches (p = .022), cantaloupe (p < .001), and spinach (p = .044) at post testing – all identified as changeable with information tailored to PARADE participants. As hypothesized, consumption of fruits and vegetables frequently and rarely eaten remained unchanged or were eaten more in two instances (potatoes, corn). Worth noting is that the majority of the vegetables remained unchanged at the end of the intervention. These results may indicate that these methods may be more relevant to fruits than vegetables. Table 4 summarizes the results by grouping all frequently, moderately, and rarely consumed FV. Average (or mean) baseline servings among PARADE intervention participants (N = 304) are consistent with the assumptions presented where those described as frequently (Mean = 1.75 servings), moderately (Mean = 0.75 servings), and rarely (Mean = 0.26 servings) consumed follow that pattern. T-test analyses indicate that mean serving differences from baseline to post test were significant for moderately eaten fruits (p < .001), however, not for vegetables (p = .312). In general, the authors conclude that applying the concept of changeability to a FV intervention among school-aged children is feasible and supported by this data. Table 4 Moderately eaten fruits improved at post intervention PARADE Intervention (N = 304) Mean Servings Baseline Mean Servings Post T-test Difference p-value Frequently Eaten Fruits 1.75 1.72 .504 .615 Moderately Eaten Fruits .746 .932 -4.15 <.001 Rarely Eaten Fruits .264 .213 .802 .423 Frequently Eaten Vegetables 1.36 1.41 -1.43 .155 Moderately Eaten Vegetables .720 .745 -1.013 .312 Rarely Eaten Vegetables .181 .159 .393 .695 Discussion Recent findings suggest that innovative research is necessary to broaden the traditional approach beyond increasing FV awareness and education. [22] Overall, this study reinforces these methods as a way to systematically identify FV to target for an intervention. Identifying and concentrating on the most changeable behavioral targets for interventions direct resources to where they will be most beneficial. Furthermore, greater specificity in program development simplifies the evaluation process (i.e., brief food frequency questionnaire). [23,24] Assessments of changeability across fruit and vegetable patterns can be made by defining person and place variables of the priority population, examining the frequency with which fruits and vegetables have been consumed, and identifying those that can be reasonably expected to change for targets of an intervention. Moreover, this approach takes into account important environmental influences upon dietary patterns. Haire-Joshu and Nanney describe individual food preference, cultural and familial influences, and home, school and community environments as having significant influences upon the food environment of children. [6] This process addresses eating behavior as a function of the varied food environments for a specific population, albeit, in a broader community based context. We developed, applied, and tested a systematic, data based approach to assess changeability and specify fruits and vegetables for an intervention among underserved school aged children. This approach will allow for further clarity of intervention effects by targeting only those changeable fruits and vegetables for intervention. Furthermore, this process can be applied to diverse populations for a variety of dietary outcomes. Additionally, larger mass media interventions like the 5 A Day campaign may benefit from this approach. More research is needed to evaluate the effectiveness and generalizability of community-based efforts that promote changeable foods for an intervention. Competing interests The author(s) declare that they have no competing interests. Authors' contributions MSN conceived of the concept of changeability and applied the methods to the intervention, developed the data collection instrument and drafted the manuscripts. DHJ conceived of the study, and participated in its design and coordination and helped to draft the manuscript. ME managed the data, performed statistical analysis, and helped to draft the manuscript. KH contributed to the conception of the study, application of the methods, and helped draft the manuscript. RB participated in the design of the study, oversaw the statistical analysis, and helped to draft the manuscript. Acknowledgements This work is based upon funding from the American Cancer Society (TURPG-00-286-01 PBP) and National Institutes of Nursing Research (USPHS-5-R01-NR05079). The authors would like to recognize Program Manager of the PARADE Project Brandye L. Mazdra and our community partners; The OASIS Institute, St. Louis area Big Brothers Big Sisters, and Girls Inc. ==== Refs Krebs-Smith SM Cook A Subar AF Cleveland L Friday J Kahle LL Fruit and Vegetable Intakes of Children and Adolescents in the United States Arch Pediatr Adolesc Med 1996 150 81 86 8542012 Munoz KA Krebs-Smith SM Ballard-Barbash RB Cleveland LE Food Intakes of US Children and Adolescents Compared With Recommendations Pediatrics 1997 100 323 329 9282700 10.1542/peds.100.3.323 Ammerman A Lindquist C Lohr KN Hersey J The Efficacy of Behavioral Interventions to Modify Dietary Fat and Fruit and Vegetable Intake: A Review of the Evidence Preventive Medicine 2002 35 25 41 12079438 10.1006/pmed.2002.1028 Baranowski T Davis M Resnicow K Baranowski J Doyle C Lin LS Smith M Wang DT Gimme 5 fruit, juice, and vegetables for fun and health: outcome evaluation [published erratum appears in Health Educ Behav 2000 Jun;27(3):390] Health Educ Behav 2000 27 96 111 10709795 Reynolds KD Franklin FA Binkley D Raczynski JM Harrington KF Kirk KA Person S Increasing the fruit and vegetable consumption of fourth-graders: results from the high 5 project Prev Med 2000 30 309 319 10731460 10.1006/pmed.1999.0630 Haire-Joshu D Nanney MS Prevention of overweight and obesity in children: Influences on the Food Environment The Diabetes Educator 2002 28 415 422 12068650 Glanz K Lewis FM Rimer BK Health behavior and health education: theory, research, and practice 1997 2 San Francisco , Jossey-Bass Green LW Kreuter MW Behavioral and environmental assessment Health Promotion and Planning 1999 3rd. California , Mayfield Publishing Company 145 Akan GE Grilo CM Sociocultural influences on eating attitudes and behaviors, body image, and psychological functioning: a comparison of African American, Asian American, and Caucasian college women International Journal of Eating Disorders 1995 18 181 187 7581421 Bronner Y Nutritional status outcomes for children: Ethnic, cultural, and environmental contexts Journal of the American Dietetic Association 1996 96 891 903 8784334 10.1016/S0002-8223(96)00242-8 Thompson B Demark-Wahnefried WD Taylor G Baseline fruit and vegetable intake among adults in seven 5 a day study centers located in diverse geographic areas J Am Diet Assoc 1999 99 1241 1248 10524389 10.1016/S0002-8223(99)00306-5 Xie B Gilliland FD Li YF Rockett HR Effects of Ethnicity, Family Income, and Education on Dietary Intake among Adolescents Preventive Medicine 2003 36 30 40 12473422 10.1006/pmed.2002.1131 DiSogra L Glanz K The 5 A Day Virtual Classroom: an on-line strategy to promote healthful eating Journal of the American Dietetic Association 2000 100 349 352 10719410 10.1016/S0002-8223(00)00106-1 Kumanyika S Tell GS Shemanski L Polak J Savage PJ Eating patterns of community-dwelling older adults: the Cardiovascular Health Study 4 1994 Ann Epidemiol Shi L Sociodemographic characteristics and individual health behaviors Southern Medical Journal 1998 91 933 941 9786288 Kirby S Baranowski J Reynolds K Taylor G Binkley D Children's fruit and vegetable intake: Sociodemographic, adult child, regional, and urban-rural influences Journal of Nutrition Education 1995 27 261 271 Hearn MD Baranowski J Doyle C Smith M Lin LS Resnicow K Environmental influences on dietary behavior among children: availability and accessibility of fruits and vegetables enable consumption J Health Educ 1998 29 26 32 Kratt P Reynolds K Shewchuk R The Role of Availability as a moderator of Family Fruit and Vegetable Consumption Health Education and Behavior 2000 27 471 482 10929754 Morland K Wing S Roux AD The Contextual Effect of the Local Food Environment of Residents' Diets: The Atherosclerosis Risk in Communities Study American Journal of Public Health 2002 92 1761 1767 12406805 Buzzard M Willett W 24-hour dietary recall and food record methods Nutritional Epidemiology 1998 2nd New York , Oxford University Press Continuing Survey of Food Intakes by Individuals (CSFII) and the Diet and Health Knowledge Survey (DHKS), 1994-96 1998 Serdula MK Gillespie C Kettel-Khan L Farris R Seymour J Denny C Trends in fruit and vegetable consumption among adults in the United States: Behavioral Risk Factor Surveillance System, 1994-2000 American Journal of Public Health 2004 94 1014 1018 15249308 Nanney MS Hessler K Kelly C Haire-Joshu D Identifying usual portion sizes for commonly eaten foods among 7-9 year olds using parent-child consensus 2003 Saint Louis University School of Public Health Nielsen SJ Popkin BM Patterns and trends in food portion sizes, 1977-1998 American Medical Association 2003 289 450 453 10.1001/jama.289.4.450	16300677	PMC1325040	CC BY	2021-01-04 16:27:12	no	Nutr J. 2005 Nov 21; 4:34	utf-8	Nutr J	2,005	10.1186/1475-2891-4-34	oa_comm
==== Front Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1431633665910.1186/1465-9921-6-143ResearchBiphasic effect of extracellular ATP on human and rat airways is due to multiple P2 purinoceptor activation Mounkaïla Boutchi [email protected] Roger [email protected] Etienne [email protected] Laboratoire de Physiologie Cellulaire Respiratoire, Université Bordeaux 2, Bordeaux, F-33076 France; Inserm, E356, Bordeaux, F-33076 France2005 8 12 2005 6 1 143 143 7 10 2005 8 12 2005 Copyright © 2005 Mounkaïla et al; licensee BioMed Central Ltd.2005Mounkaïla 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 Extracellular ATP may modulate airway responsiveness. Studies on ATP-induced contraction and [Ca2+]i signalling in airway smooth muscle are rather controversial and discrepancies exist regarding both ATP effects and signalling pathways. We compared the effect of extracellular ATP on rat trachea and extrapulmonary bronchi (EPB) and both human and rat intrapulmonary bronchi (IPB), and investigated the implicated signalling pathways. Methods Isometric contraction was measured on rat trachea, EPB and IPB isolated rings and human IPB isolated rings. [Ca2+]i was monitored fluorimetrically using indo 1 in freshly isolated and cultured tracheal myocytes. Statistical comparisons were done with ANOVA or Student's t tests for quantitative variables and χ2 tests for qualitative variables. Results were considered significant at P < 0.05. Results In rat airways, extracellular ATP (10-6–10-3 M) induced an epithelium-independent and concentration-dependent contraction, which amplitude increased from trachea to IPB. The response was transient and returned to baseline within minutes. Similar responses were obtained with the non-hydrolysable ATP analogous ATP-γ-S. Successive stimulations at 15 min-intervals decreased the contractile response. In human IPB, the contraction was similar to that of rat IPB but the time needed for the return to baseline was longer. In isolated myocytes, ATP induced a concentration-dependent [Ca2+]i response. The contractile response was not reduced by thapsigargin and RB2, a P2Y receptor inhibitor, except in rat and human IPB. By contrast, removal of external Ca2+, external Na+ and treatment with D600 decreased the ATP-induced response. The contraction induced by α-β-methylene ATP, a P2X agonist, was similar to that induced by ATP, except in IPB where it was lower. Indomethacin and H-89, a PKA inhibitor, delayed the return to baseline in extrapulmonary airways. Conclusion Extracellular ATP induces a transient contractile response in human and rat airways, mainly due to P2X receptors and extracellular Ca2+ influx in addition with, in IPB, P2Y receptors stimulation and Ca2+ release from intracellular Ca2+ stores. Extracellular Ca2+ influx occurs through L-type voltage-dependent channels activated by external Na+ entrance through P2X receptors. The transience of the response cannot be attributed to ATP degradation but to purinoceptor desensitization and, in extrapulmonary airways, prostaglandin-dependent PKA activation. ==== Body Background ATP is an extracellular messenger released by different cells that modulate lung functioning. ATP can be liberated from parasympathetic nerves as co-transmitter with acetylcholine [1], from epithelial cells [2], for example following exposure to air pollutants [3], and is released, probably from cell lysis, during lung injury [4]. ATP stimulates surfactant production by type II pneumocytes [5], Cl- secretion by epithelial cells and the activity of the mucociliary escalator [6]. ATP also acts on airway smooth muscle (ASM) cells, inducing ASM cell proliferation [7] and changes in airway contractility [8]. Receptors for ATP are classified into 2 families. P2X receptors are ionotropic receptors that, upon activation by ATP, initiate extracellular Ca2+ and Na+ influx. P2Y receptors are 7-transmembrane domain receptors that are coupled to G-proteins. When stimulated, they activate PLC leading to inositol 1,4,5-trisphosphate production and intracellular Ca2+ release via Gq/11 protein, or modulate cAMP production and PKA activity via Gs or Gi binding [9,10]. It has been shown that extracellular ATP modulates cytosolic Ca2+ response and contraction in a variety of smooth muscle. However, its effect on airway smooth muscle reactivity has not been comprehensively investigated and the results are quite controversial. In normal rat, intratracheal instillation of ATP in vivo increases airway resistance [11]. In lung slides obtained from isolated mouse lung, Bergner and co-workers have shown that ATP induced a transient contraction and cytosolic Ca2+ oscillations mediated by P2Y purinoreceptors, but has no effect on acetylcholine-induced contraction [8]. By contrast, Aksoy and Kelsen [12] have shown in isolated rabbit tracheal strips that ATP alone did not produce any contraction but rather induced relaxation on strips precontracted with acetylcholine, a mechanical response due to P2 receptor activation. A relaxant effect on precontracted isolated rings has also been reported in guinea-pig trachea, but this effect was attributed to P1 receptor stimulation [13]. When present, the contractant effect of ATP alone seems to be associated with [Ca2+]i increase. Bergner and co-workers reported, in mouse freshly ASM cells, that ATP induced an oscillating [Ca2+]i response [8], while Michoud and co-workers observed in cultured rat trachea cells a non oscillating [Ca2+]i response [14]. Both authors attributed the [Ca2+]i response to intracellular Ca2+, whereas in pig cultured ASM cells, Sawai and co-workers showed that the ATP-induced [Ca2+]i response was decreased in the presence of extracellular Ca2+ [15,16]. The aim of this study was therefore to characterize the effect of extracellular ATP on airway reactivity. Since results obtained in airways with different calibres suggest that it may act differentially along the airway tree, we compared the effect of ATP in rat trachea, extrapulmonary bronchi (EPB) and intrapulmonary bronchi (IPB) and, additionally, in human IPB. We have investigated whether ATP modulation of airway reactivity was due to an indirect or direct action on airway smooth muscle cells. We have also determined the pharmacological profile of the receptors involved in the ATP-induced response and the subsequent intracellular pathways, and, finally, we have assessed the implication of enzymatic ATP degradation in the response pattern to purinergic stimulation. Methods Preparation of rat tissues Rat airways were obtained from male Wistar rats 10–15 weeks old, weighing 300–400 g. Animals were treated and sacrificed according to national guidelines, with approval of the local ethical committee. For each experiment, a rat was stunned and killed by cervical dissociation. Heart and lungs were removed in bloc, and the trachea, the extracellular bronchi and the first left intrapulmonary bronchus were dissected under binocular control. For isometric contraction experiments, rings about 3 mm in length were obtained from 1st, 2nd and 3rd airway generations, i.e., trachea, left and right extrapulmonary and left IPB. In order to avoid possible biases due to variation in ring size, contraction was normalised to a reference functional response (see below). When needed, the epithelium was mechanically removed. Preparation of human bronchial rings Human bronchial rings were obtained from lung pieces collected for histological examination following resection for carcinoma. As in previous studies [17] specimens were selected from 15 patients whose lung function was within a normal range, i.e., whose forced expiratory volume in 1 second (FEV1) was above 80% of predicted. Quickly after resection, segments of human bronchi (3rd to 5th generation; 3–5 mm in internal diameter) were carefully dissected from a macroscopically tumour-free part of each of the histological pieces and transferred to the laboratory in an ice-cold PSS solution. Segments were then cut into rings measuring about 4–5 mm in length for isometric contraction measurements. Use of human tissues was performed according to national guidelines, in compliance with the Helsinki Declaration. Obtention of freshly isolated and cultured cells For isolated cell-experiments, the muscular strip located on the dorsal face of the rat trachea was further dissected under binocular control. The epithelium-free muscular strip was cut into several pieces and the tissue was then incubated overnight (14 h) in low-Ca2+ (200 μM) physiological saline solution (PSS; composition given below) containing 0.5 mg·ml-1 collagenase, 0.35 mg·ml-1 pronase, 0.03 mg·ml-1 elastase and 3 mg·ml-1 bovine serum albumin at 4°C. After this time, the muscle pieces were triturated in a fresh enzyme-free solution with a fire polished Pasteur pipette to release cells, which were collected by centrifugation. In control experiments, immunocytochemistry was performed using monoclonal mouse anti-smooth muscle α-actin antibodies and FITC-conjugated anti-mouse IgG antibodies to verify that the isolated cells obtained by dissociation were smooth muscle cells (data not shown). For experiments on freshly isolated cells, cells were stored for 1 to 3 h to attach on glass coverslips at 4°C in PSS containing 0.8 mM Ca2+ and used on the same day. For cell culture, coverslips with attached cells were placed in multiwell plates at 37°C in humidified air containing 5% CO2 in DMEM containing 0.5 U·mL-1 penicillin, 0.5 mg·mL-1 streptomycin and 0.25 μg·mL-1 amphotericin B, and cultured in non-proliferating and proliferating conditions. For experiments in non-proliferating conditions, cells (15000 cells·mL-1) were cultured in the above-described DMEM supplemented with insulin, and ITS medium, which maintains the cells in quiescent state. For experiments in proliferating conditions, cells (7500 cells/mL) were cultured in the above-described DMEM supplemented with 10% foetal bovine serum. After 10 days, confluent cells were detached with a 0.5% trypsin-0.02% EDTA, resuspended and stored for 1 h to attach on coverslips at 4°C before use. Isometric contraction measurement Isometric contraction was measured in isolate rings that were mounted between two stainless steel clips in vertical 5 ml organ baths of a computerized isolated organ bath system (IOX, EMKA Technologies, Paris, France) previously described [17]. Baths were filled with Krebs-Henseleit (KH) solution (composition given below) maintained at 37°C and bubbled with a 95% O2-5% CO2 gas mixture. The upper stainless clip was connected to an isometric force transducer (EMKA Technologies). Tissues were set at optimal length (Lo) by equilibration against a passive load of 1.5 g for extrapulmonary airways and 1 g for IPB. At the beginning of each experiment, supramaximal stimulation with acetylcholine (ACh, 10-3 M final concentration in the bath) was administered to each of the rings to elicit a reference response. Rings were then washed with fresh KH solution to eliminate the ACh response. After the tension returned to baseline, the organ bath was filled with the appropriate solution, and unique or non-cumulative concentrations of agonists were added to the bath and the subsequent variation in tension recorded, and expressed as a percentage of the reference response to ACh in that ring. Each type of experiment was repeated for the number of rings from different specimens indicated in the text. In epithelium-free experiments, the epithelium of isolated rings was rubbed using a plastic cylinder introduced in the lumen of the ring. Rings were frozen at the end of the experiment for histological examination of actual removal of the epithelium (data not shown). Fluorescence measurement and estimation of [Ca2+]i [Ca2+]i responses of isolated tracheal myocytes were monitored fluorimetrically using the Ca2+-sensitive probe indo-1 as previously described [18]. Briefly, freshly isolated cells were loaded with indo-1 by incubation in PSS containing 1 μM indo-1 AM for 25 min at room temperature and then washed in PSS for 25 min. Coverslips were then mounted in a perfusion chamber and continuously superfused at room temperature. A single cell was illuminated at 360 ± 10 nm. Emitted light from that cell was counted simultaneously at 405 nm and 480 nm by two photomultipliers (P100, Nikon). [Ca2+]i was estimated from the 405/480 ratio using a calibration for indo-1 determined within cells. ATP or ACh was applied to the tested cell by a pressure ejection from a glass pipette located close to the cell. No change in [Ca2+]i was observed during test ejections of PSS (data not shown). Generally, each record of [Ca2+]i response was obtained from a different cell. Each type of experiment was repeated for the number of cells indicated in the text. Solution, chemicals and drugs Normal PSS contained (in mM): 130 NaCl, 5.6 KCl, 1 MgCl2, 2 CaCl2, 11 glucose, 10 Hepes, pH 7.4. Normal KH solution contained (in mM): 118.4 NaCl, 4.7 KCl, 2.5 CaCl2·2H2O, 1.2 MgSO4·7H2O, 1.2 KH2PO4, 25.0 NaHCO3, 11.1 D-glucose, (pH 7.4). In Ca2+-free solution, Ca2+ was removed and 0.4 mM EGTA was added. In order to keep the osmotic pressure constant, in Na+-free solution, Na+ was omitted and replaced by N-methyl-D-glucamine, and, for KCl-induced contraction, KCl was substituted to NaCl for the desired concentrations. Collagenase (type CLS1) was from Worthington Biochemical Corp. (Freehold, NJ, USA). Bovine serum albumin, acetylcholine, carbachol, ATP, ATP-γ-S, α-β-methylene ATP, D600, RB2, H-89, caffeine and thapsigargin were purchased from Sigma (Saint Quentin Fallavier, France). Indo-1 AM was from Calbiochem (France Biochem, Meudon, France). Indo-1 AM and thapsigargin were dissolved in dimethyl sulphoxide which maximal concentration used in our experiments was < 0.1% and had no effect on the resting value of the [Ca2+]i (data not shown). DMEM, ITS, penicillin, streptomycin, amphotericin B and foetal bovine serum were from GIBCO-BRL (Invitrogen, Eragny-sur-Oise, France). Data analysis and statistics Data are given as mean ± SEM. The maximal contraction Fmax was taken as the apparent maximal response, i.e., the response obtained with the maximal concentration used, even though the CRC had not reached a plateau. Overall differences in CRC were performed by ANOVA test. The transient effect of ATP was estimated by TR10, the time needed for the tension value to decrease to 10% Fmax, calculated from the maximal contraction. Fmax and TR10 were compared using Student's t tests. Statistical comparisons of [Ca2+]i response of isolated cells were carried out with Student's t tests for quantitative variables and χ2 tests for qualitative variables. Results were considered significant at P < 0.05 Results Effect of ATP on rat and human isolated airways ATP induced a fast and transient contraction of rat isolated airway rings which amplitude depended on the concentration of agonist and the location along the airway tree. Original trace obtained in IPB is presented in figure 1A. Non-cumulative concentration-response curves, shown in figure 1C, indicated that the ATP-induced contraction was the greatest in IPB, and the lowest in trachea (n = 7 to 10). The time needed to return to baseline, expressed as TR10, is shown in figure 1E. As in rat airways, ATP induced a transient contractile response in human IPB, as illustrated by the original trace shown in figure 1B. The maximal response was in the same range as that observed in rat IPB (Figure 1D). However, the return to baseline was much slower in human bronchi (figure 1F) (n = 7). Figure 1 Effect on ATP isolated airway rings. A: typical trace of the effect of 10-3 M ATP on rat IPB. B: typical trace of the effect of 10-3 M ATP on human IPB. C: mean ATP-induced non-cumulative response curves in trachea (black circles) right EPB (down triangles), left EPB (up triangles) and left IPB (squares) from rat airways (n = 10). D: mean ATP-induced non-cumulative response curves in human IPB (n = 7). E: TR10 in rat trachea (black column) right (REPB) and left EPB (LEPB) (hatched columns), and left IPB (cross-hatched column). F: TR10 in human IPB (cross-hatched column) Error bars and SEM. P < 0.05. Effect of ATP on rat epithelium-free isolated airways In this set of experiments, for each rat, ATP was applied at fixed concentration (10-3 M) on epithelium-denuded rings. Measurements were repeated on 6 to 8 specimens. The response pattern was similar to that obtained in intact rings (Figure 2A). Statistical comparison showed no difference between intact and epithelium-free rings, either on the maximal contractile response or on the return to baseline (figure 2B and 2C). Figure 2 Effect on ATP on rat epithelium-free isolated airway rings. A: typical trace of the effect of 10-3 M ATP on epithelium-free rat EPB. B: Fmax to 10-3 M ATP in epithelium-free rings from trachea (n = 8), left and right EPB (n = 6), and left IPB (n = 7). Horizontal bars are Fmax in control rings. C: TR10 in rat trachea (black column) right and left EPB (hatched columns), and left IPB (cross-hatched column). Error bars are SEM. P < 0.05. Effect of ATP on freshly isolated and cultured tracheal myocytes In a first set of experiments, ATP was applied at 10-6 M (n = 33), 10-5 M (n = 65), 10-4 M (n = 97), and 10-3 M (n = 82) on myocytes freshly isolated from rat trachea. Original representative [Ca2+]i responses are shown in figure 3A, and results are summarised in figure 3B, C. ATP stimulation resulted in a transient [Ca2+]i rise followed, in some cases, by several subsequent [Ca2+]i oscillations. The percentage of responding cells, the amplitude of the [Ca2+]i peak, and the percentage of oscillating responses were concentration-dependent. Similar experiments were performed with 10-5ACh (n = 61), a concentration that induces the maximal [Ca2+]i response [18]. The percentage of responding cells was 100%, the amplitude of the [Ca2+]i peak was 627 ± 30.2 nM, the percentage of oscillating response was 39.3%, and the frequency of oscillations was 7.83 ± 0.69 oscillations/min. Compared to the cholinergic response, the percentage of responding cells to 10-3 M ATP and the frequency of oscillations were significantly lower, but not the amplitude of the peak nor the percentage of oscillating responses. Figure 3 Effect of ATP on freshly isolated rat tracheal myocytes. A: original traces of the effect of several ATP concentrations (10-6 to M 10-3 M) on freshly isolated rat tracheal myocytes (n = 33 to 97 for each concentration). B: percentage of responding cells depending on ATP concentration (left panel) and percentage of oscillating responses in responding cells. C: abscissa: log concentration of ATP (M). Ordinates: amplitude of the Ca2+ peak (left panel) in responding cells (left panel) and oscillation frequency in oscillating cells. Since some authors have observed a [Ca2+]i response to ATP only in cultured cells [15], we investigated the [Ca2+]i response to 10-3 M ATP in cells cultured for 3 days (n = 27) in non-proliferating medium and 10 days in proliferating medium(n = 35) (figure 4). Culture did not significantly alter the number of responding cells. 72 h-culture decreased the amplitude of the [Ca2+]i peak to ATP. In 10 day-cultured cells, the amplitude of the [Ca2+]i peak re-increased up to the values observed in non-cultured myocytes, and the general profile of the response dramatically altered, as shown in the original trace (figure 4B). To see whether the effect of cell culture on the [Ca2+]i response was specific to ATP, we compared the Ca2+ response to ACh in cultured cells (n = 26) with that obtained in freshly isolated cells. After 2 days of culture in non-proliferating medium, the percentage of responding cells as well as the amplitude of the [Ca2+]i peak in responding cells were significantly reduced (figure 4C and 4D), and oscillating responses were only 12.5%. Figure 4 Effect of ATP and ACh on cultured rat tracheal myocytes. A: percentage of cells responding to 10-3 M ATP, and amplitude of the [Ca2+]i peak, in cells cultured for 72 h in non-proliferating medium (black columns, n = 27) and in cells cultured for 10 days in proliferating medium (open columns, n = 35). B: typical single [Ca2+]i recording of a cell cultured for 10 days in proliferating medium stimulated with 10-3 M ATP. C: typical single [Ca2+]i response to 10-5 M ACh in tracheal myocytes freshly isolated (J0) (n = 61) and cultured for 48 h in non-proliferating medium (n = 26). D: percentage of cells responding to 10-5 M ACh, and amplitude of the [Ca2+]i peak, in freshly isolated myocytes (black columns, n = 61) and in cells cultured for 48 h in non-proliferating medium (open columns, n = 26). P < 0.05 versus responses in freshly isolated cells. Role of intracellular Ca2+ stores and extracellular Ca2+ in ATP-induced response In order to determine the implication of intracellular Ca2+ stores in the response to ATP, we performed the following experiments: in the absence of extracellular Ca2+, rings from rats airways (n = 6 to 8) were exposed to 10-6 M thapsigargin, an irreversible SERCA blocker. Ca2+ release from the SR was triggered by 5 mM caffeine application for 30 min, followed by wash up. Such a protocol ensures the emptiness of the SR, which was verified by the fact that in these conditions, the contractile response to ACh, which has been shown to act via intracellular Ca2+ release from the SR [18], is abolished (data not shown). After caffeine washout, Ca2+ (2 mM) was reintroduced in the extracellular medium. Such a re-introduction did not change the basal tension (data not shown). 10-3 M ATP was then applied to the tissues. As shown in figure 5A, the absence of intracellular Ca2+ did not modify the ATP-induced contraction. Figure 5 Role of intracellular Ca2+ stores and extracellular Ca2+ in ATP-induced response. A: Fmax to 10-3 M ATP in rings from rat trachea (black column, n = 8) left (LEPB) and right (REPB) EPB (hatched columns, n = 8), and left IPB (cross-hatched column, n = 6) after depletion of intracellular Ca2+ stores by application of thapsigargin and caffeine. Horizontal bars are Fmax in control conditions B: Fmax to 10-3 M ATP rings from rat trachea (black column, n = 8) left (LEPB, n = 8) and right (REPB, n = 7) EPB (hatched columns), and left IPB (cross-hatched column, n = 8), and in human IPB (HumIPB, cross-hatched column, n = 5) in the absence of external Ca2+. Horizontal bars are Fmax in control conditions. C: percentage of rat freshly isolated tracheal myocytes responding to 10-3 M ATP, and amplitude of the [Ca2+]i peak, in the presence (black columns, n = 61) and in the absence (grey columns, n = 30) of external Ca2+. Error bars are SEM. P < 0.05. To assess the implication of external Ca2+ influx in the response to ATP, we performed experiments on rat airways (n = 7 to 8) in the absence of extracellular Ca2+. In Ca2+-free KH solution, Fmax was significantly lower than in control conditions, and was below 10% of the ACh reference response, except in IPB where the remaining response, though significantly reduced, was above 20%. Similar experiments were performed on human IPB (n = 5). As in rat, the contractile response was significantly lower, but remained above 25%. Results are summarized in figure 5B. Experiments in the absence of external Ca2+ were also performed on freshly isolated tracheal myocytes (n = 30). Removal of extracellular Ca2+ reduced both the percentage of responding cells to 10-3 M ATP and the amplitude of the [Ca2+]i response in the responding cells, as shown in figure 5C, abolished [Ca2+]i oscillations. Role of L-type Ca2+ channels and extracellular Na+ in ATP-induced contraction Since ATP-induced response appeared to be dependent on extracellular Ca2+, we tested the effect of 10-5 M D600, an inhibitor of the L-type voltage-dependent Ca2+ channels on the contractile response to 10-3 M ATP (n = 7 to 10). As shown in figure 6A, Fmax was significantly reduced in the presence of D600. In a following series of experiments, 10-3 M ATP was applied to the rings in the absence of extracellular Na+. In these conditions, the ATP-induced response was significantly reduced in each type of rings, as shown in figure 6B (n = 7). By contrast, removal of extracellular Na+ did not modify the contractile response to the depolarizing agent KCl (30 mM) (n = 5 to 7), as shown in figure 6C. Figure 6 Effect of D600 and extracellular Na+ removal on ATP-induced response. A: Fmax to 10-3 M ATP in rat airway rings in the presence of 10 μM D600 (n = 7 to 10). B: Fmax to 10-3 M ATP in rat airway rings in the absence of extracellular Na+ (n = 7 to 8). C: Fmax to 30 mM KCl in rat airway rings in the absence of extracellular Na+ (n = 5 to 7). Trachea: black column; left (LEPB) and right EPB (REPB): hatched columns; left IPB: cross-hatched column. Horizontal bars are Fmax in control conditions. Error bars are SEM. P < 0.05. Effect of α-β-methylene ATP and RB2 on ATP-induced contraction In order to determine which type of P2 purinoreceptors was implicated in the contractile response to ATP, we tested the effect of RB2, a P2Y inhibitor, on the ATP-induced contraction and we measured the contractile response to α-β-methylene ATP, a specific agonist of P2X purinoreptors. Incubation with RB2 did not significantly modify the ATP-induced contractile response in extrapulmonary bronchi, but it significantly increased the response of trachea, and reduced that of IPB, (n = 10). RB2 also significantly reduced the contractile response of human IPB (n = 8). Results are shown in figure 7A. α-β-methylene ATP was used at 10-4 M. As with ATP at the same concentration, the α-β-methylene ATP-induced contraction was transient. The amplitude of the contractile response was not different from experiments with ATP in similar conditions in extrapulmonary airways, but was significantly reduced in IPB (figure 7B). TR10 was significantly smaller in extrapulmonary airways, whereas it was not modified in IPB, as shown in figure 7C (n = 7 to 8). Figure 7 Effect of RB2 and α-β-methylene ATP on rat airway rings. A: Fmax to 10-3 M ATP in rat airway rings (n = 8) and human IPB (HumIPB, n = 8) in the presence of 10 μM RB2. B: Fmax to 10-4 M α-β-methylene ATP in rat airway rings (N = 7 to 8). Horizontal bars are Fmax in control conditions. C: TR10 in rat airway rings stimulated with 10-4 M α-β-methylene ATP. Vertical bars are TR10 in control conditions, i.e., 10-4 M ATP. Trachea: black column; left (LEPB) and right EPB (REPB): hatched columns; left IPB: cross-hatched column. Error bars are SEM. P < 0.05. Effect of ATP-γ-S on rat isolated airways In order to evaluate a possible role of ATP degradation in the transience of the response, we assessed the effect of the non-hydrolysable ATP analogous, ATP-γ-S, from 10-7 to 10-4 M. Results are shown in figure 8. ATP-γ-S induced a fast and transient contraction which characteristics did not differ from that of ATP. The CRC were not significantly different from that obtained with ATP and neither was the TR10 (n = 5 to 10). Figure 8 Effect of ATP-γ-S on isolated airway rings. A, B & C: mean ATP-induced (black symbols) and ATP-γ-S-induced (open symbols) non-cumulative response curves in trachea (A, n = 10) left EPB (B, n = 7), and left IPB (C, n = 10) from rat airways. D: TR10 in rat trachea (black column) right (REPB) and left EPB (LEPB) (hatched columns), and left IPB (cross-hatched column) stimulated by 10-4 M ATP-γ-S. Vertical bars are TR10 in control conditions, i.e., 10-4 M ATP. Error bars are SEM. P < 0.05. Effect of indomethacin and H-89 on ATP-induced contraction in rat isolated airways In order to identify a possible implication of arachidonic acid derivatives due to cyclooxygenase activity in the response to ATP stimulation, experiments were performed with 10-5 M indomethacin. Rat tissues were incubated in the presence of indomethacin 30 min before ATP stimulation. The maximal contractile response was not significantly modified (figure 9A). By contrast, the return to baseline was significantly longer in the presence of indomethacin in extrapulmonary airways, but not in IPB (figure 9B). We tested the effect of H-89, an inhibitor of PKA, on the ATP-induced contraction. In the presence of H-89, TR10 was significantly increased in tracheal and extrapulmonary bronchial rings, but was not modified in IPB (figure 9C). Figure 9 Effect of indomethacin and H-89 on ATP-induced contraction in rat isolated airway rings. A: Fmax to 10-3 M ATP in rat airway rings in the presence of 10 μM indomethacin. Horizontal bars are Fmax in control conditions. B: TR10 in rat airway rings stimulated by 10-3 M ATP in the presence of 10 μM indomethacin (n = 5 to 8) C: TR10 in rat airway rings stimulated by 10-3 M ATP in the presence of H-89 (n = 8). Trachea: black column; left (LEPB) and right EPB (REPB): hatched columns; left IPB: cross-hatched column. Vertical bars are TR10 in control conditions. Error bars are SEM. P < 0.05. Effect of successive ATP stimulations In order to assess a possible desensitization of purinoreceptors that may explain the progressive return to baseline following the initial contraction, we performed 4 successive ATP stimulations. 10-3 M ATP was applied for 5 minutes, then washed, and stimulations were performed at 15 minute-intervals. As shown in figure 10C, the maximal responses to successive stimulations were progressively decreased. Figure 10 Effect of successive ATP stimulation in rat isolated airway rings. Fmax in response to 4 successive stimulations by 10-3 M ATP at 15 min-intervals of rat trachea (A, n = 8) left EPB (B, n = 8), and left IPB (C, n = 8). Error bars are SEM. Discussion Our results showed that extracellular ATP induced a concentration-dependent transient contraction of rat and human airways, which both amplitude and mechanisms depend on the location along the airway tree. The ATP-induced response was not modified in the absence of epithelium, and mainly depended on the presence of external Ca2+ and Na+. The response pattern was similar with the non-hydrolysable analogous ATP-γ-S. The fact that extracellular ATP alone induced a transient contractile response in airways is in agreement with previous studies that have evidenced such a response profile in mouse IPB [8] and guinea-pig trachea [19,20], though due to different mechanisms. A biphasic contractile response has also been observed in other smooth muscles, such as vesical smooth muscle [21,22]. However, in rabbit trachea, Aksoy and co-workers failed to evidence any contractile effect of ATP alone in rabbit trachea, whereas, in human isolated bronchi, Finney and co-workers reported a small contractile effect of ATP on small airway preparation [23]. It appears then that the effect of extracellular ATP on airways depends both on the location along the airway tree and the species. The contractile response observed in guinea-pig trachea has been reported, by some authors, to depend on the epithelium and/or related to arachidonic acid derivatives [19,20]. However, in rat airways including in trachea, we failed to evidence a significant involvement of the epithelium or the cyclooxygenase activity in the amplitude of the ATP-induced contractile response. Similarly, Bergner and co-workers concluded that in mouse IPB, ATP did not release sufficient quantities of prostaglandins to influence ATP-induced contraction [8]. The possible implication of epithelium-dependent prostanoid release in the ATP-induced response seems therefore to depend both on species and location alongside the airway tree. Several studies performed on airway myocytes have shown that extracellular ATP induces [Ca2+]i increase [7,8,14-16]. We also found that direct exposure of isolated tracheal myocytes to ATP results in a concentration-dependent [Ca2+]i increase. Comparison of the response to ATP with that to cholinergic stimulation obtained in this study and in previous ones [18] indicates that the Ca2+ response to ATP is smaller than that to ACh. Though the amplitude of the first peak is in the same range with the 2 agonists, the percentage of responding cells, as well as the percentage of oscillating responses and the frequency of oscillations was lower with ATP. This difference in the Ca2+ response pattern explains why the contractile response to ATP is lower than that observed upon cholinergic stimulation. We have demonstrated using both contraction measurements and [Ca2+]i recording in isolated cells that the major source of Ca2+ was extracellular Ca2+ influx, with an additional Ca2+ release from internal stores, mainly in IPB, and, to a lesser degree, in extrapulmonary airways. These results are not in accordance with some previous studies that have shown that the ATP-induced response does not depend on extracellular Ca2+ [8,14]. However, it should be noted that, in swine tracheal smooth muscle cells, the [Ca2+]i response to ATP stimulation appeared to depend on extracellular Ca2+ [16]. These discrepancies may be due to different factors including species specificity. Also, the location along the airway tree may influence the relative participation of external versus internal Ca2+. Though removal of external Ca2+ deeply reduced the contractile response to external airways, contraction of IPB remained significant even in the absence of extracellular Ca2+, a result in partial accordance with that of Bergner and co-workers [8]. Finally, results obtained on isolated cells may also differ between non cultured and cultured cells. Michoud and co-authors worked on cultured, not freshly isolated cells. Our experiments performed in both freshly isolated cells and cells cultured under several conditions indicated that cell culture, even primary culture, may alter not only the [Ca2+]i response to ATP but also to other agonists. This indicates that cell culture, even for short period, may critically modify the mechanisms responsible for Ca2+ homeostasis in airway myocytes. ATP-induced Ca2+ influx is supposed to be due to Ca2+ influx though P2X receptors. Surprisingly, in our study, the ATP-induced Ca2+ response appeared to be dependent on L-type voltage-dependent Ca2+ channels, indicating that [Ca2+]i increase was not due to a direct Ca2+ influx through P2X receptors. However, P2X are not Ca2+ specific and, hence, other cations may enter the cell through them. The fact that removal of extracellular Na+ specifically inhibited the ATP-induced contraction, without altering the contraction elicited by direct depolarization by high extracellular K+ concentration, indicates a functional coupling between ATP-activated channels and voltage-operated channels: Na+ entry through ATP-activated channels may induce membrane depolarization and subsequent opening of voltage-operated channels and Ca2+ influx. Such a coupling has been evidenced in PC-12 cells [24]. Taken together, our results about Ca2+ sources are consistent with the activation of P2X receptors, associated, at least in IPB, with the activation of P2Y receptors. The specific P2X agonist α-β-methylene ATP induced a contractile response similar to that obtained with ATP. Moreover, the P2Y specific antagonist RB2 did not modify the response to ATP, except in IPB. Hence, the pharmacological characterization of the purinoceptors involved in the ATP-induced response seems in good accordance with the determination of the sources of [Ca2+]i implicated in the response. The contraction induced by ATP is transient, with a return to baseline tension in several minutes. Previous studies have suggested that it can be ascribed to the degradation of ATP by ectonucleotidases [8]. Considering the CRC and TR10, return to baseline due to ATP degradation would require 99% ATP degradation in 3 to 6 minutes. Taking into account the size of a rat airway ring and the volume of the organ bath, such an explanation was highly improbable in our experimental conditions. This was confirmed by the fact that the contraction profile induced by ATP-γ-S, a non hydrolysable analogous of ATP, does not differ from ATP response. These results are in partial discordance with that obtained in mouse lung, where the response to ATP-γ-S was more prolonged than that to ATP [8]. However, according to the authors, although more prolonged than that obtained with ATP, the response to ATP-γ-S was transient. Previous studies have shown an relaxant effect of ATP mediated by prostanoid release [25]. Such an effect does not seem to be involved in rat IPB, since the return to baseline was not modified by indomethacin. However, indomethacin did prolong the contractile effect of ATP in extrapulmonary airways, indicating that prostaglandin pathway is partially responsible for the transient contractile effect of ATP. Prostaglandin receptors EP2 have been identified in airway smooth muscle cells and their stimulation activates cAMP production and PKA activation [25,26]. Results obtained in the presence of the PKA inhibitor H-89, which, as indomethacin, significantly prolongs the contractile effect of ATP in trachea and EPB but not in IPB, show that in extrapulmonary airways, the transient contractile effect of ATP depends, at least in part, on PKA activation, probably due to prostaglandin receptor activation. An additional mechanism accounting for the transient contraction is the desensitization of the purinoceptors, since repeated stimulations resulted in a progressive decrease in the intensity of the response both in extra- and intrapulmonary airways. It is known that α-β-methylene ATP has a greater desentizating effect than ATP. The fact that, in trachea and EPB, α-β-methylene ATP-induced return to baseline was quicker than with ATP is in accordance with rapid P2X receptor desensitization in extrapulmonary airways. In IPB, where P2Y receptor activation is effective, the relaxant effect may be due to P2Y receptor desentization, a mechanism already evidenced in vesical smooth muscle [21]. However, in addition to PKA activation and/or receptor desensitization, other mechanisms may contribute to the transience of the ATP-induced contraction. Among them, opening of K+ channels that have been identified as potential targets of purinoceptor activation may repolarize the plasma membrane and hence inhibit voltage-dependent Ca2+ entry. In rat vascular smooth muscle, glibenclamide-sensitive K+ channels have been shown to be implicated in the prolonged phase of ATP-induced vasorelaxation [27], whereas, in colonic smooth muscle cells, ATP appeared to activate Ca2+-dependent K+ channels [28]. Very recently, a delayed ATP-elicited K+ current, Ca2+- and glibenclamide-insensitive, has been identified in smooth muscle cells freshly isolated from rat aorta [29]. If present in ASM cells, these mechanisms may also contribute to the transience of the ATP-induced contraction. Taken together, these results show regional variations in the effect of ATP along the airway tree, in terms of both amplitude of the response and underlying mechanisms. This suggests a segmental difference in the distribution of purinoceptor types and/or subtypes in the airways. On the basis of pharmacological studies, regional variation in P2 receptor expression has also been hypothesized in the pulmonary vasculature [30]. The expression of P2 purinoceptors has been investigated in several smooth muscle types, but few studies have been done in airway smooth muscle. Very recently, Govindaraju and co-workers, using RT-PCR and Western blotting, have identified in cultured human airway smooth muscle cell the expression of P2Y1, P2Y2, P2Y4 and P2Y6 receptor subtypes [31], but the authors did not investigate the possible expression of P2X receptors, whereas mRNA and protein expression of both P2X and P2Y have been evidenced in human vascular smooth muscle, P2X1, P2Y2 and P2Y6 being the predominant subtypes [32]. Data available in airway smooth muscle appear then to be fragmental, and systematic screening of P2 receptor expression along the airways requires further investigation. Conclusion In conclusion, we have shown that ATP has a transient contractile effect on human and rat airways, depending on the location along the airway tree. Based on our results in rat airways, we proposed the following mechanism for the effect of ATP on airways (figure 11): ATP acts directly on airway myocytes. Opening of P2X receptors triggers external Na+ entry that depolarizes the plasma membrane and activates L-type voltage-operated Ca2+ channels. The subsequent Ca2+ influx is responsible for contraction. In IPB, in addition to these mechanisms, ATP acts on P2Y receptors and induces Ca2+ release from intracellular Ca2+ stores. The transient effect of ATP is not due to ATP degradation but can be attributed, as least partially, to purinoceptor desensitization and, in extrapulmonary airways, to PKA activation due to epithelium-independent prostaglandin release. Experiments in human IPB, though not as extensive as those performed in rat IPB, suggest that similar mechanisms are involved in human IPB. Figure 11 Mechanisms of action of extracellular ATP on airway myocytes. ATP opens P2X receptors, which triggers external Na+ entry that depolarizes the plasma membrane and activates L-type voltage-operated Ca2+ channels. The subsequent [Ca2+]i rises activates the contractile apparatus. In addition to these mechanisms, ATP acts on P2Y receptors and induces Ca2+ release from SR via protein Gq and PLC activation, mainly in IPB. The progressive return to baseline following the initial contraction is due to desensitization of the purinergic receptors associated, in extrapulmonary airways, with epithelium-independent PG. PG binds to EP receptor coupled to protein Gs and AC and hence induces the production of cAMP, which inhibits the contractile apparatus via PKA activation. List of abbreviations ACh: Acetylcholine AC: Adenylcyclase ASM: Airway Smooth Muscle ATP: Adenosine triphosphate [Ca2+]i: cytosolic Ca2+ concentration cAMP: Cyclic adenosine monophosphate CRC: Concentration-Response Curve CLS: Collagenase DMEM: Dulbecco's modified Eagle's medium D600: Methoxyverapamil EDTA: Ethylene diamine tetra-acetic acid EGTA: Ethylene glycol tetra-acetic acid EPB: Extrapulmonary bronchi Fmax: Maximal apparent contraction IPB: Intrapulmonary bronchi FEV: Forced expiratory volume ITS medium: Insulin, transferrin and selenite medium Indo-1 AM: Indo-1 acetoxymethylester KH: Krebs-Henseleit PLC: Phospholipase C PKA: Protein kinase A PSS: Physiological saline solution PG: Prostaglandin RB2: Reactive blue 2 SERCA: SarcoEndoplasmic Reticulum Ca2+ ATPase SR: Sarcoplasmic Reticulum TR10: time needed for the tension value to decrease to 10% Fmax Competing interests The author(s) declare that they have no competing interests. Authors' contributions BM carried out the contractile experiments and [Ca2+]i recording on isolated cells, participated in the analysis of the data, and helped the draft of the manuscript. RM participated in the design of the study and helped the draft of the manuscript. ER conceived the study, participated in its design, helped in [Ca2+]i recordings, performed statistical analysis and drafted the manuscript. Acknowledgements The authors thank Dr. Patrick Berger, M.D., PhD, Associate Professor of Physiology, Dr. Hughes Begueret, M.D., Ph. D. Staff Specialist of Histology, and the "Service de Chirurgie Thoracique", C.H.U. de Bordeaux, France, for the supply of human tissues, and Ms. Huguette Crevel and Mr. Pierre Téchoueyres for technical assistance. ==== Refs Novak I ATP as a signaling molecule: the exocrine focus News Physiol Sci 2003 18 12 17 12531926 Guyot A Hanrahan JW ATP release from human airway epithelial cells studied using a capillary cell culture system J Physiol 2002 545 199 206 12433960 10.1113/jphysiol.2002.030148 Ahmad S Ahmad A McConville G Schneider BK Allen CB Manzer R Mason RJ White CW Lung epithelial cells release ATP during ozone exposure: Signaling for cell survival Free Radic Biol Med 2005 39 213 226 15964513 10.1016/j.freeradbiomed.2005.03.009 Rich PB Douillet CD Mahler SA Husain SA Boucher RC Adenosine triphosphate is released during injurious mechanical ventilation and contributes to lung edema J Trauma 2003 55 290 297 12913640 Rice WR Effects of extracellular ATP on surfactant secretion Ann N Y Acad Sci 1990 603 64 74; discussion 75 2291538 Son M Ito Y Sato S Ishikawa T Kondo M Nakayama S Shimokata K Kume H Apical and basolateral ATP-induced anion secretion in polarized human airway epithelia Am J Respir Cell Mol Biol 2004 30 411 419 12972400 10.1165/rcmb.2003-0183OC Michoud MC Napolitano G Maghni K Govindaraju V Cogo A Martin JG Effects of extracellular triphosphate nucleotides and nucleosides on airway smooth muscle cell proliferation Am J Respir Cell Mol Biol 2002 27 732 738 12444033 Bergner A Sanderson MJ ATP stimulates Ca2+ oscillations and contraction in airway smooth muscle cells of mouse lung slices Am J Physiol Lung Cell Mol Physiol 2002 283 L1271 9 12388370 Fredholm BB Abbracchio MP Burnstock G Dubyak GR Harden TK Jacobson KA Schwabe U Williams M Towards a revised nomenclature for P1 and P2 receptors Trends Pharmacol Sci 1997 18 79 82 9133776 10.1016/S0165-6147(96)01038-3 Ralevic V Burnstock G Receptors for purines and pyrimidines Pharmacol Rev 1998 50 413 492 9755289 Flezar M Olivenstein R Cantin A Heisler S Extracellular ATP stimulates elastase secretion from human neutrophils and increases lung resistance and secretion from normal rat airways after intratracheal instillation Can J Physiol Pharmacol 1992 70 1065 1068 1451029 Aksoy MO Kelsen SG Relaxation of rabbit tracheal smooth muscle by adenine nucleotides: mediation by P2-purinoceptors Am J Respir Cell Mol Biol 1994 10 230 236 8110478 Piper AS Hollingsworth M ATP and beta,gamma-methylene ATP produce relaxation of guinea-pig isolated trachealis muscle via actions at P1 purinoceptors Eur J Pharmacol 1996 307 183 189 8832220 10.1016/0014-2999(96)00237-3 Michoud MC Tolloczko B Martin JG Effects of purine nucleotides and nucleoside on cytosolic calcium levels in rat tracheal smooth muscle cells Am J Respir Cell Mol Biol 1997 16 199 205 9032128 Sawai H Wang R Yamashita T Kokubun S Effects of purinoceptor agonists on cytosolic Ca2+ concentration in swine tracheal smooth muscle cells in culture Br J Pharmacol 1996 119 539 544 8894175 Sawai H Wang R Yamashita T Kokubun S The effect of extracellular Ca2+ on responses to purinoceptor agonists in cultured swine tracheal smooth muscle cells J Smooth Muscle Res 1997 33 89 98 9533820 Roux E Hyvelin JM Savineau JP Marthan R Human isolated airway contraction: interaction between air pollutants and passive sensitization Am J Respir Crit Care Med 1999 160 439 445 10430711 Roux E Guibert C Savineau JP Marthan R [Ca2+]i oscillations induced by muscarinic stimulation in airway smooth muscle cells: receptor subtypes and correlation with the mechanical activity Br J Pharmacol 1997 120 1294 1301 9105705 Advenier C Bidet D Floch-Saint-Aubin A Renier A Contribution of prostaglandins and thromboxanes to the adenosine and ATP-induced contraction of guinea-pig isolated trachea Br J Pharmacol 1982 77 39 44 7126994 Fedan JS Stem JL Day B Contraction of the guinea pig isolated, perfused trachea to purine and pyrimidine agonists J Pharmacol Exp Ther 1994 268 1321 1327 8138946 Bolego C Pinna C Abbracchio MP Cattabeni F Puglisi L The biphasic response of rat vesical smooth muscle to ATP Br J Pharmacol 1995 114 1557 1562 7599923 McMurray G Dass N Brading AF Purinoceptor subtypes mediating contraction and relaxation of marmoset urinary bladder smooth muscle Br J Pharmacol 1998 123 1579 1586 9605564 10.1038/sj.bjp.0701774 Finney MJ Karlsson JA Persson CG Effects of bronchoconstrictors and bronchodilators on a novel human small airway preparation Br J Pharmacol 1985 85 29 36 2411319 Hur EM Park TJ Kim KT Coupling of L-type voltage-sensitive calcium channels to P2X(2) purinoceptors in PC-12 cells Am J Physiol Cell Physiol 2001 280 C1121 9 11287325 Fortner CN Breyer RM Paul RJ EP2 receptors mediate airway relaxation to substance P, ATP, and PGE2 Am J Physiol Lung Cell Mol Physiol 2001 281 L469 74 11435222 Clarke DL Belvisi MG Catley MC Yacoub MH Newton R Giembycz MA Identification in human airways smooth muscle cells of the prostanoid receptor and signalling pathway through which PGE2 inhibits the release of GM-CSF Br J Pharmacol 2004 141 1141 1150 15023863 10.1038/sj.bjp.0705716 Ralevic V Mechanism of prolonged vasorelaxation to ATP in the rat isolated mesenteric arterial bed Br J Pharmacol 2001 132 685 692 11159721 10.1038/sj.bjp.0703868 Koh SD Dick GM Sanders KM Small-conductance Ca(2+)-dependent K+ channels activated by ATP in murine colonic smooth muscle Am J Physiol 1997 273 C2010 21 9435508 Serir K Hayoz S Fanchaouy M Beny JL Bychkov R A delayed ATP-elicited K(+) current in freshly isolated smooth muscle cells from mouse aorta Br J Pharmacol 2005 in press 16258525 Chootip K Ness KF Wang Y Gurney AM Kennedy C Regional variation in P2 receptor expression in the rat pulmonary arterial circulation Br J Pharmacol 2002 137 637 646 12381677 10.1038/sj.bjp.0704915 Govindaraju V Martin JG Maghni K Ferraro P Michoud MC The effects of extracellular purines and pyrimidines on human airway smooth muscle cells J Pharmacol Exp Ther 2005 315 941 948 16099930 10.1124/jpet.105.089698 Wang L Karlsson L Moses S Hultgardh-Nilsson A Andersson M Borna C Gudbjartsson T Jern S Erlinge D P2 receptor expression profiles in human vascular smooth muscle and endothelial cells J Cardiovasc Pharmacol 2002 40 841 853 12451317 10.1097/00005344-200212000-00005	16336659	PMC1325041	CC BY	2021-01-04 16:23:24	no	Respir Res. 2005 Dec 8; 6(1):143	utf-8	Respir Res	2,005	10.1186/1465-9921-6-143	oa_comm
==== Front CytojournalCytoJournal1742-6413BioMed Central London 1742-6413-2-201633667410.1186/1742-6413-2-20CommentaryPrimary fallopian tubal transitional cell carcinoma with exfoliation of malignant cells in cervical Pap smear Gupta Nalini [email protected] Radhika [email protected] Raje [email protected] Lakhbir Kaur [email protected] Department of Cytology and Gynecological Pathology, Post Graduate Institute of Medical Education and Research, Chandigarh, India2 Department of Cytology and Gynecological Pathology, Post Graduate Institute of Medical Education and Research, Chandigarh, India3 Department of Cytology and Gynecological Pathology, Post Graduate Institute of Medical Education and Research, Chandigarh, India4 Department of Obstetrics and Gynecology, Post Graduate Institute of Medical Education and Research, Chandigarh, India2005 9 12 2005 2 20 20 19 6 2005 9 12 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. ==== Body Only 0.2 – 0.5% of primary female genital malignancies are tubal, and histologically most of these are adenocarcinomas. Primary transitional cell carcinoma (TCC) accounts for about 10% of primary tubal carcinomas. [1] Primary TCC of the fallopian tube with exfoliation of malignant cells in cervical Pap smear has not been described in the literature previously. A 52-year- old lady presented with episodic spotting per vaginum. On general physical examination, the only significant finding was a 1.5 cm, firm, mobile, left supraclavicular lymph node. Fine needle aspiration (FNAC) from the left supraclavicular lymph node showed a metastatic carcinoma. The patient was investigated for detection of the primary malignancy. She was referred to the gynecologist, who took a cervical Pap smear. The Pap smear revealed mainly clusters as well as scattered cells showing moderate pleomorphism (Figure 1). The cells had moderate amount of cytoplasm, coarsely clumped granular chromatin and inconspicuous nucleoli. It was reported as positive for squamous cell carcinoma by the cytopathologist, who also advised colposcopy and biopsy confirmation. Colposcopical examination of the cervix was performed which showed no significant abnormality. A cervical biopsy and endocervical curettage were done. On microscopic examination, the endocervical curettage showed an occasional cluster of malignant cells entangled in mucus. The endocervical curettage was reported as suspicious of malignancy. The cervical biopsy was unremarkable. A repeat Pap smear also showed malignant cells as described previously. In view of the persistently positive Pap smear report, a total abdominal hysterectomy with bilateral salpingo-oophrectomy was carried out and the specimen was sent for histopathological examination. Figure 1 Clusters of malignant cells showing moderate pleomorphism in cervical Pap smear (Papanicolaou stain, #215; 250) Grossly, the cervix, endo- myometrium, bilateral ovaries and left fallopian tube were unremarkable. The right fallopian tube showed a nodule measuring about 2.5 × 2 × 2 cm occluding its fimbrial end. The cut section of this nodule was gray-white. Microscopically, a tumor was identified in the right fallopian tube. In one focus, a clump of tumor cells was seen in its lumen. The tumor showed large areas of necrosis. The tumor cells were arranged predominantly in ribbons, trabeculae and sheets (Figure 2). Histologicaly, this tumor pattern was that of a transitional cell carcinoma. The cells had "coffee bean" like nuclei and prominent nucleoli. There was no evidence of keratinisation. Mitosis was increased and the tumor was seen infiltrating transmurally. In addition, a small serosal tumor deposit was seen in the uterus. Figure 2 Microphotograph showing a tumor in the lumen of fallopian tube (H&E, 55). Primary transitional cell carcinoma (TCC) of the fallopian tube is rare. The typical signs and symptoms of invasive tubal carcinoma include vaginal bleeding, clear or serosanguinous vaginal discharge, pelvic pain and a pelvic mass. The cervical Pap smear has shown exfoliated malignant cells rarely in cases of adenocarcinoma of fallopian tube. [1-5] Grossly, the tubal lumen is usually filled and dilated by papillary or solid and necrotic tumor. Tumor at the fimbriated end, with ready access to the peritoneal cavity, may also warrant individual staging. The morphology of TCC in fallopian tube is similar to TCC of urinary bladder. To the best of our knowledge, there are less than twenty cases reported of primary TCC in the fallopian tube in the English literature[6-8]. Primary TCC of fallopian tube showing exfoliated malignant cells in cervical Pap smear has not been described in the literature previously. Therefore, if the cervical Pap smear is positive for malignant cells and the cervical biopsy is negative, the patient should be investigated for a malignancy higher up in the gynecological tract. If endometrial curettage also does not reveal malignancy, the possibility of a tubal malignancy must be excluded by appropriate investigations. Figure 3 Microphotograph showing tumor cells arranged in ribbons and trabeculae with intervening areas of necrosis (Inset, H& E 250). ==== Refs Kim JW Cho EM Kim YT Han JH A case of primary transitional cell carcinoma of the fallopian tube J Obstet Gynaecol Res 1999 25 321 6 10533326 Johnston GA Jr Primary malignancy of the fallopian tube: a clinical review of 13 cases J Surg Oncol 1983 24 304 9 6656258 Safret A Bosch B Bannwart F Rinderknecht B Hafner HU Carcinoma in situ of the fallopian tube presenting as a positive Pap smear Acta Cytologica 2004 48 462 4 15192974 Rahimpanah F Reid R Fallopian tube carcinoma detected by ThinPrep cytology smear Med J Aust 2000 172 38 10682018 Warshal DP Burgelson ER Aikins JK Rocereto TF Post- hysterectomy fallopian tube carcinoma presenting with a positive Papanicolaou smear Obstet Gynecol 1999 94 834 6 10546748 10.1016/S0029-7844(99)00241-0 Koshiyama M Konishi I Yoshida M Wang DP Mandai M Mori T Fujii S Transitinal cell carcinoma of the fallopian tube: a light and electron microscopic study Int J Gynecol Pathol 1994 13 175 80 8005739 Takeuchi S Hirano H Ichio T Taniguchi H Toyoda N A case report: rare case of primary transitional cell carcinoma of the fallopian tube J Obstet Gynaecol Res 1999 25 29 32 10067010 Rabczynski J Kochman A Hudziec P Primary transitional cell carcinoma of the fallopian tube Przegl Lek 1998 55 572 5 10216369	16336674	PMC1325042	CC BY	2021-01-04 16:24:26	no	Cytojournal. 2005 Dec 9; 2:20	utf-8	Cytojournal	2,005	10.1186/1742-6413-2-20	oa_comm
==== Front BMC Ear Nose Throat DisordBMC Ear, Nose and Throat Disorders1472-6815BioMed Central London 1472-6815-5-111633666210.1186/1472-6815-5-11Research ArticleC.35delG/ GJB2 and del(GJB6-D13S1830) mutations in Croatians with prelingual non-syndromic hearing impairment Medica Igor [email protected] Gorazd [email protected] Manuela [email protected] Borut [email protected] Division of Medical Genetics, Department of Obstetrics and Gynaecology, University Medical Centre Ljubljana, Slovenia, Slajmerjeva 3 1000 Ljubljana, Slovenia2 Outpatient Paediatric Clinic Pula, Croatia, Istarska 13 52100 Pula, Croatia2005 8 12 2005 5 11 11 26 9 2005 8 12 2005 Copyright © 2005 Medica 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 C.35delG/GJB2 mutation is the most frequent genetic cause of deafness in Caucasians. Another frequent mutation in some Caucasian populations is del(GJB6-D13S1830). Both GJB2 and GJB6 genes belong to the same DFNB1 locus and when the two mutations are found in combination in a hearing-impaired person, a digenic pattern of inheritance is suggested. Methods We examined 63 Croatian subjects (25 familial and 38 sporadic cases) with prelingual non-syndromic hearing impairment by polymerase chain reaction for the presence of the c.35delG/GJB2 and the del(GJB6-D13S1830) mutations. Results Of the 63 unrelated hearing-impaired subjects, the mutation c.35delG/GJB2 was found in 21 subjects (33.3%). In 5 of them the mutation was found in the heterozygous state, all of them being compound heterozygotes, as sequencing revealed a second mutation within the coding region of the gene in 3 subjects, and a splice site mutation in 2 subjects. The del(GJB6-D13S1830) mutation was not found in the investigated hearing-impaired Croatian subjects. Conclusion Our results contribute to the knowledge of geographic distribution and population genetics of the GJB2 and GJB6 mutations in the Europeans. ==== Body Background The identification of genes causing non-syndromic hearing impairment has partially resolved the puzzle of clinical and genetic heterogeneity of deafness [1]. Among these genes the gene with the most significant impact on the population genetics and genetic counselling is the GJB2 gene with the mutation c.35delG that accounts for the majority of mutations in deaf Caucasians [1-3]. Studies published so far have reported the differences in frequency of the mutation in different populations, and its variability in clinical impact on hearing impairment [4,5]. However, although the mutations in the GJB2 gene are responsible for up to 69% of autosomal recessive non-syndromic hearing impairment [2], a problem emerges when patients are identified with only one GJB2 mutant allele. Recently, in the same DFNB1 locus, a 309-kb deletion implicating the GJB6 gene, del(GJB6-D13S1830), has been identified, and was found to be very common in non-syndromic hearing-impaired patients from Spain, France, Israel, the United Kingdom and Brazil, suggesting also a possible GJB2 / GJB6 digenic pattern of inheritance of deafness [6,7]. There are no reports on frequencies of the c.35delG/GJB2 and del(GJB6-D13S1830) mutation in either the hearing impaired in Croatia, or in the Slavic populations of Central-Eastern Europe. The aim of our study was to establish the frequencies of c.35delG/GJB2 and del(GJB6-D13S1830) mutations in Croatians with prelingual non-syndromic hearing impairment. Methods Patients with hearing impairment were ascertained through two regional otorhinolaryngology centres (General Hospital Pula and University Medical Centre Rijeka). The inclusion criteria were: prelingual, non-syndromic hearing loss of unknown aetiology. Both the patients with familial history of deafness and sporadic cases were included. The study group consisted of 66 patients: 29 males and 37 females, with ages ranging from 3 months to 73 years; 28 cases had familial history of deafness, 38 were sporadic. Of the 66 subjects, 63 were unrelated. Fifty adult patients were audiologically evaluated by pure tone audiometry with a diagnostic audiometer in a soundproof room according to the International Organization for Standardization (ISO). The conductive component of hearing loss was excluded by tympanometry. The air conduction binaural mean pure tone threshold (dB hearing loss) was recorded for 500, 1000, 2000, and 4000 Hz. Hearing impairment was classified as mild when the average threshold was between 21 and 40 db, moderate if 41–70 db, severe if 71–95 db, and profound if more than 95 db. In 16 children, aged from 3 months to 16 years, the testing was performed by auditory brainstem response [8,9]. After DNA isolation from the peripheral blood by standard procedures, all patients were molecularly evaluated for the presence of the c.35delG mutation by allele-specific polymerase chain reaction method (PCR) [10]. Sequencing was performed in patients heterozygous for the mutation at GENDIA – Genetic Diagnostic Network, Antwerp, Belgium. All patients were also evaluated for the presence of del(GJB6-D13S1830) mutation by the PCR method described by del Castillo et al [7]. The study was approved by the institutional medical ethics committee, and was performed after the informed consent of the patients or their parents was obtained. Results Mutation analysis Of the 63 unrelated investigated patients, the c.35delG/GJB2 mutation was found in 21 subjects (33.3%): 12 males and 9 females. In 16 out of 63 subjects the mutation was found in the homozygous state, in 5 in the heterozygous state. In the latter cases, sequencing of the coding exon, in search for the second mutation was performed. In three patients the sequencing revealed, besides the c.35delG mutation, three different second mutations already known as pathogenic: c.235delC, p.E47X and p.R143W respectively. In remaining two patients the sequencing revealed a known splice site mutation which is also already disease-associated: c.IVS1+1G>A. Twelve homozygous and 1 heterozygous patients out of the 21 had a positive familial history of deafness, 4 homozygous and 4 heterozygotes had no familial history of deafness. In the group of 25 subjects with positive familial history, 12 were homozygous for the c35delG mutation (48%), 1 heterozygous (4%), while in the group of 38 sporadic cases 4 were homozygous (12.5%) and 4 heterozygous (12.5%). None of the 63 patients was found to carry the mutation del(GJB6-D13S1830) in either the homozygous or heterozygous state. Phenotypic analysis When considering all 66 investigated patients, the c.35delG/GJB2 mutation was found in 24: 13 males and 11 females. In 19 out of 66 subjects the mutation was found in the homozygous state, in 5 in the heterozygous state. Of the 19 patients with hearing impairment due to the c.35delG/GJB2 mutation, 16 had profound deafness, one had severe, and two had moderate hearing loss. All patients had sensorineural bilateral hearing loss. All but 2 patients had symmetric hearing impairment. The 3 patients – compound heterozygotes: c. [35delG] + [235delC], c. [35delG] + p. [E47X], c. [35delG] + p. [R143W] had also profound deafness while 2 patients compound heterozygotes c. [35delG] + [IVS1+1G>A] had severe and moderate deafness respectively. Discussion The c.35delG mutation in the GJB2 gene has been identified as the most important among genetic causes of hearing impairment as it accounts for the majority of mutations in deaf Caucasians, whether in the homozygous state or as one of the mutations in compound heterozygosity (within GJB2 or in GJB6 gene) [4,7]. The studies have been published reporting the frequency of the mutation in different populations showing a south-to-north European gradient, where a high prevalence of the mutation in the populations of Southern Europe has been explained by the founder effect [4,5,11,12]. The share of the c.35delG involvement in hearing impairment varies from 28 to 63% [4]. There are only a few reports on its frequency in the populations of Slavic origin [13,14], but no mutation frequency data have been published on the Croatian or other Slavic Central-Eastern European populations. The del(GJB6-D13S1830) mutation has been identified as the second most frequent mutation causing non-syndromic, prelingual, autosomal recessive hearing impairment in the European populations of Spain, France, the United Kingdom, but also in Brazilians and Jews [7]. Like the c.35delG mutation, this deletion is also population specific as it was established that it is very rare in Czech population, while it is absent in the Austrian population [15,16], showing a west-to-east European gradient and indicating again a founder effect as it is the case of c.35delG/GJB2 mutation. There are no reports on the share of this mutation in hearing impaired Slavic Central-Eastern European populations. Our results suggest the importance of the mutation c.35delG in the Croatian population, the data being comparable to those from the neighbouring areas [4,17,18]. Besides, results such as ours suggest the introduction of molecular genetics diagnosis of the mutation into clinical practice with significant impact on genetic counselling. However, although the involvement of the mutation in the aetiology of deafness has been found in 21 out of 63 hearing-impaired Croatian subjects, a possibility should be mentioned that some GJB2-related cases of deafness due to two less prevalent GJB2 mutations should escape detection in our investigation. In our study, 16 of the 19 subjects with the c.35delG mutation involvement had profound deafness, one had severe and two had moderate hearing impairment. The results are in agreement with the majority of studies on the severity of hearing impairment related to the mutation, where mainly the patients with severe to profound hearing loss are reported, due to an important truncation of the protein after deletion [19-21]. Patients with mild and moderate hearing loss have been rarely reported [22,23]. Our data on the c.35delG/GJB2 frequency in Croatia contribute to the geographic distribution of the mutation in Europe, confirming the south-to-north gradient. On the other hand, our results on the involvement of del(GJB6-D13S1830) mutation in the aetiology of hearing impairment show that the mutation is of no epidemiological and clinical importance in Croatians, the results being in agreement with the results of the frequency of the mutation in populations from neighbouring areas, i.e. from Central European populations [14-16]. Competing interests The author(s) declare that they have no competing interests. Authors' contributions IM recruited patients, performed data interpretation, and wrote the manuscript. MB and GR carried out the molecular genetic studies and participated in data interpretation. BP designed, coordinated, and supervised the study and reviewed the completed manuscript. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements The study was performed with financial support of the Health Department of the Istria County, Croatia, and the Ministry of Education, Science and Sport, Republic of Slovenia, Grant no. J3-2370. The authors wish to thank Ms M. Pirc and Mrs P. Duff for revising the English text and Mrs S. Piljan Pauletic for technical assistance. ==== Refs Hereditary Hearing Loss Denoyelle F Weil D Maw MA Wilcox SA Lench NJ Allen-Powell DR Osborn AH Dahl HH Middleton A Houseman MJ Dode C Marlin S Boulila-ElGaied A Grati M Ayadi H BenArab S Bitoun P Lina-Granade G Godet J Mustapha M Loiselet J El-Zir E Aubois A Joannard A Levilliers J Garabedian EN Mueller RF McKinlay Gardner RJ Petit C Prelingual deafness: high prevalence of a 30delG mutation in the connexin 26 gene Hum Mol Genet 1997 6 2173 2177 9336442 10.1093/hmg/6.12.2173 Zelante L Gasparini P Estivill X Melchionda S D'Agruma L Govea N Mila M Monica MD Lutfi J Shohat M Mansfield E Delgrosso K Rappaport E Surrey S Fortina P Connexin 26 mutations associated with the most common form of non-syndromic neurosensory autosomal recessive deafness (DFNB1) in Mediterraneans Hum Mol Genet 1997 6 1605 1609 9285800 10.1093/hmg/6.9.1605 Gasparini P Rabionet R Barbujani G Melchionda S Petersen M Brondum-Nielsen K Metspalu A Oitmaa E Pisano M Fortina P Zelante L Estivill X Genetic Analysis Consortium of GJB2 35delG. High carrier frequency of the 35delG deafness mutation in European populations Eur J Hum Genet 2000 8 19 23 10713883 10.1038/sj.ejhg.5200406 Kenneson A Van Naarden Braun K Boyle C GJB2 (connexin 26) variants and nonsyndromic sensorineural hearing loss: a HuGE review Genet Med 2002 4 258 274 12172392 del Castillo I Villamar M Moreno-Pelayo MA del Castillo FJ Alvarez A Telleria D Menendez I Moreno F A deletion involving the connexin 30 gene in nonsyndromic hearing impairment N Engl J Med 346 243 249 11807148 10.1056/NEJMoa012052 del Castillo I Moreno-Pelayo MA Del Castillo FJ Brownstein Z Marlin S Adina Q Cockburn DJ Pandya A Siemering KR Chamberlin GP Ballana E Wuyts W Maciel-Guerra AT Alvarez A Villamar M Shohat M Abeliovich D Dahl HH Estivill X Gasparini P Hutchin T Nance WE Sartorato EL Smith RJ Van Camp G Avraham KB Petit C Prevalence and evolutionary origins of the del (GJB6-D13S1830) mutation in the DFNB1 locus in hearing-impaired subjects: a multicenter study Am J Hum Genet 2003 73 1452 1458 14571368 10.1086/380205 International Organization for Standardization n. ISO 8253-1: Acoustics: audiometric test methods, I: basic pure tone air and bone conduction threshold audiometry 1989 Geneva: International Organization for Standardization GENDEAF European Thematic Network on Genetic Deafness Scott DA Kraft ML Carmi R Ramesh A Elbedour K Yairi Y Srisailapathy CR Rosengren SS Markham AF Mueller RF Lench NJ Van Camp G Smith RJ Identification of mutations in the connexin 26 gene that cause autosomal recessive nonsyndromic hearing loss Hum Mutat 1998 11 387 394 9600457 10.1002/(SICI)1098-1004(1998)11:5<387::AID-HUMU6>3.0.CO;2-8 Rothrock CR Murgia A Sartorato EL Leonardi E Wei S Lebeis SL YU LE Elfenbein JL Fisher RA Friderici KH Connexin 26 35delG does not represent a mutational hotspot Hum Genet 2003 113 18 23 12684873 Van Laer L Coucke P Mueller RF Caethoven G Flothmann K Prasad SD Chamberlin GP Houseman M Taylor GR Van de Heyning CM Fransen E Rowland J Cucci RA Smith RJ Van Camp G A common founder for the 35delG GJB2 mutation in connexion 26 hearing impairment J Med Genet 2001 38 515 518 11483639 10.1136/jmg.38.8.515 Seeman P Malikova M Raskova D Bendova O Groh D Kubalkova M Sakmaryova I Seemanova E Kabelka Z Spectrum and frequencies of mutations in the GJB2 (Cx26) gene among 156 Czech patients with pre-lingual deafness Clin Genet 2004 66 152 157 15253766 10.1111/j.1399-0004.2004.00283.x Markova TG Megrelishvilli SM Zaitseva NG Shagina IA Poliakova AV DNA diagnosis in congenital and early childhood hypoacusis and deafness Vestn Otorinolaringol 2002 6 12 15 12501766 Seeman P Bendova O Raskova D Malikova M Groh D Kabelka Z Double heterozygosity with mutations involving both the GJB2 and GJB6 genes is possible, but very rare, cause of congenital deafness in the Czech population Ann Hum Genet 2005 69 9 14 15638823 10.1046/j.1529-8817.2003.00120.x Günther B Steiner A Nekahm-Heis D Albegger K Zorowka P Utermann G Janecke A The 342-kb deletion in GJB6 is not present in patients with non-syndromic hearing loss from Austria Hum Mutat 2003 22 180 10.1002/humu.9167 Löffler J Nekahm D Hirst-Stadlmann A Günther B Menzel HJ Utermann G Janecke AR Sensorineural hearing loss and the incidence of Cx26 mutations in Austria Eur J Hum Genet 2001 9 226 230 11313763 10.1038/sj.ejhg.5200607 Frei K Szuhai K Lucas T Weipoltshammer K Schöfer C Ramsebner R Baumgartner WD Raap AK Bittner R Wachtler FJ Kirschhofer K Connexin 26 mutations in cases of sensorineural deafness in eastern Austria Eur J Hum Genet 2002 10 427 432 12107817 10.1038/sj.ejhg.5200826 Cryns K Orzan E Murgia A Huygen PLM Moreno F del Castillo I Parker Chamberlin G Azaiez H Prasad S Cucci RA Leonardi E Snoeckx RL Govaerts PJ Van de Heyning PH Van de Heyning CM Smith RJ Van Camp G A genotype-phenotype correlation for GJB2 (connexin26) deafness J Med Genet 2004 41 147 154 14985372 10.1136/jmg.2003.013896 Azaiez H Chamberlin GP Fischer SM Welp CL Prasad SD Taggart RT del Castillo I Van Camp G Smith RJ GJB2: the spectrum of deafness-causing allele variants and their phenotype Hum Mutat 2004 24 305 311 15365987 10.1002/humu.20084 Gualandi F Ravani A Berto A Sensi A Trabanelli C Falciano F Trevisi P Mazzoli M Tibiletti MG Cristofari E Burdo S Ferlini A Martini A Calzolari E Exploring the clinical and epidemiological complexity of GJB2-linked deafness Am J Med Genet 2002 112 38 45 12239718 10.1002/ajmg.10621 Cohn ES Kelley PM Fowler TW Gorga MP Lefkowitz DM Kuehn HJ Schaefer GB Gobar LS Hahn FJ Harris DJ Kimberling WJ Clinical studies of families with hearing loss attributable to mutations in the connexin26 gene (GJB2/DFNB1) Pediatrics 1999 103 546 550 10049954 10.1542/peds.103.3.546 Denoyelle F Marlin S Weil D Moatti L Chauvin P Garabedian EN Petit C Clinical features of the prevalent form of childhood deafness, DFNB1, due to a connexin-26 gene defect: implications for genetic counselling Lancet 1999 353 1298 1303 10218527 10.1016/S0140-6736(98)11071-1	16336662	PMC1325043	CC BY	2021-01-04 16:03:26	no	BMC Ear Nose Throat Disord. 2005 Dec 8; 5:11	utf-8	BMC Ear Nose Throat Disord	2,005	10.1186/1472-6815-5-11	oa_comm
==== Front Thromb JThrombosis Journal1477-9560BioMed Central London 1477-9560-3-201633667010.1186/1477-9560-3-20Original Clinical InvestigationAnticoagulant treatment at a specialized outpatient anticoagulant therapy unit, a descriptive study Ekblom Kim [email protected] Johan [email protected] Bo [email protected] Tage [email protected] Department of Medical Biosciences, Clinical Chemistry, Umeå University Hospital, 901 85 Umeå, Sweden2 Department of Public Health and Clinical Medicine, Medicine, Umeå University Hospital, 901 85 Umeå, Sweden2005 8 12 2005 3 20 20 10 7 2005 8 12 2005 Copyright © 2005 Ekblom et al; licensee BioMed Central Ltd.2005Ekblom 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 indications for continuous oral anticoagulant treatment, the target interval and the procedures for withdrawing treatment have changed in the last 10 years. Methods Patients on continuous oral anticoagulant treatment at the Outpatient Anticoagulant Clinic at Umeå University Hospital in 2002 were included in a descriptive study (n = 900). 263 of those had a mechanical heart valve prosthesis. Only patient records for patients with other indications than mechanical heart valve prosthesis were examined. 582 of those records were found. In the 55 remaining patients some clinical information was retrieved from the computerised warfarin dosage database. These latter, more unsure clinical data, are presented separately. Anticoagulant treatment was discontinued if lack of proper indication or presence of too high risk for hemorrhagic complications were found. Results The prevalence of continuous oral anticoagulant treatment in the uptake area was 0.65%. The most common target interval was INR 2.1–3.0, but patients with a mechanical heart valve prosthesis were often treated more aggressively, i.e. with a higher INR target interval. Of the patients on continuous treatment, 26.6% of the INR values were outside 2.0–3.0. The most common reasons for oral anticoagulant treatment were atrial fibrillation or mechanical heart valve prosthesis, in contrast to earlier findings in studies of our population in 1987 and 1990. We found 90 patients (10.0%) without proper indication for oral anticoagulant treatment or too high risk, and their treatment was discontinued. Conclusion In patients on oral anticoagulant therapy, re-evaluation of indications and risks resulted in a substantial number of treatment withdrawals. There have been major changes in treatment indications during the last decade, possibly due to rapid development of knowledge in the field of thrombosis risk factors. Treatment should be re-considered once a year. ==== Body Background Treatment with warfarin or other coumarin derivatives is an established method of secondary prevention after venous or arterial thromboembolic events, as well as for primary prevention. In 1997 the prevalence of oral anticoagulant treatment in Sweden was 0.8% [1]. That figure is comparable with those in Denmark, were the prevalence was 0.663% in 1997 and 0.784% in 1999 [2]. In 2002 19,823 kDDD (thousand defined daily doses) of Waran® 2.5 mg tablets (sodium warfarin) were sold in Sweden. In 2003 the number rose to 21,387 kDDD. In 1998 Apekumarol® (dicoumarol) was withdrawn from the Swedish market. Marcoumar® (phenprocoumon) can be prescribed on licence when warfarin cannot be used [3]. Nearly all patients now receive sodium warfarin in Sweden. Oral anticoagulant treatment is a potentially dangerous medication. According to a study in Mölndal, Sweden, 4.5 per 100 patient years resulted in a serious bleeding complication leading to hospitalisation or persisting sequele. 0.5 per 100 patient years resulted in a patient dying of a bleeding complication [4]. Because of the possibility of serious, and sometimes lethal, complications it is necessary to ensure that each patient has a valid indication for continuous anticoagulant treatment. Every patient should be re-evaluated yearly according to the Swedish National Board of Health and Welfare. In 1987–1990 a study compared the quality of oral anticoagulant treatment in the primary health care and in the Department of Internal Medicine at Umeå University hospital [5]. In 1990, 290 patients were treated in primary health care and 175 via the Department of Internal Medicine. Nowadays most patients in Umeå and the surrounding municipalities are treated via the specialized Outpatient Anticoagulant Clinic within the Department of Internal Medicine at Umeå University Hospital. Only a few patients are managed through primary health care centres. The aims of this study were to describe patients on oral anticoagulant therapy in our outpatient clinic, to compare the results with a previous investigation, and re-evaluate indications and risks of complications in these patients. Methods The cohort consisted of all patients receiving continuous oral anticoagulant therapy with either warfarin or coumarin derivatives on October 15, 2002, registered at the Outpatient Anticoagulant Clinic at the University Hospital of Umeå. It is the only hospital in the reception area. A small number of patients treated by general practitioners (n = 57) were not included in the study. In most cases monitoring by general practitioners is due to long distances in our catchment area. Due to the same reason, it may be estimated that the number of patients treated by other hospitals is negligible. During the time of the study, only one patient living in our area used a self-monitoring device, and he was supervised by the Outpatient Anticoagulant Clinic, and thus is included in our study. All patients in the register with a treatment time defined as "indefinite" were considered to be on continuous oral anticoagulant therapy. The patient records from relevant clinics were examined. Personal data, start date of anticoagulant treatment, and all thromboembolic events in the patient's history that came to our knowledge were recorded. If previously unknown events were found in personal communication with the patient, these events were also recorded. For all patients without mechanical heart valve prosthesis, the reason for the ongoing anticoagulant treatment was recorded. Mechanical valve prosthesis, dilated cardiomyopathia, or venous thrombosis, in combination with one of the following coagulopathies, were defined as absolute indications for continuous warfarin treatment: antithrombin deficiency, protein C deficiency, protein S deficiency, homozygote for F V Leiden (FV 1691 AA), homozygote for the prothrombin mutation (FII 20210 AA), heterozygote for both APC-resistance (FV 1691 GA) and prothrombin mutation (FII 20210 GA). If no problems with the treatment were registered by the nurses at the Outpatient Anticoagulant Clinic, these patients were not subjected to any further investigation in this study. Patients with atrial fibrillation were divided into two groups with low or high risk for cardioembolic events. A patient with atrial fibrillation at low risk was defined one who had none of the following: age more than 60 years, previous cerebrovascular event, dilated cardiomyopathia, mitral stenosis, marked mitral insufficiency, or marked enlargement of the left atrium. All other patients with atrial fibrillation were defined as high risk. High risk was considered a legitimate reason for continuous anticoagulant treatment as long as the risk for bleeding complications was considered low. Low risk patients were invited for a medical examination, and discontinuation of anticoagulant treatment was considered, with a possible switch to other medications. Patients with recurrent venous thromboses (three or more) without known risk factors were offered a medical examination and the presence of coagulopathy was examined. Anticoagulant treatment of these patients was continued if there were no contraindications. All other legitimate indications were considered as relative indications, and they were weighed against the risk of bleeding complications. Risk factors and possible contraindications were: documented serious bleeding complications, as well as factors increasing risk of bleeding such as problems with compliance, fluctuating INR, balance problem with documented slip or fall accidents, dementia, liver failure and high age (>85 years of age). Patients with unclear indication or with known risk factors were invited for further investigation. The indication for treatment was re-evaluated after laboratory tests and medical examination was performed. If no valid indication was found, or if the risk was found to be too high, anticoagulant treatment was discontinued or replaced with other appropriate medication whenever possible. INR was determined with STA® – SPA 50 kit (Diagnostica Stago, Asnieres-sur-Seine, France) on a Sysmex® CA-6000 automatic coagulation instrument (Sysmex Corporation, Kobe, Japan). In mid-November INR values for the subjects were recorded. The first INR value dated on or after October 15, 2002 was recorded. In absence of such a value, the first INR value before this date was used. All values, except one, were determined within two months before this date. Demographic data was collected from the SCB, Statistics Sweden website, , accessed January 6, 2005. Information about warfarin use in Sweden was retrieved from the database PharmX supplied by Läkemedelsinformation AB. Data was collected using a data sheet produced with SPSS Data Entry Builder 3.0, and SPSS™ version 12.0 (SPSS inc, Chicago, IL, USA) was used for statistical analysis. For comparison between groups, one-way ANOVA with Bonferoni post-hoc testing was used. A p-value <0.05 was considered statistically significant. All statistical tests and corresponding p-values were two-sided. Ethical permission was granted by the local ethical committee of Umeå University. Results We found 900 patients on continuous oral anticoagulant medication at the Oral Anticoagulant Clinic. On September 30, 2002 the total number of patients, both on short and long term treatment, registered in the outpatient clinic was 998. In addition, there were 57 patients monitored by general practitioners. On November 1, 2002, the population in the uptake area was 138 240 persons. Therefore, the total prevalence of oral anticoagulant treatment in the district was approximately 0.76%. In the study by Jansson et al 1987–1990, the two most common indications for anticoagulant treatment were valvular heart disease (including mechanical valve prosthesis) and arterial thromboembolism [5]. In our study the most common indication for treatment was atrial fibrillation followed by mechanical heart valve prosthesis. Arterial thromboembolism is no longer one of the main indications (Table 1). Table 1 Indications for warfarin treatment in 1987, 1990 (both continuous and temporary) and 2002 (continuous) at the Department of Internal Medicine and primary health care units in the Umeå district. Data extracted from Jansson et al. [5]. Only one indication was recorded. Hierarchical order 1987 and 1990: pulmonary embolism, deep venous thrombosis, ischaemic cerebrovascular disease, peripheral thromboembolism, valvular heart disease, atrial fibrillation and miscellaneous. Hierarchical order 2002: valvular heart disease, atrial fibrillation, arterial thromboembolism, venous thromboembolism and miscellaneous. When a patient had both arterial and venous events, the most recent event decided the group. Patients with missing records are not included in the table. Indications 1987 1990 2002 n (%) n (%) n (%) Valvular heart disease 103 (42) 119 (41) 314 (37) Arterial thromboembolism 70 (29) 82 (28) 11 (1) Atrial fibrillation 18 (7) 21 (7) 339 (40) Venous thromboembolism 47 (19) 52 (18) 60 (7) Miscellaneous 5 (2) 16 (6) 120 (14) Total 243 (100) 290 (100) 845 (100) Mechanical heart valve prosthesis was present in 263 patients. Patient records of these patients were not examined, leaving 637 records to be found. We managed to find 582 patient records. Some indications and the mean target interval for these patients are presented in Table 2. Patient records of 55 subjects could not be retrieved. For these subjects some information was found through the warfarin dosage system, treatment indications of there are presented in Table 3. After evaluation of indications and contraindications, treatment was discontinued in 90 of the 582 patients due to dubious reasons for treatment or too high risk. Basic facts on these patients are presented in Table 4. Of these 90, the treatment of 23 patients was discontinued before the patient was invited for examination and by a physician who was not involved in this study (Table 5). All discontinuations, except one case that had atrial fibrillation, were in accordance with the method of re-evaluation. In four cases without a valid indication the medication was continued either because the patient or the patient's physician strongly opposed the discontinuation of warfarin treatment. Table 2 Indications for oral anticoagulant treatment, age of patient, duration of treatment, mean target and, actual INR October 15, 2002. For mean target and INR, 10th and 90th percentiles are presented. All groups are compared with the Mechanical heart valve prosthesis group. Indication n (%) Age (years) Duration of treatment (years) Mean target (10–90%) INR mean (10–90%) Mechanical heart valve prosthesis 263 (28.4) 68.3 7.43 2.59(2.55–2.85) 2.68(2.10–3.40) Atrial fibrillation 399(44.3) 75.4 a 3.92 a 2.51 (2.40–2.55) a 2.54(1.90–3.20) c Other 183(20.3) 65.3 n.s. 5.20 a 2.50 (2.25–2.55) a 2.50(1.90–3.20) b Record missing d 55 (6.9) 70.9 n.s. 4.67 a 2.50(2.15–2.55) a 2.45 (1.83–3.10) n.s. All indications 900 (100) 71.0 5.22 2.53 (2.40–2.55) 2.57 (1.90–3.20) ap < 0.001, bp < 0.01, cp < 0.05, n.snonsignificant dcf. Table 3 Table 3 Treatment indications for patients with missing patient record, data retrieved from warfarin dosage database Indication n % Atrial fibrillation 30 55 Thrombosis NUD 11 20 Valvular disease NUD 5 9 Cerebrovascular disease 3 5 Biological heart valve prosthesis 3 5 Myocardial infarction 2 4 Cardiomyopathia 1 2 Total 55 100 Table 4 Basic facts on patients whose treatment was continued vs. those whose treatment was discontinued. Continued (n = 492) Discontinued (n = 90) Mean (SD) Min Max Mean (SD) Min Max p Age on October 15, 2002 (years) 72.4 (11.0) 20.4 94.7 71.2 (15.1) 20.6 92.6 0.370 Age at start of treatment (years) 68.2 (11.8) 17.2 89.1 66.1 (16.1) 13.7 89.7 0.149 Duration of treatment (years) 4.2 (3.6) 0.0 23.5 5.1 (4.5) 0.1 19.0 0.044 Mean target (INR) 2.5 (0.1) 1.7 3.2 2.5 (0.1) 1.7 3.2 0.249 Table 5 Reasons for discontinuation of oral anticoagulant treatment. Reason n % Lack of indication 46 51 Too high risk 21 23 Discontinued by physician outside the study 23 26 Total 90 100 Of all INR values recorded, using a cross-section method, 11.4% were below 2.0 and 15.2% above 3.0 in all long-term treated patients. For the subgroups continuously treated patients with mechanical valve prosthesis, atrial fibrillation, other indications and those with record missing the percentages were 7.0 and 22.7, 12.5 and 12.5, 13.7 and 12.0, and 16.1 and 11.3, respectively. Using a cumulative method for the period from January 1, to December 31, 2002, the percentages of patients in the total group, including short-term treatments that were outside the individual target intervals were higher, as well as the percentages outside INR 2.0–3.0 (Tables 6 and 7). Table 6 Number of INR samples during 2002 under, within and over the patient's personal target interval 2002 Under personal interval n (%) Within personal interval n (%) Over personal interval n (%) January-March 1274 (20.3) 3863 (61.7) 1125 (18.0) April-July 1257 (20.7) 3791 (62.3) 1036 (17.0) August-September 1299 (21.0) 3894 (62.8) 1005 (16.2) October-December 1273 (20.5) 397 (63.8) 977 (15.7) Total 5103 (20.6) 15521 (62.7) 4143 (16.7) Table 7 Number of INR samples during 2002 <2.0, 2.0–3.0 and <3.0 2002 INR <2.0 n (%) INR 2.0–3.0 n (%) INR >3.0 n (%) January-March 1032 (16.5) 4093 (65.4) 1137 (18.2) April-July 1026 (16.9) 4040 (66.4) 1018 (16.7) August-September 1033 (16.7) 4185 (67.5) 980 (15.8) October-December 1047 (16.8) 4232 (68.0) 944 (15.2) Total 4138 (16.7) 16550 (67.2) 4079 (16.5) Discussion The total prevalence of 0.76% oral anticoagulant treatment in the reception area of the oral anticoagulant outpatient clinic was slightly lower than that reported, around 0.8% reported in Sweden in general [1] and in Denmark [2]. The population in the uptake area is young as compared with the rest of Sweden, 33% of the population was under 25 years of age, and 67% of the population was less than 50 years of age, as compared with 30% and 63% respectively, in the total Swedish population. We have not found any studies on continuous oral anticoagulant treatment to compare with our results. A high number of patients with dubious indications and/or unacceptably high risk for treatment were found. The reason for this may be the fear of discontinuing anticoagulant medication. Those patients or their physicians may prefer the risk of a bleeding complication to the risk for a thromboembolic event. In addition, valid indications for oral anticoagulant treatment have changed, and the significance of coagulopathies may have been under- or overestimated in the past. This may also be the case for other indications. It takes time for new information to be implemented in everyday medical practice. In some cases, the patient or physician prefer to continue treatment despite lack of indication; this reflects some of the difficulties in routine oral anticoagulant monitoring. A possible explanation for the high number of patients on questionable oral anticoagulant treatment may be that yearly re-evaluation is not always done due to lack of resources. Sometimes it may not be obvious who is in charge of the patient's oral anticoagulant treatment. In many cases the re-evaluation is made by the patient's primary health care doctor who doesn't always have access to the reasons for initiating the treatment or information about possible risk factors. The change in indications as compared with the study done in 1987–1990 is striking (Table 1). Arterial thrombosis is no longer one of the main indications for oral anticoagulant therapy. A series of clinical trials that began in the mid-1980s provided substantial evidence for the effectiveness of warfarin in prevention of stroke in patients with atrial fibrillation [6-11]. The change may reflect the increasing awareness of this fact. The most common target treatment range in this study was INR 2.1–3.0. After the completion of the study the target range had been changed to 2.0–3.0 in most of these cases. In a study by Samsa et al, 45.0–66.8% of the INR values in atrial fibrillation patients were outside the interval 2.0–3.0 [12]. In our study only 33.2% of all INR values during 2002 were outside this interval. In 1987 and 1990 Jansson et al found that less than 20% of the INR values at the oral anticoagulant clinic at Umeå University Hospital were outside the desired treatment interval 2.1–4.2 during that time period [5]. That was a much broader interval than we use nowadays. They also excluded patients whose observation time was too short, i.e., less than five INR values. We did not exclude any patients in our study due to short observation time. We also had several patients with actual target intervals different from the most common target interval 2.1–3.0. We therefore believe that the treatment quality in our study was at least as good as it was in the study by Jansson et al [5]. In that study PT ratio was used, but comparison is possible because our laboratory was responsible for all the analyses, using the same method principle (Owren), with the same reagents thus enabling us to calculate the INR values properly. There has been consensus on the target range 2.0–3.0 in atrial fibrillation and also in most of the other indications [13,14]. Some patients, especially those with mechanical heart valve prostheses, have a higher target INR. This was seen especially among those whose treatment started a long time ago (data not shown). Some elderly people have lower target treatment range than INR 2.1–3.0. There is still a debate about the benefits and disadvantages of low-dose warfarin treatment after deep venous thrombosis [15-17]. In some cases a lowering of the treatment range might be an optimal alternative. In others it may be just an excuse for not discontinuing oral anticoagulant treatment. More studies are needed on this issue. This study had some limitations: The prevalence of anticoagulant treatment may have been slightly underestimated since a few patients are treated through primary health care centres. However these patients are very few (n = 57). Complete patient records could not be found for 55 patients (8.6%) in the study, not counting those with mechanical heart valve prostheses. However, treatment indications of these patients were similar to those with records found. It can not be excluded that some bias of the results may be caused by the patients with missing records. The number of missing records will be lower when computerised medical records are introduced. However, in studies based on medical records, drop out rates higher than we found are common. The difficulty of retrieving medical records in a medical setting has been described [18]. In that study up to 30% of the patient records were unavailable at the time of the medical consultation. Some thromboembolic events in the past may have escaped our attention because they had been treated at another hospital. There is a problem with the oldest of the old: The risk of thromboembolic events rises with age. So do the problems with maintaining the treatment within the target interval, and the risk of bleeding complications [19,20]. Attempts to prospectively assess risk factors in warfarin treatment have been made [21]. In medical practice, risk assessment is often subjective in the absence of accurate, objective, easy-to-use protocols. The balance between the risk for a thromboembolic event and the risk for a bleeding complication in this group is still unclear and further studies need to be done on this issue. Thrombin inhibitors are a new group of drugs on the market. The fear of liver complications when using these drugs may prevent the wider usage of these drugs for many years, leaving anti vitamin K drugs an important option for secondary thrombosis prevention [22]. Conclusion In patients on oral anticoagulant therapy, re-evaluation of indications and risks resulted in a substantial number of treatment withdrawals. Treatment should be re-considered once a year, due to the rapid development of knowledge in the field of thrombosis risk factors. The indications for oral anticoagulant treatment have changed: Atrial fibrillation is now the most common indication while arterial thrombosis is no longer a common reason. Competing interests The author(s) declare that they have no competing interests. Authors' contributions KE, JH, BC and TS designed the study. KE examined the patient records and the patients. KE and TS made the clinical decisions on the patients. KE and JH analysed the data. KE, JH, BC and TS wrote the paper. Acknowledgements A special thanks to the nurses Ann-Marie Jonsson and Sofia Långström, and the archivists Sonja Gustavsson and Karin Appelvik at the Anticoagulant Outpatient Clinic at Umeå University Hospital for their help and assistance. ==== Refs Johnsson H [Anticoagulant therapy can be safer. An active follow-up and written care programs reduce the number of complications] Lakartidningen 1999 96 3388 3390 10479787 Holm T Lassen JF [How many patients are on oral anticoagulant therapy in Denmark? Methods to estimate the number] Ugeskr Laeger 2003 165 1871 1875 12772397 Stigendal L Johnsson H Apekumarol ersätts nu med waran, exempel på patientinformation Information från Läkemedelsverket 1998 9 Medical Products Agency Taghavi A Jonson T Stockelberg D [Survey of complications following treatment with anticoagulants. A computerized search for hemorrhagic complications completes manual reporting] Lakartidningen 1999 96 3421 3424 10479797 Jansson JH Westman G Boman K Nilsson T Norberg B Oral anticoagulant treatment in a medical care district--a descriptive study Scand J Prim Health Care 1995 13 268 274 8693211 The effect of low-dose warfarin on the risk of stroke in patients with nonrheumatic atrial fibrillation. The Boston Area Anticoagulation Trial for Atrial Fibrillation Investigators N Engl J Med 1990 323 1505 1511 2233931 Stroke Prevention in Atrial Fibrillation Study. Final results Circulation 1991 84 527 539 1860198 Secondary prevention in non-rheumatic atrial fibrillation after transient ischaemic attack or minor stroke. EAFT (European Atrial Fibrillation Trial) Study Group Lancet 1993 342 1255 1262 7901582 Connolly SJ Laupacis A Gent M Roberts RS Cairns JA Joyner C Canadian Atrial Fibrillation Anticoagulation (CAFA) Study J Am Coll Cardiol 1991 18 349 355 1856403 Ezekowitz MD Bridgers SL James KE Carliner NH Colling CL Gornick CC Krause-Steinrauf H Kurtzke JF Nazarian SM Radford MJ Warfarin in the prevention of stroke associated with nonrheumatic atrial fibrillation. Veterans Affairs Stroke Prevention in Nonrheumatic Atrial Fibrillation Investigators N Engl J Med 1992 327 1406 1412 1406859 Petersen P Boysen G Godtfredsen J Andersen ED Andersen B Placebo-controlled, randomised trial of warfarin and aspirin for prevention of thromboembolic complications in chronic atrial fibrillation. The Copenhagen AFASAK study Lancet 1989 1 175 179 2563096 10.1016/S0140-6736(89)91200-2 Samsa GP Matchar DB Goldstein LB Bonito AJ Lux LJ Witter DM Bian J Quality of anticoagulation management among patients with atrial fibrillation: results of a review of medical records from 2 communities Arch Intern Med 2000 160 967 973 10761962 10.1001/archinte.160.7.967 Albers GW Dalen JE Laupacis A Manning WJ Petersen P Singer DE Antithrombotic therapy in atrial fibrillation Chest 2001 119 194S 206S 11157649 10.1378/chest.119.1_suppl.194S Deitcher SR Carman TL Deep Venous Thrombosis and Pulmonary Embolism Curr Treat Options Cardiovasc Med 2002 4 223 238 12003721 Kearon C Ginsberg JS Kovacs MJ Anderson DR Wells P Julian JA MacKinnon B Weitz JI Crowther MA Dolan S Turpie AG Geerts W Solymoss S van Nguyen P Demers C Kahn SR Kassis J Rodger M Hambleton J Gent M Comparison of low-intensity warfarin therapy with conventional-intensity warfarin therapy for long-term prevention of recurrent venous thromboembolism N Engl J Med 2003 349 631 639 12917299 10.1056/NEJMoa035422 Ridker PM Long-term low-dose warfarin use is effective in the prevention of recurrent venous thromboembolism: yes J Thromb Haemost 2004 2 1034 1037 15219181 10.1111/j.1538-7836.2004.00813.x Ridker PM Goldhaber SZ Danielson E Rosenberg Y Eby CS Deitcher SR Cushman M Moll S Kessler CM Elliott CG Paulson R Wong T Bauer KA Schwartz BA Miletich JP Bounameaux H Glynn RJ Long-term, low-intensity warfarin therapy for the prevention of recurrent venous thromboembolism N Engl J Med 2003 348 1425 1434 12601075 10.1056/NEJMoa035029 Tufo HM Speidel JJ Problems with medical records Med Care 1971 9 509 517 5138986 Fihn SD Callahan CM Martin DC McDonell MB Henikoff JG White RH The risk for and severity of bleeding complications in elderly patients treated with warfarin. The National Consortium of Anticoagulation Clinics Ann Intern Med 1996 124 970 979 8624064 Froom P Miron E Barak M Oral anticoagulants in the elderly Br J Haematol 2003 120 526 528 12580973 10.1046/j.1365-2141.2003.04110.x Beyth RJ Quinn LM Landefeld CS Prospective evaluation of an index for predicting the risk of major bleeding in outpatients treated with warfarin Am J Med 1998 105 91 99 9727814 10.1016/S0002-9343(98)00198-3 Thompson CA Ximelagatran data fail to impress FDA Am J Health Syst Pharm 2004 61 2472, 2474 5, 2480 15595219	16336670	PMC1325044	CC BY	2021-01-04 16:36:24	no	Thromb J. 2005 Dec 8; 3:20	utf-8	Thromb J	2,005	10.1186/1477-9560-3-20	oa_comm
==== Front Reprod Biol EndocrinolReproductive Biology and Endocrinology1477-7827BioMed Central London 1477-7827-3-681633226210.1186/1477-7827-3-68ResearchCloning and expression of two new prolactin-related proteins, prolactin-related protein-VIII and -IX, in bovine placenta Ushizawa Koichi [email protected] Toru [email protected] Misa [email protected] Kanako [email protected] Kazuyoshi [email protected] Reproductive Biology and Technology Laboratory, Developmental Biology Department, National Institute of Agrobiological Sciences, 2 Ikenodai, Tsukuba, Ibaraki 305-8602, Japan2 Department of Veterinary Medicine, Faculty of Agriculture, Iwate University, 3-18-8 Ueda, Morioka, Iwate 020-8550, Japan3 Department of Technology, National Livestock Breeding Center, 1 Odakurahara, Odakura, Nishigo, Fukushima 961-8511, Japan2005 7 12 2005 3 68 68 7 10 2005 7 12 2005 Copyright © 2005 Ushizawa et al; licensee BioMed Central Ltd.2005Ushizawa 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 Prolactin-related proteins (PRPs) are specific proteins of the growth hormone/prolactin (GH/PRL) family in bovine placenta. This study reports the identification and sequencing of a full-length cDNA for two new members of bovine PRPs, bPRP-VIII and -IX, and their localization and quantitative expression in bovine placenta. Methods New bPRP-VIII and -IX were identified from bovine placentome. Localization and quantitative gene expression in the placenta were respectively investigated by in situ hybridization and real-time RT-PCR methods. Recombinant proteins of these genes were produced by a mammalian HEK293 cell expression system. Results Full-length bPRP-VIII and -IX cDNA were respectively cloned with 909 and 910 nucleotide open-reading-frames corresponding to proteins of 236 and 238 amino acids. The predicted bPRP-VIII amino acid sequence shared about 40 to 70% homology with other bPRPs, and bPRP-IX had about 50 to 80 % homology of others. The two new bPRPs were detected only in the placenta by RT-PCR. mRNA was primarily expressed in the cotyledon and intercotyledonary tissues throughout gestation. An in situ hybridization analysis revealed the presence of bPRP-VIII and -IX mRNA in the trophoblastic binucleate and/or trinucleate cells. bPRP-VIII mRNA was observed in the extra-embryonic membrane on Day 27 of gestation, however, no bPRP-IX mRNA was observed in the extra-embryonic membrane in the same stage of pregnancy by quantitative real-time RT-PCR analysis. Both new bPRP genes were possible to translate a mature protein in a mammalian cell expression system with approximately 28 kDa in bPRP-VIII and 38 kDa in bPRP-IX. Conclusion We identified the new members of bovine prolactin-related protein, bPRP-VIII and -IX. Localization and quantitative expression were confirmed in bovine placenta by in situ hybridization or real-time PCR. Their different temporal and spatial expressions suggest a different role for these genes in bovine placenta during gestation. ==== Body Background Prolactin-related proteins (PRPs) are members of the growth hormone/prolactin (GH/PRL) family [1]. PRPs are highly expressed in binucleate cells of bovine trophoblasts, although their function is still obscure. No new member had been identified between early 1990 and this year [2]. We recently found a new PRP member through a comprehensive analysis using a bovine utero-placental cDNA microarray [3]. Seven bovine PRP (bPRP) genes have been identified in placentomal tissues, whereas only two of those genes have been shown to be translated. The expression pattern of these genes is spatially and temporally different. bPRP-I indicates the considerable importance of regulation of implantation and formation of the placentome in bovines [4,5]. However, bPRP-VII was expressed in intercotyledon and cotyledon with rather weak compare to those of bPRP-I during the gestation [6]. Placental lactogens (PLs) are known as classical members and PRPs are categorized as the non-classical members of the GH/PRL family in bovine [7]. Various non-classical members are common in rodents, such as mice and rats, and some of them exhibit known biological functions, such as proliferin and prolactin-like protein-A (PLP-A) [7-12]. However, as a whole the biological function of Prolactin-related genes is still not known in various species including bovine. Two research groups identified bovine PRPs mRNA individually; bPRP-I to -III were identified by the Schuler group [13,14], and bPRP-IV to -VI were identified by the Nakashima group [15-17]. We recently identified another one named as bPRP-VII [6]. Our previous study suggested our bovine placental cDNA library contained other bPRP genes [3]. We reveal here the full-length sequences of two new members of bPRPs, bPRP-VIII and -IX, and their localization and quantitative expression. We confirmed the possibility of translation of them and the accuracy of these gene sequences to produce a recombinant protein using the HEK293 cell-transfecting system. Materials and methods Animals and tissues Placental tissues for cDNA cloning and mRNA expression were collected from Japanese Black cows. The extra-embryonic tissues, placenta, and endometrium were collected at a local slaughterhouse on days 27 to 28, 56 to 64, 144 to 149, and 245 to 258 after artificial insemination (day 1). The tissues were separated into four portions: the cotyledon (COT); intercotyledon (the area between the cotyledonary villous (ICOT)); the caruncle area, including the maternal placentomal septa in the endometrium (CAR); and the intercaruncle area (ICAR). It was difficult to divide the COT and ICOT on days 27 to 28, and thus the COT contained very few villi. Tissues from two different cows on day 27 and one cow on day 28 of gestation (n = 3) were used as Day 27 extra-embryonic membrane (Day 27EEM), Day 27 caruncle (Day 27CAR), and Day 27 intercaruncular endometrium (Day 27ICAR). Placentomal tissues were collected on days 56, 58, and 64 (in total, n = 3) and designated as Day 60COT, CAR, ICOT, and ICAR. Sample materials from days 144, 148, and 149 (n = 3) and days 245 (two samples) and 252 (one sample) were marked as Day 150COT, CAR, ICOT, ICAR, Day 250COT, CAR, ICOT, and ICAR. The cotyledonary and caruncular part was separated mechanically, and each part may contained a counter part of tissue. The collected samples were stored at -80°C until RNA extraction. The placentomes of day 60 were fixed in 3.7% formaldehyde PBS at pH 7.4 and then embedded in paraffin wax and stored at 4°C until in situ hybridization. All procedures for these animal experiments were carried out in accordance with the guidelines and ethics approved by the Animal Ethics Committee of the National Institute of Agrobiological Sciences for the use of animals. Cloning of full-length bPRP-VIII and -IX cDNA The new full-length bPRP-VIII and -IX cDNA were isolated from bovine cotyledonary tissue by the 3'-rapid amplification of cDNA ends (RACE) method. In brief, a complete RNA was isolated from a bovine placentome on day 60 of gestation using ISOGEN (Nippon Gene, Toyama, Japan). We performed 3'-RACE using a 3'-full RACE core set (Takara, Kyoto, Japan) with bPRP-VIII-specific forward primer (5'-CCACAGTCAACAGGAGTCCTCA-3') and bPRP-IX-specific forward primer (5'-CCAACAGAGAGTCCTCACCCTGCGA-3'). The bPRP-VIII and -IX primer was designed from bovine EST accession number AW464912 and BP108069, respectively. The 3'-RACE products were sequenced using an ABI Prism 370 automatic sequencer (Applied Biosystems, Foster City, CA, USA) after cloning into a pGEM-T Easy vector (Promega, Madison, WI, USA). Phylogenetic analysis Alignments of deduced protein sequences were performed with the multiple alignment software Clustal W 1.83 on the DDBJ web site. Clustal W was also employed to calculate trees using the Neighbor-Joining (NJ) method [18]. TreeView was used to display the phylogenetic tree [19]. The values represent bootstrap scores for 1,000 trials, indicating the credibility of each branch. Except the bPRP-VIII and -IX sequences, all the bPRPs and bPLs protein sequences were obtained from GenBank. Their GenBank accession numbers are: bPRP-I (J02944), bPRP-II (M27239), bPRP-III (M27240), bPRP-IV (M33269), bPRP-V (AB239755), bPRP-VI (X59504), bPRP-VII (AB187564), bPL-Ala (J02840), and bPL-Val (M33268). Three-dimensional structure prediction by FAMS We predicted the three-dimensional (3D) structure of bPRP-VIII and -IX by using the FAMS (Fully automated homology modeling system; ) [20]. FAMS is the software which predicts 3D model of the target protein from the structural known protein of high homology. In case of bPRP-VIII and -IX, the 3D structure was constructed based on the human prolactin (hPRL) 3D structure (Protein Data Bank ID: 1N9D) in the element. FAMS program only requires an amino acids sequence as input, and constructs 3D model structures automatically. Visualization of the 3D structure was performed using the RasMol 2.7.3 software [21]. RT-PCR Tissue distribution of bPRP-VIII and -IX expression was studied by RT-PCR. Bovine GAPDH was used as a positive control for the PCR. Details of the RT-PCR method were described in previous reports [6,22]. The total RNA in a total reaction mixture was used for reverse transcription and template cDNA synthesis using oligo(dT) primer and Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA, USA) at 42°C for 50 min. Each PCR contained the cDNA template, primers, deoxynucleotide triphosphate mixture (dNTP), MgCl2, 10 × PCR buffer II, autoclaved milliQ water, and AmpliTaq gold DNA polymerase (Applied Biosystems). Amplification conditions included denaturation at 95°C for 30 sec and extension at 72°C for 1 min. Twenty-seven cycles were performed for all samples. The annealing temperature was set at 60°C for 30 sec. A single denaturation step at 95°C for 10 min before the first PCR cycle and a final extension step at 72°C for 10 min after the last PCR cycle were also performed. The PCR products were analyzed by agarose gel electrophoresis and visualized by ethidium bromide staining. The primers encoding for the bPRP-VIII and -IX sequences were designed using the sequence illustrated in Fig. 1. The primer sequence was selected to include a mismatch for other bPRPs. The designated primers are listed in Table 1. All the primers were commercially synthesized (Espec Oligo Service, Tsukuba, Japan). Figure 1 Nucleotide and deduced amino acid sequences of bPRP-VIII. The arrow indicates the putative primary cleavage site of the signal peptide. The potential N-glycosylation site is underlined with a dotted line. The asterisks indicate the termination codon. The polyadenylation signal is underlined with a solid line. Table 1 Oligonucleotide primers used for RT-PCR Gene Primer Sequence Position bPRP-VIII Forward 5'-AGAACATGCCCGTTCTGCTGT-3' 155–175 (AB196438) Reverse 5'-TTAGCACGTGTTGTTGATTCG-3' 757-737 bPRP-IX Forward 5'-AACTCATGCCCATCCTGTGGT-3' 150–170 (AB204881) Reverse 5'-TTAGCACTTTTTGCGGATTCG-3' 758-738 GAPDH Forward 5'-CCTTCATTGACCTTCACTACATGGTCTA-3' 71–98 (U85042) Reverse 5'-GCTGTAGCCAAATTCATTGTCGTACCA-3' 927-901 In situ hybridization Full-length cDNA of bPRP-VIII and -IX was used as a template for hybridization probe synthesis. Digoxigenin (DIG)-labeled antisense and sense-complementary RNA probes were prepared as described in previous studies [6,23]. The placentomes were sectioned into 7 μm-thick sections for hybridization. In situ hybridization was performed using automated Ventana HX System Discovery with a RiboMapKit and BlueMapKit (Ventana, Tucson, AZ, USA) [6]. Briefly, sections were hybridized with DIG-labeled probes in RiboHybe (Ventana) hybridization solution at 65°C for 6 hours. The sections were washed three times in RiboWash (Ventana) (65°C, 6 min) after hybridization and were fixed in RiboFix (Ventana) (37°C, 10 min). The hybridization signals were then detected with monoclonal-anti-digoxin biotin conjugate (Sigma, Saint Louis, MI, USA). The hybridized glasses were observed after preparation with a Nikon ECLIPSE E800 photomicroscope (Nikon, Tokyo, Japan). Real-time RT-PCR Gene expression of bPRP-VIII and -IX was confirmed quantitatively at each stage of gestation (Days 27FM, 27CAR, 27ICAR, 60COT, 60CAR, 60ICOT, 60ICAR, 150COT, 150CAR, 150ICOT, 150ICAR, 250COT, 250CAR, 250ICOT, and 250ICAR) by real-time RT-PCR analysis. Details of the real-time RT-PCR procedures have been described in previous reports [6,24]. Briefly, fifty ng of the total RNA were reverse transcribed into cDNA for 30 min at 48°C by MultiScribe™ reverse transcriptase with an Oligo dT primer, dNTP mixture, MgCl2 and RNase inhibitor. Primer pairs and oligonucleotide probes labeled with a reporter fluorescent dye at the 5'-end and a quencher fluorescent dye at the 3'-end were designed using the Primer Express computer software program (Applied Biosystems). The primers and probes for each gene are listed in Table 2. The thermal-cycling conditions included initial sample incubation at 50°C for 2 min and at 95°C for 10 min, followed by 40 cycles at 95°C for 15 sec and at 60°C for 1 min. The cycle threshold values (CT) indicate the quantity of the target gene in each sample and were determined in real time using an ABI Prism 7700 sequence detector (Applied Biosystems). The relative difference in the initial amount of each mRNA species (or cDNA) was determined by comparing the CT values. The standard curves for each gene were generated by serial dilution of plasmid containing bPRP-VIII, -IX, or GAPDH cDNA to quantify the mRNA concentrations. The ratio of bPRP-VIII and -IX mRNA to GAPDH mRNA was calculated to adjust for any variations in the RT-PCR reaction. All values are presented as mean ± SD. Statistical analysis was performed using one-way ANOVA followed by the Tukey-Kramer multiple comparison test. Differences were considered significant at P < 0.05. Table 2 Oligonucleotide primers and TaqMan probes used for real-time RT-PCR analysis Gene Primer or TaqMan probe Sequence Position bPRP-VIII Forward 5'-CAAGGGTCCTGCCTGCTG-3' 98–115 (AB196438) Reverse 5'-GGCATGTTCTGCCTTGGC-3' 164–344 Probe 5'-TGCTGCTGGTCATGTCAAATCTGCTCC-3' 117–143 bPRP-IX Forward 5'-ATATGCCCAGGGCAAACTGT-3' 287–306 (AB204881) Reverse 5'-TCGGGAGCATGGAAGGAAT-3' 358-340 Probe 5'-TATCAATGCCACCAACAGCTGCCACA-3' 311–336 GAPDH Forward 5'-AAGGCCATCACCATCTTCCA-3' 178–197 (U85042) Reverse 5'-CCACTACATACTCAGCACCAGCAT-3' 253-230 Probe 5'-AGCGAGATCCTGCCAACATCAAGTGG-3' 200–225 TaqMan probes of bPRP-VIII and -IX have 5'-FAM dye and 3'-TAMRA dye TaqMan probe of GAPDH has 5'-VIC dye and 3'-TAMRA dye Production of recombinant proteins The bPRP-I, -VIII, -IX, and bPL sequences encoding the mature protein region, which combined the FLAG and 6 × His epitope tag sequences, were inserted into the pFLAG-CMV-3 vector (Sigma). The constructed plasmid was transfected into HEK 293 cells using FuGENE 6 (Roche Diagnostics, Basel, Switzerland) for transient transfection. Stably transfected HEK 293 cells were adapted to the suspension culture in a spinner flask using 293 SFM II medium (Invitrogen, Gibco) and cultured in an atmosphere of 5% CO2 in air at 37°C for 3 days. The medium was separated by centrifugation and stored at -30°C. Western blot analysis The 10 μg of proteins from the HEK293 cell conditioned media were loaded on each lane, separated by SDS-PAGE, and electrophoretically transferred onto a polyvinylidene difluoride membrane [25]. Western blotting was performed by the method of Towbin et al. [26]. Briefly, the membrane was blocked in 10% skim milk overnight, incubated with anti-FLAG M2 (Sigma) for 1 h at room temperature, followed by incubation with anti-mouse IgG conjugated with alkaline phosphatase (Sigma) (diluted 1:3000) for 1 h at room temperature. Immunopositive bands were stained using NBT (Bio-Rad, Hercules, CA, USA) and BCIP (Bio-Rad). Results Sequences of bPRP-VIII and -IX cDNA Full-length bPRP-VIII and -IX were cloned from bovine placentome. The 906- and 910-nucleotide sequences were isolated in bPRP-VIII and -IX, respectively (Fig. 1 and 2). The protein sequence regions (CDSs) were composed of 711 nucleotides in bPRP-VIII and 717 nucleotides in bPRP-IX. The 3'-untranslated region contains one AATAAA polyadenylation signal start 20 and 26 bases upstream from the poly (A) addition site in bPRP-VIII and -IX, respectively. The amino acid sequences deduced from full-length bPRP-VIII and bPRP-IX cDNA are amino acids 236 and 238. The homology of predicted amino acid sequences of bPRP-VIII and -IX protein were shown in Fig. 3. The predicted sequence of bPRP-VIII protein was 69% homologous to that of bPRP-VI, 66% homologous to that of bPRP-VII, 61% homologous to that of bPRP-I and -III, 58% homologous to that of bPRP-IV and -V, 57% homologous to that of bPRP-IX, 42% homologous to that of bPRP-II, and 39% homologous to that of bPL-Ala (Fig. 3). The predicted sequence of bPRP-IX protein was 81% homologous to that of bPRP-IV, 76% homologous to that of bPRP-I, 70% homologous to that of bPRP-II, 60% homologous to that of bPRP-VII, 57% homologous to that of bPRP-VI and -VIII, 53% homologous to that of bPRP-III and -V, and 40% homologous to that of bPL-Ala (Fig. 3). In the phylogenetic analysis, it was shown that bPRP-VIII was close to bPRP-III, bPRP-VI, and bPRP-VII sides in the phylogenetic tree and bPRP-IX was close to bPRP-II and bPRP-IV sides in the phylogenetic tree (Fig. 4). The N-terminal regions of the bPRP-VIII and -IX proteins were rich in hydrophobic amino acid residue, which is characteristic of the signal peptide. bPRP-VIII had two consensus sequences for N-glycosylation and Asn-X-Ser/Thr at the positions of 60 to 62 and 233 to 235 (Fig. 1). bPRP-IX also had four consensus sequences for N-glycosylation at the positions of 70 to 72, 92 to 94, 146 to 148, and 160 to 162 (Fig. 2). Another atypical N-glycosylation site, Asn-X-Cys was found in only bPRP-IX at the position of 95 to 97, and this region is identified in bPLs. The TAA stop codon was used in both bPRP-VIII and -IX, and appeared after the sequence TGC, which was present in other bPRPs except for bPRP-VI and bPLs that encode C-terminal cysteine residue [17]. The predicted 3D structures of bPRP-VIII and -IX mature region are shown in Fig. 5. The structural differences of N-glycosylation site, disulfide bond (-S-S-) and each atomic configuration were confirmed. We submitted these sequences to the DNA Data Bank of Japan (DDBJ). The DDBJ/GenBank accession Nos. are AB196438 and AB204881. Figure 2 Nucleotide and deduced amino acid sequences of bPRP-IX. The arrow indicates the putative primary cleavage site of the signal peptide. The potential N-glycosylation site is underlined with a dotted line. The asterisks indicate the termination codon. The polyadenylation signal is underlined with a solid line. Figure 3 Comparison of amino acid sequences of bPRP-VIII and -IX with other members of bPRPs and bPLs. Residue present in both bPRP-VIII and -IX is shown in black boxes. Residue present in only bPRP-VIII is designated in red boxes. Residue present in only bPRP-IX is designated in green boxes. Amino acid sequences were aligned with assistance from Clustal W 1.83 on the DDBJ web site. The arrow indicates the putative primary cleavage site of the signal peptide of bPRP-VII. LD1, 2, 3, and 4 refer to the conserved domain in the prolactin family. Figure 4 Phylogenetic tree of bPRPs and bPLs. The tree was constructed using TreeView following alignment of the protein sequences by the Clustal W 1.83 algorithm. The numbers at the base of each branch division represent bootstrap values after 1000 repeats. Scale bar represents 0.1 amino acid replacements per amino acid site. For GenBank/DDBJ accession numbers, refer to Materials and Methods. Figure 5 The predicted 3D structure of (A) bPRP-VIII and (B) bPRP-IX mature protein. The 3D structure were predicted by FAMS software. bPRP-VIII structure was able to construct from Cys43 to Cys236 amino acid region. bPRP-IX structure was able to construct from Pro40 to Ile234 amino acid region. Disulfide bonds refer to white solid line. Predicted disulfide bonds refer to white dot line. N-GLY refer to potential N-glycosylation site. Expression of bPRP-VIII and -IX mRNA in bovine placenta The distribution of bPRP-VIII and -IX mRNA was examined in various bovine tissues by RT-PCR. Both bPRP mRNAs appeared only in the placental tissue (Fig. 6). Figure 6 Expression of bPRP-VIII and -IX mRNA in bovine tissues. Heart, liver, lung, kidney, and spleen were used by RT-PCR. Cotyledonary tissue at day 148 of gestation was used as a placental sample. GAPDH expression in each tissue is presented as standard data. bPRP-VIII and -IX mRNA localization was determined by in situ hybridization in the bovine placentome on day 60 of gestation (Fig. 7). DIG-labeled bPRP-VIII and -IX anti-sense RNA probes specifically detected the mRNA transcript in the placentome and intercotyledonary membrane. Both bPRPs appeared in the binucleate cells in the cotyledon and intercotyledonary membrane, and bPRP-IX mRNA was also detected in trinucleate cells (Figs. 7D and 7F). No significant signals were detected with sense probes in any genes (Fig. 7B and 7E). Figure 7 Localization of bPRP-VIII and -IX in the bovine placentome on day 60 of gestation. (A, B, C) bPRP-VIII and (D, E, F, G) bPRP-IX mRNA were detected by in situ hybridization. (A, C, D, F, G) DIG-labeled anti-sense cRNA probes were used. (B, E) DIG-labeled sense cRNA probes were used. Seven-microgram sections of bovine placentome were hybridized with each probes. Scale bar = 100 μm on the (A, B, D, E) and 20 μm on the (C, F, G). Quantitative real-time RT-PCR analysis indicated that bPRP-VIII mRNA appeared primarily in the extra-embryonic membrane of Day 27, cotyledon and intercotyledonary membrane of Day 60 to 250 (Fig. 8). The expression intensity increased approximately four-fold in the cotyledon from Day 27 to Day 60, and then slightly decreased until Day 250 (approximately 0.6-fold). The expression intensity increased approximately three-fold in the intercotyledon from Day 27 to Day 60 and was maintained from Day 60 to Day 150. Then it then decreased slightly until Day 250 (approximately 0.7-fold). In the caruncular area, the expression increased by about four to five times by Day 60, and the expression level was maintained to late gestation. However no significant expression was detected in the intercaruncle. bPRP-IX mRNA was also detected in the cotyledonary placentome and intercotyledonary membrane after Day 60 of gestation, but it did not appear in the extra-embryonic membrane on Day 27 of gestation. The mRNA expression intensity increased approximately 1.2-fold from Day 60 to Day 150 in the cotyledon and approximately 1.9-fold the intercotyledon, and then those values were maintained until Day 250. No expression of bPRP-IX was observed in the caruncular area on Day 27, but stable expression intensities were detected from Day 60 to Day 250. These expression intensities were rather strong compared to the intercotyledonary and cotyledonary areas, even though the sources for this gene may be binucleate and trinucleate cells. Figure 8 Quantitative real-time RT-PCR analysis of (A) bPRP-VIII and (B) bPRP-IX mRNA in bovine placenta. The total RNA was extracted from cotyledons containing extra-embryonic membrane (EEM), cotyledonary placenta (COT), intercotyledonary fetal membrane (ICOT), caruncular placenta (CAR), and intercaruncular endometrium (ICAR) on Day 27, Day 60, Day 150, and Day 250 of gestation. The expressions of these mRNAs were normalized to the expression of GAPDH measured in the same RNA preparation. Values are means ± SD. Values with different letters are significantly different (P < 0.05). The recombinant proteins of bPRP-VIII and -IX raised with HEK293 cell were detected by Western blot analysis (Fig. 9). The intense bands of bPRP-VIII with a FLAG epitope tag migrated to approximately 28 kDa molecular weight, while those of bPRP-IX with a FLAG epitope tag migrated to 38 kDa, 34 kDa and 26 kDa (Fig. 9). Figure 9 Production of recombinant bPRP-VIII, bPRP-IX, bPRP-I, and bPL proteins. Conditioned media from HEK 293 cells transiently transfected with each gene were collected, and the proteins (10 μg) were loaded in separate lanes. The proteins were separated by SDS PAGE and specific proteins were detected by Western blot analysis using an anti-FLAG tag. MW Marker: molecular weight marker. NC: Negative control (transfected vector only). Discussion In the present study, we identified two new members of bPRP cDNAs, bPRP-VIII and -IX. Totally nine bPRP genes have been identified at this moment in bovine. bPRP-VIII and -IX cDNA sequences had following features, respectively. The position of the polyadenylation signal (AATAAA) in the bPRP-VIII started 20 bases upstream from the poly(A) additional site. This configuration is specific to bPRP-VIII. The number of bases in bPRP-VIII from the polyadenylation signal to poly(A) site are the fewest in known bPRPs. The position of the polyadenylation signal in the bPRP-IX started 26 bases upstream from the poly(A) additional site. This position coincides with those in bPRP-II and bPRP-IV [14,15]. The signal peptide sequences in the N-terminal regions of both bPRP-VIII and -IX were conserved similarly to those in other bPRPs/bPLs (Fig. 3). The signal cleavage site was predicted to be between the Glu-36 and Arg-37 in bPRP-VIII, and between Glu-36 and Asn-37 in bPRP-IX by homology to the N-terminal regions of the bPRPs/bPLs (Fig. 3). As a result, bPRP-VIII is predicted to be a mature protein composed of 200 amino acids, while bPRP-IX has 202 amino acids. bPRP-VIII lacked two amino acids (positions 94 and 183), which is a sequence characteristic, since other known bPRPs have 202 amino acids. bPRP-VIII had two potential N-glycosylation sites (amino acid portions 60 to 62 and 233 to 235, Figs. 1, 3 and 5). The position of 60 to 62 coincided with those of bPRP-III, -VI, and -VII [6,14,17], however, the second N-glycosylation site was found at 233 to 235 in amino acids. In contrast, bPRP-IX had four potential typical N-glycosylation sites of Asn-X-Ser/Thr (amino acid portions 70 to 72, 92 to 94, 146 to 148, and 160 to 162, Figs. 2, 3 and 5). All these configurations of Asn-X-Ser/Thr coincide with those of bPRP-II and -IV [14,15]. However, it is characteristic that only the bPRP-IX have atypical N-glycosylation site of Asn-X-Cys (amino acid portions 95 to 97). The conserved domains (LD1 to LD4, Fig. 3) were suggested by Yamakawa et al. [15] in other prolactin-related genes. The bPRP-VIII and -IX sequences also revealed these LD domains. Cysteine residues (PRP-VIII positions 96, 213, and 230, and PRP-IX positions 97, 215, and 232) in the LD1 and LD4 domains were conserved in bPRPs/bPLs, except for bPRP-VI. bPRP-VIII had four more cysteines (positions 38, 41, 42, and 236, Fig. 3, a total of seven residues in mature protein), and bPRP-IX had three more cysteines (positions 38, 41, and 238, Fig. 3, a total of six residues in mature protein). These configurations indicate common features in bPRPs/bPLs. In particular, the configuration of the cysteine residue in bPRP-VIII coincided exactly with that of bPRP-III in the mature sequence region. The cysteine residue in bPRP-IX was also the same as that in bPRP-I, -II, and -IV in the mature sequence region. Therefore, both bPRP-VIII and -IX may have three disulfide bonds similar to those of other bPRPs, except for bPRP-VII [6]. The primary mRNA expression of bPRP-VIII and -IX was observed in the binucleate cells (Fig. 7). Both genes could independently produce mature recombinant proteins in the mammalian cell expression system (Fig. 9). The binucleate cells may have produced the bPRP-VIII and -IX proteins simultaneously because bPRP-VIII and -IX mRNA expressed in the binucleate cells, and HEK293 cells translated these mRNA to each protein individually. Binucleate cells are also primary expression cells for bPRP-I and bPL [27,28] and may have a specific function for implantation in the fetomaternal interface [11,29]. bPRP-VIII mRNA was also expressed in the extra-embryonic membrane just after the implantation period in the present study (Day 27 of gestation) (Fig. 8). bPRP-VIII may also be related to the implantation process like bPRP-I and -VII since bPRP-I was expressed during early gestation. Compare to these expression, no bPRP-IX was present in the extra-embryonic membrane on Day 27 of gestation (Fig. 8). This mRNA expression pattern implies that bPRP-IX has a different role from bPRP-I and bPL in placental formation [6]. bPRP-IX mRNA was also detected in the trinucleate cells. A trinucleate cell is generated by migration of a binucleate cell to an endometrial epithelial cell with the progress of the pregnancy [30]. bPRP-IX may be a necessary molecule during middle to late gestation. Although both bPRP-VIII and -IX may be necessary molecules for gestation, in situ hybridization demonstrated that both genes were primarily expressed in binucleate and trinucleate cells. This means both genes derive from fetal side origin, but both were stably contained in the caruncular area throughout gestation, except bPRP-IX in early gestation. There are two potential causes of this result. One is contamination after separation. A second possibility is cell migration [30]. The present study and our previous study suggest that binucleate and trinucleate cells may migrate deeply into the endometrium after implantation and maintain their functions, which may be related to immunomodulation or stress during bovine gestation [6,7,30,31]. The molecular size of the recombinant proteins completely differed between bPRP-VIII and -IX. Transcripted bPRP-VIII protein is of relatively smaller size in the known bPRP family members, and the bPRP-IX protein is among the largest members [6,32,33]. The sugar chain addition in post-translational modification will give large effect in the difference of both molecular weights. It was found that the predicted 3D structure differed clearly in both molecules, because configuration of disulfide bonds and sugar chain addition dynamics are different. The tertiary structure of bPRP-VIII and -IX are similar to that of hPRL, so they may share the same receptor. However there is no information and the other evidence may not support this speculation. Only the bPRP-I gene has been known to produce protein in the placenta, however, the protein bound to the alpha2-macroglobulin [32], so it might have a biological function in paracrine status. Since all other bPRPs have a similar characteristic, it is difficult to say bPRPs are hormones secreted into circulation from the placenta, such as bPRL and bPL. It was possible to predict a part of their functions from the molecular weight, 3D structure, and post-translational modification. bPRP-VIII and -IX proteins may have different functions, since their temporal and special expressions were different. Various difficulties remain for understanding PRP functions, however, many of the same genes and molecules are expressed in bovine placenta as in rodent placenta [7]. Recently, bovine genome projects have addressed various shotgun sequence data and five known bPRPs, bPL, and bPRL, a total of seven prolactin related genes, hit in the NCBI genome database. They are located in Bos taurus chromosomes 12 and/or 23. bPRP-III, -VII and bPL were detected on chromosome 23 with bPRL. On the other hand, bPRP-I, -V, and -VI are placed on chromosome 12. Rodents such as mice and rats also have various PLPs [7,34], and they have been clustered on one chromosome, number 13 in mice and number 17 in rats [34,35]. bPRP-III and -VII were in comparatively close positions on the phylogenetic tree in Fig. 4, and they are clustered in the same chromosome 23. bPRP-VIII may be placed on chromosome 23 because bPRP-VIII was close to bPRP-III or -VII on the phylogenetic tree, although the bootstrap value is low between bPRP-VIII and bPRP-III/-VII. bPRP-I, -V, and -VI, however, were comparatively close on the phylogenetic tree in Fig. 4; these genes were clustered on chromosome 12. bPRP-IX may be placed on chromosome 12 because bPRP-IX was close to bPRP-I on the phylogenetic tree. The actual configuration of the bPRP-VIII and -IX genome will be addressed in the future. Therefore, prolactin-related genes in mammals may have similar roles for placental function, but they may evolve through different phylogenetic processes, may be two separate pathways. For example, GH and PRL functions are different in various species, namely, GH uses the PRL receptor in some species, but not in others [7,36]. In conclusion, we identified two new members of bPRPs, bPRP-VIII and -IX. bPRP-VIII was expressed in binucleate cells in bovine trophoblast tissue and placentome. The expression appeared as well in bPRP-I, -VII, and bPL, and it expressed from the implantation period to late in gestation. bPRP-IX was expressed in binucleate cells and trinucleate cells, and was expressed somewhat late in gestation compared to other PRPs. These data indicate that various PRP genes in bovine placenta have coordination roles for gestation as evidenced in rodents. Acknowledgements The authors thank Prof. Yukio Tsunoda of Kinki University and Dr. Tomoyuki Tokunaga of the National Institute of Agrobiological Sciences for their help. This research was supported by a grant from the Bio-oriented Technology Research Advancement Institution (BRAIN), a grant of Research Project for Utilizing Advanced Technologies (05–1770), a grant-in-aid (HC-04-2261-1) from the Ministry of Agriculture, Forestry, and Fisheries of Japan, grants (Hoga-kenkyu 16658105; Kiban-kenkyu C 17580284; Kiban-kenkyu B 17380172) from the Ministry of Education, Culture, Sport, Science, and Technology of Japan, and a grant of Animal Remodeling Project (05–201) from National Institute of Agrobiological Sciences. ==== Refs Dietz AB Georges M Threadgill DW Womack JE Schuler LA Somatic cell mapping, polymorphism, and linkage analysis of bovine prolactin-related proteins and placental lactogen Genomics 1992 14 137 143 1358791 10.1016/S0888-7543(05)80296-4 Schuler LA Kessler MA Tanaka M Nakashima K Nomenclature clarification for the bovine placental prolactin-related hormones Endocrinology 1991 126 2057 1915086 Ushizawa K Herath CB Kaneyama K Shiojima S Hirasawa A Takahashi T Imai K Ochiai K Tokunaga T Tsunoda Y Tsujimoto G Hashizume K cDNA microarray analysis of bovine embryo gene expression profiles during the pre-implantation period Reprod Biol Endocrinol 2004 2 77 15560851 10.1186/1477-7827-2-77 Kessler MA Duello TM Schuler LA Expression of prolactin-related hormones in the early bovine conceptus, and potential for paracrine effect on the endometrium Endocrinology 1991 129 1885 1895 1915073 Yamada O Todoroki J Kizaki K Takahashi T Imai K Patel OV Schuler LA Hashizume K Expression of prolactin-related protein I at the fetomaternal interface during the implantation period in cows Reproduction 2002 124 427 437 12201816 10.1530/rep.0.1240427 Ushizawa K Kaneyama K Takahashi T Tokunaga T Tsunoda Y Hashizume K Cloning and expression of a new member of prolactin-related protein in bovine placenta: bovine prolactin-related protein-VII Biochem Biophys Res Commun 2005 326 435 441 15582596 10.1016/j.bbrc.2004.11.045 Soares MJ The prolactin and growth hormone families: Pregnancy-specific hormones/cytokines at the maternal-fetal interface Reprod Biol Endocrinol 2004 2 51 15236651 10.1186/1477-7827-2-51 Muller H Liu B Croy BA Head JR Hunt JS Dai G Soares MJ Uterine natural killer cells are targets for a trophoblast cell-specific cytokine, prolactin-like protein-A Endocrinology 1999 140 2711 2720 10342862 10.1210/en.140.6.2711 Ain R Tash JS Soares MJ Prolactin-like protein-A is a functional modulator of natural killer cells at the maternal-fetal interface Mol Cell Endocrinol 2003 204 65 74 12850282 10.1016/S0303-7207(03)00125-4 Lin J Linzer DIH Induction of megakaryocyte differentiation by a novel pregnancy-specific hormone J Biol Chem 1999 274 21485 21489 10409714 10.1074/jbc.274.30.21485 Bittorf T Jaster R Soares MJ Seiler J Brock J Friese K Muller H Induction of erythroid proliferation and differentiation by a trophoblast-specific cytokine involves activation of the JAK/STAT pathway J Mol Endocrinol 2000 25 253 262 11013351 10.1677/jme.0.0250253 Bhattacharyya S Lin J Linzer DIH Reactivation of a hematopoietic endocrine program of pregnancy contributes to recovery from thrombocytopenia Mol Endocrinol 2002 16 1386 1393 12040023 10.1210/me.16.6.1386 Schuler LA Hurley WL Molecular cloning of a prolactin-related mRNA expressed in bovine placenta Proc Natl Acad Sci USA 1987 84 5650 5654 3475696 Kessler MA Milosavljevic M Zieler CG Schuler LA A subfamily of bovine prolactin-related transcripts distinct from placental lactogen in the fetal placenta Biochemistry 1989 28 5154 5161 2765528 10.1021/bi00438a036 Yamakawa M Tanaka M Koyama M Kagesato Y Watahiki M Yamamoto M Nakashima K Expression of new members of the prolactin growth hormone gene family in bovine placenta. Isolation and characterization of two prolactin-like cDNA clones J Biol Chem 1990 265 8915 8920 2341410 Tanaka M Yamakawa M Watahiki M Yamamoto M Nakashima K Isolation of a novel prolactin-like cDNA clone from bovine placenta: occurrence of new family members Biochim Biophys Acta 1989 1008 193 197 2736246 Tanaka M Minoura H Ushiro H Nakashima K A novel cDNA clone encoding a prolactin-like protein that lacks the two C-terminal cysteine residues isolated from bovine placenta Biochim Biophys Acta 1991 1088 385 389 2015300 Saitou N Nei N A neighbor-joining method: a new method for constructing phylogenetic tree Mol Biol Evol 1987 44 406 425 3447015 Page RDM TREEVIEW: An application to display phylogenetic trees on personal computers Comput Appl Biosci 1996 12 357 358 8902363 Ogata K Umeyama H An automatic homology modeling method consisting of database searches and simulated annealing J Mol Graphics Mod 2000 18 258 272 10.1016/S1093-3263(00)00037-1 Bernstein HJ Recent changes to RasMol, recombining the variants Trends Biochem Sci 2000 25 453 455 10973060 10.1016/S0968-0004(00)01606-6 Herath CB Shiojima S Ishiwata H Katsuma S Kadowaki T Ushizawa K Imai K Takahashi T Hirasawa A Tsujimoto G Hashizume K Pregnancy-associated changes in genome-wide gene expression profiles in the liver of cow throughout pregnancy Biochem Biophys Res Commun 2004 313 666 680 14697243 10.1016/j.bbrc.2003.11.151 Kizaki K Yamada O Nakano H Takahashi T Yamauchi N Imai K Hashizume K Cloning and localization of heparanase in bovine placenta Placenta 2003 24 424 430 12657517 10.1053/plac.2002.0909 Kizaki K Nakano H Nakano H Takahashi T Imai K Hashizume K Expression of heparanase mRNA in bovine placenta during gestation Reproduction 2001 121 573 580 11277877 10.1530/rep.0.1210573 Laemmli UK Cleavage of structural proteins during the assembly of the head of bacteriophage T4 Nature 1970 227 680 685 5432063 10.1038/227680a0 Towbin H Staehelin T Gordon J Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications Proc Natl Acad Sci USA 1979 76 4350 4354 388439 Milosavljevic M Duello TM Schuler LA In situ localization of two prolactin-related messenger ribonucleic acids to binucleate cells of bovine placentomes Endocrinology 1989 125 883 889 2752982 Nakano H Takahashi T Imai K Hashizume K Expression of placental lactogen and cytokeratin in bovine placental binucleate cells in culture Cell Tissue Res 2001 303 263 270 11291772 10.1007/s004410000316 Anthony RV Liang R Kayl EP Pratt SL The growth hormone/prolactin gene family in ruminant placentae J Reprod Fertil Suppl 1995 49 83 95 7623351 Wooding FBP Beckers JF Trinucleate cells and the ultrastructural localisation of bovine placental lactogen Cell Tissue Res 1987 247 667 673 3568109 10.1007/BF00215761 Ain R Dai G Dunmore JH Godwin AR Soares MJ A prolactin family paralog regulates reproductive adaptations to a physiological stressor Proc Nat Acad Sci USA 2004 101 16543 16548 15545614 10.1073/pnas.0406185101 Kessler MA Schuler LA Purification and properties of placental prolactin-related protein-I Placenta 1997 18 29 36 9032807 10.1016/S0143-4004(97)90068-0 Takahashi T Aso H Hashizume K Immunological and biological activities of bovine placental lactogen in placental explant culture J Reprod Dev 2001 47 63 67 10.1262/jrd.47.63 Wiemers DO Shao LJ Ain R Dai G Soares MJ The mouse prolactin gene family locus Endocrinology 2003 144 313 25 12488360 10.1210/en.2002-220724 Öztürk A Fresnoza A Savoie A Duckworth HW Duckworth ML Defining regulatory regions in the rat prolactin gene family locus usinga large P1 genomic clone Endocrinology 2003 144 4742 4754 12959976 10.1210/en.2003-0591 Talamantes F Ogren L Markoff E Woodard S Madrid J Phylogenetic distribution, regulation of secretion, and prolactin-like effects of placental lactogens Fed Proc 1980 39 2582 2587 6991292	16332262	PMC1325045	CC BY	2021-01-04 16:23:28	no	Reprod Biol Endocrinol. 2005 Dec 7; 3:68	utf-8	Reprod Biol Endocrinol	2,005	10.1186/1477-7827-3-68	oa_comm
==== Front BMC ImmunolBMC Immunology1471-2172BioMed Central London 1471-2172-6-241632975910.1186/1471-2172-6-24Methodology ArticleGeneration of functional HLA-DR1101 tetramers receptive for loading with pathogen or tumour derived synthetic peptides Moro Monica [email protected] Virginia [email protected] Chiara [email protected] Eliana [email protected] Barbara [email protected] Massimo [email protected] Nicholas [email protected] Maria Pia [email protected] Paolo [email protected] Giulia [email protected] Experimental Immunology Unit, Cancer Immunotherapy and Gene Therapy Program, Dept. of Oncology, DIBIT San Raffaele Scientific Institute, 20132 Milano, Italy2 Biocrystallography Unit, DIBIT San Raffaele Scientific Institute, 20132 Milano, Italy3 INSERM E0344, Universite de Nice-Sophia Antipolis, Valbonne, France4 Tumour Immunology Unit, Cancer Immunotherapy and Gene Therapy Program, Dept. of Oncology, DIBIT San Raffaele Scientific Institute, 20132 Milano, Italy2005 5 12 2005 6 24 24 26 9 2005 5 12 2005 Copyright © 2005 Moro 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 MHC class I-peptide tetramers are currently utilised to characterize CD8+ T cell responses at single cell level. The generation and use of MHC class II tetramers to study antigen-specific CD4+ T cells appears less straightforward. Most MHC class II tetramers are produced with a homogeneously built-in peptide, reducing greatly their flexibility of use. We attempted the generation of "empty" functional HLA-DR1101 tetramers, receptive for loading with synthetic peptides by incubation. No such reagent is in fact available for this HLA-DR allele, one of the most frequent in the Caucasian population. Results We compared soluble MHC class II-immunoglobulin fusion proteins (HLA-DR1101-Ig) with soluble MHC class II protein fused with an optimised Bir site for enzymatic biotynilation (HLA-DR1101-Bir), both produced in insect cells. The molecules were multimerised by binding fluorochrome-protein A or fluorochrome-streptavidin, respectively. We find that HLA-DR1101-Bir molecules are superior to the HLA-DR1101-Ig ones both in biochemical and functional terms. HLA-DR1101-Bir molecules can be pulsed with at least three different promiscuous peptide epitopes, derived from Tetanus Toxoid, influenza HA and the tumour associated antigen MAGE-3 respectively, to stain specific CD4+ T cells. Both staining temperature and activation state of CD4+ T cells are critical for the binding of peptide-pulsed HLA-DR1101-Bir to the cognate TCR. Conclusion It is therefore possible to generate a soluble recombinant HLA-DR1101 backbone that is receptive for loading with different peptides to stain specific CD4+ T cells. As shown for other HLA-DR alleles, we confirm that not all the strategies to produce soluble HLA-DR1101 multimers are equivalent. ==== Body Background The direct ex-vivo visualization and quantification of antigen-specific T cells is key to the characterisation of complex immune responses. The antigen specificity of T cells is determined by the highly specific interaction between their TCR and the cognate MHC-peptide complex. The low affinity and fast off-rate of this interaction, however, precludes its exploitation to identify specific T cells [1-5]. Tetramer technology allows to circumvent this limitation, since the overall increased avidity of the MHC multimers compensates for the low affinity of the TCR-peptide/MHC interaction [6,7]. MHC class I tetramers are now indeed largely used to characterise CD8+ T cells in many basic or clinical settings [8-13]. By contrast, the generation of functional soluble peptide/MHC class II multimers, to characterise CD4+ T cell responses, seems somewhat less straightforward. Structural differences between MHC class I and class II molecules might account for the greater difficulties to standardise the generation and production of MHC class II tetramers. In MHC class I molecules, the α chain alone contains the complete peptide-binding groove and it is stable in the soluble form when associated with the β2-microglobulin chain [14]. In contrast, both the α and β chain of the MHC class II molecules contributes to the formation of the peptide-binding site, and molecular "zippers" are necessary to guide the correct inter-chain dimerisation to stabilize the soluble recombinant αβ heterodimers. Moreover, owing to the different structure of the peptide-binding groove, MHC class I and II molecules bind peptides stably and in an homogeneous frame or unstably and in heterogeneous frames, respectively [15,16]. A variety of approaches have been used to overcome these hurdles and, in general, it appears that there does not exist a universal way to produce soluble MHC class II molecules. Most of the adopted strategies, however, generate MHC-class II molecules endowed with little or no flexibility in terms of peptide loading. For instance, HLA-DR1 and H2-IEk have been successfully produced in bacteria [17-19]. In analogy with the production of soluble MHC class I molecules, their α and β chains are separately produced in E. coli as denatured proteins, which are allowed to refold correctly as αβ heterodimer in the presence of a single peptide. Once refolded, however, the MHC class II heterodimer cannot exchange the peptide any longer. Another way to produce functional class II tetramers relies on the generation of soluble MHC class II αβ heterodimers engineered with a covalently linked peptide at the N-terminus of either chain, via a flexible amino acid linker of various lengths. This approach results in a homogeneous binding of the target peptides to the soluble MHC class II heterodimer [20-23]. This strategy has been successfully applied to both human and mouse MHC class II molecules produced in the prokaryotic as well as the eukaryotic expression system. The latter system is generally based on insect or mammalian cells, which offer the advantage over bacteria of producing proteins in native conformation and in glycosylated form. Although the presence of a covalently linked peptide improves the yield of functional soluble MHC class II tetramers, it negatively impacts on their flexibility of use, forcing the separate production of a soluble MHC II molecule for each peptide of interest. The generation of soluble recombinant MHC class II molecules without covalently linked peptide has been successfully used with HLA-DR0401 and HLA-DR0404 molecules [23-26]. This approach permits an increased flexibility in the usage of recombinant HLA-DR monomers, which are loaded with the desired synthetic peptides by in vitro incubation. Moreover, no univocal strategy to multimerise soluble recombinant MHC class II monomers has been yet devised. For this purpose, in fact, two main modifications are currently introduced at the C-terminus of the MHC class II monomer. In the first approach, the extracellular domains of either MHC class II α or β chain are linked to the Ig constant region. Chimeric MHC-class II-Ig molecules are generally multimerised by using protein A, which is bivalent, thus generating a tetrameric (dimer of dimers) peptide/MHC complex [21,22,27,28]. Alternatively, the extracellular domains of MHC class II α or β chain are linked to an enzymatic biotinylation site (Bir) [19,20,24,29]. After labelling with biotin, tetramerisation of MHC class II-Bir monomers is accomplished by coupling with fluorochrome-conjugate streptavidin [19,20,24]. We were interested in generating HLA-DR1101 tetramers to characterise CD4+ T cell responses to several peptide epitopes derived from either environmental or tumour-associated antigens. This molecule represents one of the most frequent HLA-DR allele in Caucasians, being expressed in up to 20% of the population, we therefore decided to attempt the production of HLA-DR1101 heterodimers without covalently linked built-in peptide, receptive for loading the desired peptide by incubation. Since no information is currently available on functional empty tetramers made of this HLA-DR allele, and given the premises outlined above, we compared two strategies for the production of soluble HLA-DR1101 molecules in insect cells: the first generating soluble chimeric HLA-DR1101-Ig molecules, the second HLA-DR1101-Bir ones. We show here that it is indeed possible to produce functional HLA-DR1101 tetramers, which can be loaded with the desired peptide after the purification of the protein. For this purpose, HLA-DR1101-Bir molecules resulted superior to the HLA-DR1101-Ig chimeric constructs. Results Production and biochemical characterisation of soluble HLA-DR1101 molecules To produce HLA-DR1101-Ig and HLA-DR1101-Bir molecules, the cDNAs coding the extracellular domains of the HLA-DR α and the HLA-DR1101 β chain were fused in frame at the 5' end with the sequence encoding for the Drosophila leader peptide Bip, and at the 3' end with the acidic (AZ) and basic (BZ) leucin zipper sequences, respectively (Figure 1a). A sequence coding a hexahistidine tag (His tag) was subsequently added at the 3' end of the HLA-DR1101 β chain cDNA (Figure 1a). To generate a chimeric HLA-DR1101-Ig molecules, the DR α chain was linked at the 3' end with a cDNA encoding for the hinge and Fc fragment of human IgG1 (Figure 1a). To allow targeted biotin-labelling, the HLA-DRα-AZ chain was fused at the 3' end to a sequence coding a peptide recognized by the BirA enzyme (Figure 1a). Each construct was stably transfected into S2 Drosophila melanogaster cells, that were initially grown in selective medium as bulk culture. To stabilise and optimise the production of both recombinant soluble HLA-DR1101 molecules, bulk cultures of transfected S2 cells were cloned by limiting dilution, and clones producing highest quantity of soluble molecules were selected and expanded. The HLA-DR1101-Ig recombinant molecule was secreted in the culture supernatant as dimers of αβ heterodimers, owing to the presence of an inter-chain disulfide bond between two Ig hinge regions, and was purified using an affinity chromatography with protein A-sepharose (Figure 1b). The formation of the disulfide bond between two chimeric HLA-DRα-Ig molecules was confirmed by SDS-PAGE of the purified molecules under reducing and non-reducing condition, followed by western blot analysis using anti-His tag and anti-Ig antisera (Figure 1c). The HLA-DR1101-Bir recombinant molecule was secreted as αβ heterodimer by S2 cells and purified by immunoaffinity chromatography with the human HLA-DRα chain-specific mAb L243 (Figure 1d). Noteworthy, the L243 mAb binds a conformational epitope on HLA-DRα which depends on the correct folding of the αβ heterodimer [30,31]. The immunoprecipitation of HLA-DR1101-Ig and HLA-DR1101-Bir molecules with the mAb L243 shows that only a little fraction of HLA-DR1101-Ig molecule can be bound by the mAb, and most of the protein remains in the culture supernatant, suggesting that the secreted HLA-DR1101-Ig molecules have lost the native conformation recognised by L243 (Figure 2a). A second immunoprecipitation of HLA-DR1101-Ig molecules from the same supernatant with fresh L243-conjugated beads did not precipitated additional soluble molecule, ruling out that the little protein precipitated was the result of insufficient amounts of mAb-conjugated beads (data not shown). By contrast, as shown in Figure 2a, almost all the HLA-DR1101-Bir molecule contained in the culture supernatant was immuno-precipitated by the L243-conjugated beads, suggesting that this type of soluble recombinant protein exhibit a correct conformation for mAb binding. Based on the densitometric analysis of the immunoprecipitated molecules, we assumed that only 30% of the HLA-DR1101-Ig molecule present in the supernatant displayed a correct conformation, in contrast with nearly 100% of correct conformation displayed by the HLA-DR1101-Bir molecule (Figure 2b). Considering that we have selected transfected S2 cell clones that secreted similar amounts of either soluble recombinant HLA-DR1101 molecule (not shown), we would need to produce three times more supernatant of HLA-DR1101-Ig than HLA-DR1101-Bir to purify the same amount of correctly folded molecule. To further verify the biochemical properties and stability of the two types HLA-DR1101 molecules, we performed analytical size-exclusion chromatography on the purified proteins. Figure 2c–d shows the elution profile of HLA-DR1101-Ig and of HLA-DR1101-Bir molecules, respectively. The HLA-DR1101-Ig molecule was eluted in two peaks at elution volume of 7,5 ml, corresponding to the void volume of the column, and 10,6 ml, corresponding to >900 and 500 Kda respectively. Since the expected molecular weight for the HLA-DR1101-Ig molecule is about 300 KDa, this result suggested a high propensity of the molecule towards aggregation. Moreover, a dot blot analysis could detect the DR1101 protein in the two main elution peaks only when an anti His-tag antibody, but not when the L243 antibody, was utilised. These results further indicated that the HLA-DR1101-Ig molecule was unstable and tended to progressively aggregate after purification. On the contrary, the elution profile of HLA-DR1101-Bir (Figure 2d) showed a single symmetric peak at an elution volume of 13,6 ml, corresponding to the expected molecular weight of 80 KDa of the HLA-DR1101-Bir αβ heterodimer. In this case, the molecule could be detected in a dot-blot with both the anti His-tag mAb and the L243 antibody, indicating that the molecule was correctly folded. Altogether these data indicate that the HLA-DR1101-Bir exhibited superior biochemical quality to HLA-DR1101-Ig molecule both in terms of tertiary/quaternary structure and stability. Peptide loading of soluble HLA-DR11 molecules The relative disadvantage in the production of HLA-DR1101-Ig molecules could be compensated by possible peculiar physical properties displayed by this molecule in comparison with HLA-DR1101-Bir molecules. We speculated in fact that, once pulsed with peptide and tetramerised, the presence of the flexible Ig hinge region in the HLA-DR1101-Ig construct could represent an advantage and facilitate the binding of cognate TCRs, in comparison with the more rigid HLA-DR1101-Bir molecules. For this reason, we investigated the functional properties of both HLA-DR1101-Ig and HLA-DR1101-Bir molecules. The purified soluble HLA-DR1101 molecules were loaded with the Tetanus Toxoid-derived peptide TT830–845 (p2), a very well characterized promiscuous peptide binding with high affinity to different HLA-DR alleles [32]. Both purified HLA-DR1101-Ig and HLA-DR1101-Bir molecules were incubated for 72 hours at 37°C in a mild acidic buffer (pH 5) in the presence of a 50 fold molar excess of p2 peptide. The efficacy of peptide loading was then evaluated by a SDS-resistance assay [33]. The high affinity binding of a peptide to the HLA-DR groove stabilize in fact the association between the HLA-DR α and β chains, making in some cases the αβ heterodimer resistant to the denaturing action of SDS at r.t. [33]. As shown in Figure 3a, the "empty" (nil) HLA-DR1101-Bir molecules were denatured upon incubation in SDS-containing buffer at r.t., and the DRα and β chains migrated as separated chains in both boiled (B) and not boiled (NB) samples. On the contrary, the binding of p2 substantially stabilised the HLA-DR αβ heterodimer, leading to the appearance of an 80 KDa band in the gel, corresponding to the peptide-MHC complex. Unlike HLA-DR1101-Bir molecules, the incubation in SDS at r.t. did not induce dissociation of HLA-DR1101-Ig α and β chains even when the p2 peptide was not added (Figure 3b). These results indicated that the HLA-DR1101-Ig molecule displayed a stronger interaction between the α and β chains, as compared to HLA-DR1101-Bir molecules, precluding the quantification of peptide loading by this method. To characterise further the loading of both soluble recombinant HLA-DR1101 molecules with the exogenous peptides, we relied on an indirect functional assay, in which p2-loaded HLA-DR1101 molecules were utilised in vitro to activate p2-specific CD4+ T cells. Since the assay was performed in the absence of antigen presenting cells, a sub-optimal dose of phorbol myristate acetate (PMA) was added as costimulatory signal for T cells. As shown in Figure 3c, both TT830–845-loaded HLA-DR1101-Ig and HLA-DR1101-Bir molecules were able to induce IFN-γ production by p2-specific CD4+ T cells, but not by CD4+ T cells specific for an irrelevant peptide or restricted for another HLA-DR allele. These results indicated that both HLA-DR1101-Ig and HLA-DR1101-Bir molecules could be loaded in vitro with a peptide of interest. Staining of specific CD4+ T cells with peptide-loaded HLA-DR1101 tetramers We finally compared the staining capacity of both soluble HLA-DR1101 molecules, loaded with different CD4 peptide epitopes. We selected for the staining experiment two p2-specific CD4+ T cell clones displaying low and high affinity for the peptide respectively, as shown in Figure 4a. HLA-DR1101-Ig or HLA-DR1101-Bir were loaded with p2 peptide, multimerized and used for the staining at 4°C or 37°C, according to previous data suggesting a significant temperature effect on the interaction between the TCR and the peptide-pulsed MHC-class II tetramers [25,34]. As shown in Figure 4b, HLA-DR1101-Ig tetramers did never stain either p2-specific T cell clones at any temperature. By contrast, HLA-DR1101-Bir tetramers stained both low and high affinity p2-specific T cells, although the low affinity T cells clone could only be stained at 37°C, while the high affinity one could be stained at both temperature, though with a much higher fluorescence intensity at 37°C. Collectively these results indicated that HLA-DR1101-Bir tetramers stain peptide-specific CD4+ T cells, and required an optimal temperature of staining of 37°C. By contrast, HLA-DR1101-Ig could not stain peptide-specific T cells, despite their capacity to be loaded with peptides. Staining with HLA-DR1101-Bir tetramers depends on the activation state of CD4+ T cells In an independent series of experiments, we noticed that the staining level of CD4+ T cells by peptide-pulsed HLA-DR1101-Bir molecules was consistently higher when T cells had been recently restimulated with either cognate peptide or polyclonal stimuli in the presence of PBMCs as APCs. This observation prompted a more systematic investigation of the phenomenon. This is in fact a relevant issue when considering the use of HLA-DR1101 tetramers for ex vivo staining, since primary CD4+ T cells might display a whole range of different activation states, which in turn could affect the "detectability" of the whole peptide-specific T cell repertoire by the HLA-DR1101 tetramers. We therefore compared the capacity of HLA-DR1101-Bir tetramer to stain the high affinity p2-specific CD4+ T cell clone 162 at different time points from restimulation. As shown in Figure 5, the 162 T cell clone could be stained progressively less by p2-HLADR1101 tetramers moving from day 5 to day 14 post-restimulation, even though the mean fluorescence value of the cell-surface CD3-TCR complex was not substantially modified in the three conditions. This finding confirmed that the activation state of the target CD4+ T cells influences the binding of HLA-DR1101-Bir tetramers to the TCR. HLA-DR1101-Bir tetramers stain polyclonal CD4+ T cells specific for p2 or Influenza Hemagglutinin HA peptides We next verified the capacity of HLA-DR1101-Bir tetramers to detect polyclonal antigen-specific CD4+ T cells. Two different helper epitopes from proteins expressed by environmental pathogens were selected: TT-p2 and Influenza Hemagglutinin HA306–318 (HA) [35], another well characterised and commonly utilised promiscuous CD4 epitope. PBMCs from healthy donors were stained with p2-pulsed or HA-pulsed HLA-DR1101-Bir tetramers either ex vivo or at different time points after in vitro stimulation with the corresponding peptide. In line with published data [24,26,36], p2- or HA-specific CD4+ T cells could be hardly detected ex vivo by HLA-DR1101-Bir tetramers pulsed with either peptides (data not shown), with frequencies of tetramer positive cells in the range of those obtained with peptide-unloaded (empty) HLA-DR1101 tetramer (0,05–0,1%). After 4 days of stimulation in vitro with the specific peptide, however, a sizeable fraction of CD4+ T cells could be clearly detected with either peptide-pulsed-HLA-DR1101-Bir tetramer (Figure 6a,c), with a background staining with peptide-unloaded tetramers in the order of 0,05–0,07% in the two T cell cultures (data not shown). The percentage of CD4+ T cells stained with the tetramers paralleled fairly accurately the percentage of cells that were able to produced IFN-γ upon in vitro stimulation with the corresponding peptide at the same time point (Figure 6b,d). Again, the percentage of CD4+ T cells stainable with peptide-HLA-DR1101-Bir tetramers declined with time following the in vitro stimulation with the peptide, confirming at the polyclonal level what observed with T cell clones specific for p2, although the decline of staining of HA-specific CD4+ T cells was more rapid than that of p2-specific ones. HLA-DR1101-Bir tetramers stain CD4+ T cells specific for the MAGE-3 tumour antigen derived M3191–205 (p39) peptide We were also interested in determining whether HLA-DR1101-Bir tetramers could stain CD4+ T cells specific for the naturally processed, promiscuous p39 peptide derived from the MAGE-3 TAA [37]. As shown in Figure 7, the p39-specific CD4+ T cell clone 2C3.35 could be stained with p39-pulsed HLA-DR1101-Bir tetramers. The affinity of this CD4+ T cell clone for the p39 peptide, as shown in a dose response experiment of IFN-γ production (Figure 7b), was even lower to that of the low affinity T cell clone 51 for p2 peptide, explaining the low mfi value obtained by staining the 2C3.35 T cell clone with p39-HLA-DR1101-Bir tetramers. Thus, HLA-DR1101-Bir tetramers could be pulsed also with a naturally processed promiscuous tumour-derived peptide, and might prove useful in characterising naturally arising or therapeutically induced CD4+ T cells responses in HLA-DR1101 patients with tumours expressing MAGE-3. Discussion We have shown the generation of a versatile HLA-DR1101 tetramer. A single soluble recombinant protein backbone can bind at least three different promiscuous peptide epitopes, derived from environmental pathogens or tumour associated antigens, by a simple incubation step. This feature will greatly facilitate the use of this reagent for the study of CD4+ T cells responses specific for disparate antigens. The usefulness of this approach is evident also from the studies performed with HLA-DR0401 and HLA-DR0404 tetramers [23-26], in which the same protein backbone was loaded with either viral or self-derived peptide. The only HLA-DR1101 tetramers described so far are produced in E coli with a covalently linked HCV-derived peptide [38] and, although functional, are limited in their use to the single peptide specificity. Therefore, the production of versatile HLA-DR1101 tetramers is extremely valuable considering the frequency of expression of this allele in the Caucasian population. It is interesting to note that the same extracellular portion of the HLA-DR allele behaves differently in terms of stability when engineered with different C-terminal moieties. The presence of an Ig constant domain at the C-terminus seems in fact to destabilise the "empty" HLA-DR1101 soluble recombinant molecule significantly more than the presence of the Bir sequence. The most straightforward explanation for this phenomenon holds that the extended intermolecular interactions introduced by the pairing of two Ig Fcs may disturb the native conformation of the chimeric HLA molecules. In fact our results show that the addition of a short Bir sequence at the C-terminus does not impact on the overall molecular architecture of this human HLA-DR allele. Evidence for a conformational alteration in the HLA-DR1101-Ig molecule stems from the immuno-precipitation experiments with the anti-HLA-DRα chain-specific monoclonal antibody L243, which recognizes a conformational epitope that is present only when the HLA-DRα and β chain are correctly paired [30,31]. L243 immunoprecipitates very inefficiently HLA-DR1101-Ig molecules, whereas it efficiently binds HLA-DR1101-Bir, suggesting differences in the structures of the two proteins or a masked epitope in the former. Several mouse MHC class II alleles have nevertheless been successfully produced as chimeric MHC-Ig fusion proteins, with or without a linked peptide, and it is possible that HLA-DR alleles other than DR1101 might be produced in this chimeric form. It is clear from our study that the affinity of the TCR expressed by the CD4+ T cells for the cognate peptide-HLA-DR1101 complex will dictate the success of detection of specific primary lymphocyte population by the fluorescent tetramer. Moreover, we document that the activation state of the target CD4+ T cells affects the binding of peptide-HLA-DR tetramers to the cognate TCR, suggesting that biochemical pathways linked to CD4+ T cell activation modify the avidity of the TCR for the HLA-DR tetramer. The fact that the binding of the TCR by the cognate peptide-HLA-DR1101 tetramers requires an active metabolism and intact membrane trafficking is suggested further by the 37°C temperature requirement for optimal CD4+ T cell staining. Collectively, these observations might help to explain why the detection ex-vivo of primary CD4+ T cells by MHC class II tetramers seems more difficult than the detection of specific CD8+ T cells by MHC class I tetramers. Further study are needed to determine whether modifying rationally the structure of the helper peptide epitope to increase its affinity for the MHC class II allele and/or the cognate TCR, and manipulating the T cell activation pathways to increase the avidity of the TCR for HLA-DR tetramers, my lead to improved ex-vivo detection of specific CD4+ T cells. Conclusion In conclusion, we report here the successful engineering of stable empty HLA-DR1101-Bir tetramers. These molecules are novel powerful reagents for the study of CD4+ T cell responses towards diverse peptide antigens. Our comparative study on the production of two different "empty" HLA-DR1101 constructs, together with the evidences present in the literature, suggests that every MHC class II isotype, and even allele, might display distinct biochemical characteristics, requiring the empiric definition of the optimal strategy for the production as functional soluble recombinant molecules. Methods Construction of soluble recombinant HLA-DR1101-Ig and HLA-DR1101-Bir molecules for insect cells expression All the DR α and β constructs were cloned in the pMT/Bip/V5-His vector (Invitrogen, Groningen, the Nederlands) in frame with the Drosophila BiP secretion signal, under the control of the Drosophila metallothionein promoter. The extracellular region of the HLA-DRα chain, deprived of the leader sequence, was cloned by RT-PCR using the following primers: oligo-1-Up DRα (BglII) 5'GGGAGATCTATCAAAGAAGAACATGTGATCATCCAG3', oligo-2-Dw DRα (BamHI) 5'GGATCCTCCACCTCCGTTCTCTGTAGTCTCTGGG3'. The PCR product was cloned into the PCR2.1 vector (Invitrogen) and sequenced. To generate the HLA-DRα-Bir chain, a cassette containing the sequence for the Acidic Leucine Zipper-Bir A target peptide was generated by fusion PCR with the following primers, as illustrated in Figure 1: Oligo-3-Up AZ (BamHI) 5'GGAGGATCCACTACAGCTCCATCAGCTCAG3', Oligo-4-Dw AZ (Bir) 5'ACCACCACCGTCCCCACCCTGAGCCAGTTC3', Oligo-5-Up Bir (AZ) 5'GGTGGGAGCGGTGGTGGTCTGAACGATATTTTTG3' and Oligo-6-Dw Bir (NotI) 5'CTAGCGGCCGCTATTCATGCCATTCGAT3'. The PCR product was cloned into the PCR2.1 vector and sequenced, then subcloned in frame with the HLA-DRα fragment separated by a five-aminoacid linker (GGGGS). The resulting sequence encoding the HLA-DRα-AZ-Bir fusion protein was excised with BglII and NotI restriction enzymes, and cloned into the Drosophila melanogaster expression vector pMT/Bip/V5/His. To generate the HLA-DRα-Ig chain, a cassette containing the sequence for the Acidic Leucine Zipper-hIgG1 constant region was generated by fusion PCR as illustrated in Figure 1, using the following primers: oligo-3-upAZ(BamHI) described above, oligo-9-Dw AZ (Ig) AGATTTGGGCTCAGATGCTGCCTGAGCCAGTTC, oligo-10-Up Ig (AZ) GCAGCATCTGACGCCCAAARCTTGTGACAAAACTCAC and oligo-11-Dw IgG1 (NotI) 5'TGGCGGCCGCCGCACTCATTTACCCGGAGA. The PCR product was cloned into the PCR2.1 vector, sequenced and then subcloned in frame with the HLA-DR1101α chain fragment. The resulting sequence encoding for the HLA-DR1101α-AZ-IgG1 fusion protein was excised with BglII and NotI, and cloned in the Drosophila melanogaster expression vector pMT/Bip/V5/His. The HLA-DR1101β chain, shared by both type of soluble recombinant HLA-DR1101 molecules, was generated as follows. The extracellular region of the HLA-DR1101β, chain deprived of its leader sequence, was cloned by RT-PCR using the following primers: oligo-7-Up DRβ (BglII) 5'GGGAGATCTGGGGACACCAGACCACGTTTC3', oligo-8-Dw DRβ (BamHI) 5'GTGGATCCTCCACCTCCCTTGCTCTGTGCAGATTC3'. The PCR product was cloned into the PCR2.1 vector and sequenced. The HLA-DR1101β cDNA was subcloned in frame with a Basic Leucine Zipper-His tag cassette derived from the pCO268 plasmid spaced by a five-aminoacid linker (GGGGS) [21]. The resulting coding sequence was then excised with BglII and SalI restriction digestion, and cloned into the Drosophila melanogaster expression vector pMT/Bip/V5/His, digested with BglII and XhoI. Expression of HLA-DR1101 in Drosophila cells S2 cells were grown in SFX medium (Hy-clone, South Logan, UT, USA) supplemented with 10% heat inactivated foetal calf serum (Euroclone, Milano, Italy), 50 u/ml penicillin, 25 μg/ml streptomycin, 25 μg/ml Kanamycin (Gibco) and 1 μg/ml Amphotericin B (Gibco, Paisley, Scotland, UK) at 27°C to a density of 2–4 × 106 cells/ml. The α- and β-chain expressing vectors (15 μg each) were co-transfected with 0,5 μg of the selection plasmid pCoHYGRO using a calcium phosphate transfection kit (Invitrogen). Three days after the transfection, selection medium containing 300 μg/ml of Hygromicin B (Roche, Indianapolis, IN, USA) was added to cells. Transfected cells were cultured in SFX medium supplemented with 2% of foetal calf serum and 300 μg/ml of Hygromicin B. Cells were kept in the exponential growth phase (7–20 × 106 cells/ml) and the protein production was induced by the addition of 1 mM of CuSO4 when the cells were at a density of 107/ml. Protein production was monitored by dot blot analysis with either mouse anti-human HLA-DR α monoclonal antibody L243 (ATCC, Manassas, VA, USA) or rabbit anti-His tag polyclonal antibody His-probe (Santa Cruz, Santa Cruz, CA, USA), followed by an HRP-labelled goat anti mouse or anti rabbit antibody (Southern Biothecnology, Birmingham, AL, USA). The assay was developed with ECL (Amersham, Uppsala, Sweden). To improve the production of recombinant proteins, transfected cells were cloned by limiting dilution. S2 cells expressing the desired HLA-DR1101 molecule were diluted to 2.5 cell/ml in a suspension of irradiated (8000 rad) S2 wild type cells at the concentration of 3 × 105 cells/ml in medium+Hygromicin B, and plated in flat bottomed 96 w plates. The best producer clones were selected by dot blot analysis on the culture supernatant as described above, upon induction with 1 mM of CuSO4 of the same number of cells, and used for routine protein production. Purification of soluble HLA-DR1101 molecules Both HLA-DR1101-Ig and HLA-DR1101-Bir molecules were purified from the S2 cells supernatant by immunoaffinity chromatography using ProtA-Sepharose (Amersham Bioscience) and HLA-DRα-specific L243 mAb-Sepharose, respectively. The HLA-DR1101-Ig molecules were stored at -80°C at a concentration of 1 mg/ml in PBS buffer. HLA-DR1101-Bir molecules were dialyzed against 10 mM Tris buffer, pH 8.0, and concentrated up to 3 mg/ml on Vivaspin2 concentrator (Vivascience, Hannover, Germany). The protein was biotinylated using the BirA enzyme according to the manufacturer's instructions (Avidity, Denver, CO, USA), then was diluted to 1 mg/ml and stored at -80°C. The formation of molecular aggregates was assessed by size exclusion chromatography on a Superdex 200 HR 10/30 column using a ÅKTA FPLC chromatography system (Amersham Bioscience). The calibration of the Superdex 200 column was performed with the Gel Filtration Calibration Kit high molecular weights (Amersham Bioscience), following the manufacturer indications. Partition coefficients were calculated using the relation Kd = (Ve-Vo)/(Vc-Vo), where Ve, Vc and Vo are the elution, column and void volumes, respectively. Peptide loading into soluble HLA-DR1101 molecules Both HLA-DR1101-Ig and HLA-DR1101-Bir were loaded with synthetic peptides by incubation at 37°C for 72 hours with 50 fold molar excess of either Tetanus Toxoid peptide residues 829–844 (p2), Hemagglutinin peptide residues 306–318 (HA) or MAGE-3 tumour associated antigen peptide residues 191–205 (p39) in 10 mM Tris, 50 mM Glycine pH 5.0 with 0.2% of n-octyl-β-D-glucopyranoside (Sigma-Aldrich, St. Louis, MO, USA). Tetramerization of soluble HLA-DR1101 molecules and staining of specific T cells Tetramerization of HLA-DR1101-Ig and HLA-DR1101-Bir were achieved by adding FITC-labelled protein A (Molecular Probes, Leiden, The Netherlands) at 5:1 molar ratio, or PE-labelled streptavidin (Molecular Probes) at 5:1 molar ratio, respectively. Tetramer staining was performed for 2 hours at 37°C in complete medium. The cells are then washed twice, transferred on ice and stained for the other surface markers. Immediately before the analysis TOPRO-3 (Molecular Probes) is added following the manufacturer's instructions to exclude dead cells. The analysis is performed on CD3+CD4+ viable cells. Generation of specific CD4+ T cell lines and clones 20 million of freshly purified PBMCs from a HLA-DR1101/1104 healthy donor were cultured for 7 days in RPMI 1640 (Gibco) supplemented with 10% of human serum, 2 mM Glutamax I (Gibco), 100 u/ml penicillin and 50 μg/ml streptomycin (Gibco), in the presence of the specific peptide at a concentration ranging from 1 to 10 μg/ml. After seven days, blasts were separated from small T cells on a 30–70% discontinuous Percoll gradient [39], and put in culture in complete medium with 10–20 u/ml of human recombinant IL-2 (Roche). Peptide specific T cells were kept in culture by weekly re-stimulation with the same amount of peptide and irradiated (4000 rad) autologous PBMCs as antigen presenting cells. Peptide-specific T cell clones were generated by limiting dilution of Percoll-purified T cell blasts blasts as follows. Blasts were diluted at 25 cells/ml in a suspension containing 0.5 × 106 cells/ml of a mixture of two different allogeneic PBMCs supplemented with 1 μg/ml of PHA-L (Roche) and 200 u/ml of human recombinant IL-2 (Roche). The cloning mix is dispensed in 25 μl in flat-bottomed Terasaki plates and incubated at 37°C for 10–20 days. Growing cells are then tested for the recognition of peptide pulsed HLA matched EBV cell lines. T cells functional assays To test IFN-γ production, 104 T cells were plated in U bottom 96 well plates in RPMI 1640 supplemented with 10% of human serum, 2 mM Glutamax I, 100 u/ml penicillin, 50 μg/ml streptomycin, together with 5 × 104 irradiated (6000 rad) HLA-DR1101 LCL cells as antigen presenting cells, and the specific peptide at the concentrations detailed in the figure legends. IFN-γ production in the supernatant was measured by ELISA after 48 hours of incubation (IFN-γ detection kit-Endogen). For the intracellular analysis of IFN-γ production, T cells were stimulated for 6 hours with HLA-DR1101 LCL of at 1:5 ratio in three conditions: i. without peptide; ii. with 5 μg/ml of peptide; and iii. with 50 ng/ml of PMA (Sigma) plus 500 ng/ml of Ionomycin (Sigma). After the first hour of stimulation, Brefeldin A was added at the final concentration of 10 μg/ml. At the end of the stimulation, the cells were harvested, fixed with PFA1%, permeabilized with 0,5% saponin and stained for intracellular cytokine expression. For the T cell activation with peptide-loaded soluble recombinant HLA-DR molecules, 1 μg of peptide loaded HLA-DR1101-Ig or HLA-DR11-Bir molecules in 50 μl of PBS were incubated in U bottomed 96 well plates for 3 hours at 37°C. After washing, 2 × 104 specific T cells were added to the wells and incubated for 48 hours at 37°C in the presence of 1 ng/ml of PMA. Cells incubated either with PMA alone or in the presence of 1 μg of plate coated TR66 anti-CD3 mAb are used as negative and positive control, respectively. IFN-γ production in the supernatant is measured by ELISA (Endogen, Woburn, MA, USA) after 48 hours of culture. Antibodies and flow cytometry The following antibodies have been used: FITC- and APC-labelled mouse anti human CD3 (Becton Dickinson, San Jose, CA, USA), Quantum Red-conjugated mouse anti-human CD4 (SIGMA), PE-labelled rat anti-human IFN-γ (Becton Dickinson). Data were collected using a Becton Dickinson FACScalibur flow cytometer and analysed using the Cell Quest software (Becton Dickinson) Authors' contributions M.M, P.D., G.C designed experiments. M.M, V.C., C.M., E.D. produced and characterised the tetramers. B.G., M.D. contributed to the purification of the HLA-DR monomers. N.G., M.P.P contributed reagents and expertises. M.M, V.C., P.D., G.C. analysed data. M.M, P.D., G.C. wrote the paper. All authors read and approved the final manuscript. Acknowledgements Monica Moro is supported by a fellowship from American Italian Cancer Foundation. The study was supported in part by funds from Associazione Italiana per la Ricerca sul Cancro (AIRC), Cancer Research Institute, Fondazione Guido Berlucchi, Compagnia di San Paolo, Fondazione Cariplo to P.D. and G.C., and Fondazione Italiana Sclerosi Multipla Onlus (2003/R/19) to M.D. Figures and Tables Figure 1 Construction of soluble recombinant HLA-DR1101-Ig and HLA-DR1101-Bir molecules. Structure and homogeneity of purified soluble HLA-DR1101 recombinant molecules. (a) Schematic representation of HLA-DR1101-Ig and HLA-DR1101-Bir constructs. The extracellular region of HLA-DR1101β is fused with the Basic Zipper (BZ) and His tag. The HLA-DRα chain is fused with the Acid Zipper (AZ) and the human (h)IgG1 constant region or the biotynilation site BirA, respectively. All constructs are cloned into the pMT/V5/His Drosophila expression vector, in frame with the Drosophila leader sequence Bip, under the control of a metallotioneine promoter, inducible by CuSO4 addition. The HLA-DR1101-Ig molecule is secreted into the supernatant of Drosophila cells as dimer of HLA-DR1101α/β heterodimers, because of a disulphide bond between two hIgG domains. (b) SDS-PAGE of HLA-DR1101-Ig molecules secreted into the supernatant of S2 cells (sup) after CuSO4 induction, and after purification by protein A affinity chromatography (Prot A). Gels were run either in reducing (+βm) or non-reducing (-βm) conditions. Indicated are the bands corresponding to the migration of the HLA-DRα-hIg homodimers, HLA-DRα-hIg monomers and HLA-DR1101β-His tag proteins. (c) Western blotting analysis of HLA-DR1101-Ig molecules, separated on SDS-PAGE as shown in (b). The blotted filter was probed at the same time with the anti-His tag and anti-hIg probes, as indicated in figure. Empty circle and asterisk indicate the migration of the HLA-DRα-hIg and HLA-DR1101β-His tag, respectively. (d) SDS-PAGE analysis of the HLA-DR1101-Bir molecule contained into the supernatant (sup) of S2 cells after CuSO4 induction, and after purification by immunoaffinity chromatography with L243 mAb (L243). Figure 2 Soluble recombinant HLA-DR1101-Ig and HLA-DR1101-Bir molecules display different biochemical properties. Characterization of the native structure and stability of both soluble recombinant HLA-DR1101 molecules. (a) Immunoprecipitation of HLA-DR1101-Ig and HLA-DR1101-Bir molecule with 30 μl of L243-sepharose beads from 1 ml of S2 cells culture supernatant after CuSO4 induction. The lanes contain the following material: 1. sup, 30 μl of culture supernatant before immunoprecipitation; 2. bound, the immunoprecipitated soluble recombinant HLA-DR1101; and 3. not bound, 30 μl of culture supernatant after immunoprecipitation. Proteins were separated on SDS-page, transferred to filter and revealed with anti his-tag antibody. (b) Densitometric analysis of the protein bands displayed in panel (a), showing the elative efficiency of immunoprecipitation of the two soluble recombinant HLA-DR1101 proteins with L243-Sepharose beads, as an indirect indicator of the percentage of correctly folded molecule. (c) Size exclusion chromatography of HLA-DR1101-Ig molecules, after purification of ProtA affinity chromatography. The elution profile of the molecule from a Superdex 200 gel filtration column is shown. Inset shows the dot-blot analysis performed on the protein contained in the indicated elution peaks. Spotted proteins are probed with either anti-His tag antibody, to verify the presence of the HLA-DR1101 molecule, or L243 mAb (Anti-HLA-DR) to verify the correct conformation of the molecule. (d) Elution profile from Superdex 200 gel filtration column and dot-blot analysis on eluted peaks of HLA-DR11-Bir molecules, performed as described in (c). (e) Calibration profile of the Superdex 200 gel filtration column. Figure 3 HLA-DR1101-Ig and HLA-DR1101-Bir molecules can be loaded with synthetic peptides. Purified soluble HLA-DR1101-Ig and HLA-DR1101-Bir molecules were loaded with equal amount of p2 peptide. Stability of the peptide-loaded HLA-DR1101 αβ heterodimers was determined by running complexes in reducing condition on SDS gels without boiling, or by testing their capacity to elicit IFN-γ production in specific CD4+ T cell clones. (a) SDS-resistance assay of HLA-DR1101-Bir molecules. Indicated are the protein bands corresponding to the migration of either the HLA-DR1101-Bir αβ heterodimer, or the single HLA-DRα-Bir or HLA-DR1101β chains. Lane nil: unloaded HLA-DR1101-Bir molecules. Lane p2: p2-loaded HLA-DR1101-Bir molecules. (b) SDS-resistance assay of HLA-DR1101-Ig molecules. Indicated are the protein bands corresponding to the migration of either the HLA-DR1101-Ig αβ heterodimer, or the single HLA-DRα-Ig or HLA-DR1101β chains. Lane nil and p2 are the same as per HLA-DR1101-bir. (c) IFN-γ production by the p2-specific CD4+ T cell clone (TCC) 162 in response to p2-loaded HLA-DR1101-Bir or HLA-DR1101-Ig molecules, attached to plastic. The response of the MAGE-3 p39-specific CD4+ TCC 2C3.35 is shown as control. T cells were cultured in the presence of the indicated peptide-pulsed HLA-DR1101 recombinant molecules, in the presence of a costimulatory dose of PMA. As control, T cells were cultured in the presence only of sub-optimal doses of PMA (PMA), or activated by plastic-bound anti-CD3 mAb (Anti-CD3). The release of IFN-γ in the culture supernatant was determined by ELISA 48 h later. Figure 4 HLA-DR1101-Bir, but not HLA-DR1101-Ig, stain specific CD4+ T cells, and display a temperature-dependence. CD4+ TCC specific for p2, endowed with different affinity of recognition, and an irrelevant CD4+ TCC were stained with p2-loaded HLA-DR1101-Bir tetramers at different temperature. (a) The relative affinity for the p2-HLA-DR1101 displayed by the two TCC clones 162 and 51 is determined in an IFN-γ-releasing assay following production in response to different doses of p2 peptide. 104 T cells are incubated with 5 × 104 HLA-DR1101 LCL cells and the indicated doses of p2 peptide. After 48 hours, the concentration of IFN-γ in the culture supernatant is measured by ELISA. Indicated are the concentration of p2 peptide required to elicit half-maximum release of IFN-γ in the two TCC. (b) Staining of TCC162, TCC51 and the irrelevant TCC with either p2-loaded HLA-DR1101-Ig or p2-loaded HLA-DR1101-Bir tetramers. Staining is performed for 2 hours at the indicated temperatures with 10 μg of tetramer per sample. The tetramer staining on CD3+CD4+ gated cells is shown. The amount of surface TCR, obtained by staining with anti-CD3 mAb, expressed by the TCC in the different conditions is shown by the mean fluorescent intensity (CD3 mfi). Figure 5 HLA-DR1101-Bir tetramer staining correlates with the activation state of the target CD4+ T cells. The p2-specific CD4+ TCC 162 was stained with p2-loaded HLA-DR1101-Bir tetramers at different time after in vitro restimulation with p2 and APCs. The histograms in the upper row display the amount of surface TCR (mfi) expressed by the T cells at the time of tetramer staining, as determined by anti-CD3 staining. The histograms in the lower row show the staining with the p2-loaded HLA-DR1101-Bir tetramers (mfi) performed at the indicated time after restimulation. Figure 6 Peptide-loaded HLADR1101-Bir tetramers stain polyclonal T cells specific for p2 and HA. In vitro expanded T cells were stained with HLA-DR1101-Bir tetramers loaded with either p2 or HA peptides. (a) Staining of PBMCs from a HLA-DR1101 healthy donor after one round of in vitro stimulation with 5 μM of p2-peptide and 40 u/ml of IL-2. (b) Intracellular production of IFN-γ by T cells contained in the p2-enriched PBMC. T cells were stimulated for 6 hours with p2 in the presence of HLA-DR1101+ LCL cells at 37°C. Brefeldin A was added after the first hour of stimulation. The cells were then fixed, permeabilised and stained with anti-IFN-γ mAb. Cells unstimulated (nil) and stimulated policlonally with PMA-Ionomycin (PMA/ionomycin) are shown as controls. Numbers in quadrants indicate the percentage of T cells stained with anti-IFN-γ mAb (c) Staining of PBMCs from HLA-DR1101 healthy donor with HA-loaded HLA-DR1101-Bir tetramers. T cells were after three rounds of in vitro stimulation with 1 μg/ml of HA peptide and 40 u/ml of IL-2. The staining is performed at the indicated days after the third stimulation. Numbers in quadrants indicate the percentage of T cells stained by HLA-DR tetramers. (d) Intracellular production of IFN-γ by CD4+ T cells contained in the HA-specific T cell line at day 8 from the third in vitro stimulation. Activation and staining of CD4+ T cells was performed as described in (b). Figure 7 HLA-DR1101-Bir tetramers stain CD4+ T cells specific for the MAGE-3 tumour antigen derived p39 epitope. Ability of p39-loaded HLA-DR1101-Bir tetramers to stain a specific CD4+ TCC. (a) Clone M3 2C3.35 was assayed for IFN-γ production to different doses of p39 peptide. 104 T cells are incubated for 48 hours with 5 × 104 HLA-DR1101+ LCL cells and the indicated doses of p39 peptide. IFN-γ in the supernatant is measured by ELISA. (b) Staining of a MAGE-3-specific T cell clone. The cells are stained at day 19 from re-stimulation with 10 μg of either CLIP-loaded or p39-loaded HLA-DR1101-Bir tetramers for 2 hours at 37°C. ==== Refs Corr M Slanetz AE Boyd LF Jelonek MT Khilko S al-Ramadi BK Kim YS Maher SE Bothwell AL Margulies DH T cell receptor-MHC class I peptide interactions: affinity, kinetics, and specificity Science 1994 265 946 949 8052850 Matsui K Boniface JJ Steffner P Reay PA Davis MM Kinetics of T-cell receptor binding to peptide/I-Ek complexes: correlation of the dissociation rate with T-cell responsiveness Proc Natl Acad Sci U S A 1994 91 12862 12866 7809136 Garcia KC Tallquist MD Pease LR Brunmark A Scott CA Degano M Stura EA Peterson PA Wilson IA Teyton L Alphabeta T cell receptor interactions with syngeneic and allogeneic ligands: affinity measurements and crystallization Proc Natl Acad Sci U S A 1997 94 13838 13843 9391114 10.1073/pnas.94.25.13838 Liu CP Parker D Kappler J Marrack P Selection of antigen-specific T cells by a single IEk peptide combination J Exp Med 1997 186 1441 1450 9348301 10.1084/jem.186.9.1441 Garboczi DN Utz U Ghosh P Seth A Kim J VanTienhoven EA Biddison WE Wiley DC Assembly, specific binding, and crystallization of a human TCR-alphabeta with an antigenic Tax peptide from human T lymphotropic virus type 1 and the class I MHC molecule HLA-A2 J Immunol 1996 157 5403 5410 8955188 Altman JD Moss PA Goulder PJ Barouch DH McHeyzer-Williams MG Bell JI McMichael AJ Davis MM Phenotypic analysis of antigen-specific T lymphocytes Science 1996 274 94 96 8810254 10.1126/science.274.5284.94 Dunbar PR Ogg GS Oligomeric MHC molecules and their homologues: state of the art J Immunol Methods 2002 268 3 7 12213337 10.1016/S0022-1759(02)00194-1 Appay V Rowland-Jones SL The assessment of antigen-specific CD8+ T cells through the combination of MHC class I tetramer and intracellular staining J Immunol Methods 2002 268 9 19 12213338 10.1016/S0022-1759(02)00195-3 Callan MF Tan L Annels N Ogg GS Wilson JD O'Callaghan CA Steven N McMichael AJ Rickinson AB Direct visualization of antigen-specific CD8+ T cells during the primary immune response to Epstein-Barr virus In vivo J Exp Med 1998 187 1395 1402 9565632 10.1084/jem.187.9.1395 Gillespie GM Wills MR Appay V O'Callaghan C Murphy M Smith N Sissons P Rowland-Jones S Bell JI Moss PA Functional heterogeneity and high frequencies of cytomegalovirus-specific CD8(+) T lymphocytes in healthy seropositive donors J Virol 2000 74 8140 8150 10933725 10.1128/JVI.74.17.8140-8150.2000 Migueles SA Sabbaghian MS Shupert WL Bettinotti MP Marincola FM Martino L Hallahan CW Selig SM Schwartz D Sullivan J Connors M HLA B5701 is highly associated with restriction of virus replication in a subgroup of HIV-infected long term nonprogressors Proc Natl Acad Sci U S A 2000 97 2709 2714 10694578 10.1073/pnas.050567397 Tan LC Gudgeon N Annels NE Hansasuta P O'Callaghan CA Rowland-Jones S McMichael AJ Rickinson AB Callan MF A re-evaluation of the frequency of CD8+ T cells specific for EBV in healthy virus carriers J Immunol 1999 162 1827 1835 9973448 Romero P Dunbar PR Valmori D Pittet M Ogg GS Rimoldi D Chen JL Lienard D Cerottini JC Cerundolo V Ex vivo staining of metastatic lymph nodes by class I major histocompatibility complex tetramers reveals high numbers of antigen-experienced tumor-specific cytolytic T lymphocytes J Exp Med 1998 188 1641 1650 9802976 10.1084/jem.188.9.1641 Garboczi DN Hung DT Wiley DC HLA-A2-peptide complexes: refolding and crystallization of molecules expressed in Escherichia coli and complexed with single antigenic peptides Proc Natl Acad Sci U S A 1992 89 3429 3433 1565634 Hackett CJ Sharma OK Frontiers in peptide-MHC class II multimer technology Nat Immunol 2002 3 887 889 12352960 10.1038/ni1002-887 Latek RR Petzold SJ Unanue ER Hindering auxiliary anchors are potent modulators of peptide binding and selection by I-Ak class II molecules Proc Natl Acad Sci U S A 2000 97 11460 11465 11016975 10.1073/pnas.210384197 Cameron TO Norris PJ Patel A Moulon C Rosenberg ES Mellins ED Wedderburn LR Stern LJ Labeling antigen-specific CD4(+) T cells with class II MHC oligomers J Immunol Methods 2002 268 51 69 12213343 10.1016/S0022-1759(02)00200-4 Savage PA Boniface JJ Davis MM A kinetic basis for T cell receptor repertoire selection during an immune response Immunity 1999 10 485 492 10229191 10.1016/S1074-7613(00)80048-5 Gutgemann I Fahrer AM Altman JD Davis MM Chien YH Induction of rapid T cell activation and tolerance by systemic presentation of an orally administered antigen Immunity 1998 8 667 673 9655480 10.1016/S1074-7613(00)80571-3 Crawford F Kozono H White J Marrack P Kappler J Detection of antigen-specific T cells with multivalent soluble class II MHC covalent peptide complexes Immunity 1998 8 675 682 9655481 10.1016/S1074-7613(00)80572-5 Malherbe L Filippi C Julia V Foucras G Moro M Appel H Wucherpfennig K Guery JC Glaichenhaus N Selective activation and expansion of high-affinity CD4+ T cells in resistant mice upon infection with Leishmania major Immunity 2000 13 771 782 11163193 10.1016/S1074-7613(00)00075-3 Appel H Gauthier L Pyrdol J Wucherpfennig KW Kinetics of T-cell receptor binding by bivalent HLA-DR.Peptide complexes that activate antigen-specific human T-cells J Biol Chem 2000 275 312 321 10617620 10.1074/jbc.275.1.312 Nepom GT Buckner JH Novak EJ Reichstetter S Reijonen H Gebe J Wang R Swanson E Kwok WW HLA class II tetramers: tools for direct analysis of antigen-specific CD4+ T cells Arthritis Rheum 2002 46 5 12 11817608 10.1002/1529-0131(200201)46:1<5::AID-ART10063>3.0.CO;2-S Novak EJ Liu AW Nepom GT Kwok WW MHC class II tetramers identify peptide-specific human CD4(+) T cells proliferating in response to influenza A antigen J Clin Invest 1999 104 R63 7 10606632 Reichstetter S Ettinger RA Liu AW Gebe JA Nepom GT Kwok WW Distinct T cell interactions with HLA class II tetramers characterize a spectrum of TCR affinities in the human antigen-specific T cell response J Immunol 2000 165 6994 6998 11120826 Danke NA Kwok WW HLA class II-restricted CD4+ T cell responses directed against influenza viral antigens postinfluenza vaccination J Immunol 2003 171 3163 3169 12960344 Lebowitz MS O'Herrin SM Hamad AR Fahmy T Marguet D Barnes NC Pardoll D Bieler JG Schneck JP Soluble, high-affinity dimers of T-cell receptors and class II major histocompatibility complexes: biochemical probes for analysis and modulation of immune responses Cell Immunol 1999 192 175 184 10087186 10.1006/cimm.1999.1441 Arnold PY Vignali KM Miller TB La Gruta NL Cauley LS Haynes L Scott Adams P Swain SL Woodland DL Vignali DA Reliable generation and use of MHC class II:gamma2aFc multimers for the identification of antigen-specific CD4(+) T cells J Immunol Methods 2002 271 137 151 12445737 10.1016/S0022-1759(02)00343-5 Schatz PJ Use of peptide libraries to map the substrate specificity of a peptide-modifying enzyme: a 13 residue consensus peptide specifies biotinylation in Escherichia coli Biotechnology (N Y) 1993 11 1138 1143 7764094 Stockel J Meinl E Hahnel C Malotka J Seitz R Drexler K Wekerle H Dornmair K Refolding of human class II major histocompatibility complex molecules isolated from Escherichia coli. Assembly of peptide-free heterodimers and increased refolding-yield in the presence of antigenic peptide J Biol Chem 1994 269 29571 29578 7961943 Arimilli S Cardoso C Mukku P Baichwal V Nag B Refolding and reconstitution of functionally active complexes of human leukocyte antigen DR2 and myelin basic protein peptide from recombinant alpha and beta polypeptide chains J Biol Chem 1995 270 971 977 7529765 10.1074/jbc.270.2.971 Panina-Bordignon P Tan A Termijtelen A Demotz S Corradin G Lanzavecchia A Universally immunogenic T cell epitopes: promiscuous binding to human MHC class II and promiscuous recognition by T cells Eur J Immunol 1989 19 2237 2242 2481588 Sadegh-Nasseri S Germain RN A role for peptide in determining MHC class II structure Nature 1991 353 167 170 1653903 10.1038/353167a0 Cameron TO Cochran JR Yassine-Diab B Sekaly RP Stern LJ Cutting edge: detection of antigen-specific CD4+ T cells by HLA-DR1 oligomers is dependent on the T cell activation state J Immunol 2001 166 741 745 11145645 Eckels DD Sell TW Bronson SR Johnson AH Hartzman RJ Lamb JR Human helper T-cell clones that recognize different influenza hemagglutinin determinants are restricted by different HLA-D region epitopes Immunogenetics 1984 19 409 423 6202629 10.1007/BF00364644 Lucas M Day CL Wyer JR Cunliffe SL Loughry A McMichael AJ Klenerman P Ex vivo phenotype and frequency of influenza virus-specific CD4 memory T cells J Virol 2004 78 7284 7287 15194806 10.1128/JVI.78.13.7284-7287.2004 Consogno G Manici S Facchinetti V Bachi A Hammer J Conti-Fine BM Rugarli C Traversari C Protti MP Identification of immunodominant regions among promiscuous HLA-DR-restricted CD4+ T-cell epitopes on the tumor antigen MAGE-3 Blood 2003 101 1038 1044 12393675 10.1182/blood-2002-03-0933 Harcourt GC Lucas M Sheridan I Barnes E Phillips R Klenerman P Longitudinal mapping of protective CD4+ T cell responses against HCV: analysis of fluctuating dominant and subdominant HLA-DR11 restricted epitopes J Viral Hepat 2004 11 324 331 15230855 10.1111/j.1365-2893.2004.00516.x Protti MP Manfredi AA Straub C Wu XD Howard JFJ Conti-Tronconi BM Use of synthetic peptides to establish anti-human acetylcholine receptor CD4+ cell lines from myasthenia gravis patients J Immunol 1990 144 1711 1720 1968487	16329759	PMC1325046	CC BY	2021-01-04 16:03:32	no	BMC Immunol. 2005 Dec 5; 6:24	utf-8	BMC Immunol	2,005	10.1186/1471-2172-6-24	oa_comm
==== Front J Transl MedJournal of Translational Medicine1479-5876BioMed Central London 1479-5876-3-421631863410.1186/1479-5876-3-42ResearchProtein kinase activity is associated with CD63 in melanoma cells Iida Joji [email protected] Amy PN [email protected] James B [email protected] Keith M [email protected] Department of Laboratory Medicine and Pathology, The University of Minnesota Medical School, Minneapolis, MN 55455, USA2 Department of Medicine, The University of Minnesota Medical School and the Masonic Cancer Center, Minneapolis, MN 55455, USA2005 30 11 2005 3 42 42 12 8 2005 30 11 2005 Copyright © 2005 Iida et al; licensee BioMed Central Ltd.2005Iida 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 tetraspan protein CD63, originally described as a stage-specific melanoma antigen but also present in a number of normal cells, regulates melanoma cell growth in nude mice, motility in serum containing media, and adhesion to several extracellular matrix proteins. CD63 has been reported to associate with β1 and β2 integrins, but the mechanism of signal transduction by CD63 is not clear. This study examined whether CD63 is associated with protein kinase and can transmit signals in melanoma cells. Methods Immunoprecipitation and radiolabeling were used to test for association of protein kinase activity with CD63. Adhesion of cells to monoclonal antibodies immobilized to microtiter plates was used to examine the ability of CD63 to transmit signals. Results CD63 was capable of transmitting a signal in melanoma cells that required extracellular calcium. In the absence of extracellular calcium at the time of binding to the CD63 mAb, the cell was no longer responsive to stimulation by CD63. Immunoprecipitation studies demonstrated protein kinase activity associated with CD63, and phosphoamino acid analysis revealed that most of this protein kinase activity was due to serine kinase activity. Conclusion The current study suggests that serine protein kinase activity associated with CD63 may play a role in signaling by CD63 in melanoma cells. ==== Body Background CD63 was initially described as the ME491 antigen of melanoma cells [1-8]. CD63 is not expressed in normal melanocytes, but is expressed in nevi and many melanoma cells. CD63 is also expressed in a number of normal cells [1-5,9-18]. A number of studies have suggested that CD63 may regulate melanoma cell functions. Transfection with CD63 of NIH-3T3 cells transformed with H-ras, resulted in cells with lower growth rates following subcutaneous injection in nude mice [19]. Transfection of the CD63 negative melanoma cell line KM3 with genomic CD63 resulted in cells with similar in vitro growth rates as control cells, but much slower growth rates in vivo when cells were injected intradermally in nude mice [20]. In addition, intravenous injection of CD63 transfected KM3 cells resulted in fewer peritoneal and subcutaneous metastases [20]. These data, therefore, suggest that CD63 may regulate in vivo growth and metastatic capability [20]. Finally, transfection of the highly motile KM3 cells with CD63 resulted in suppression of in vitro motility in serum-containing media that was potentiated by CD63 mAbs, and increased adhesion and migration on fibronectin, laminin, and collagen [21]. In contrast, transfection of antisense CD63 cDNA into melanoma cells endogenously expressing CD63, resulted in increased cell motility and invasiveness in vitro[22]. Thus, CD63 appears capable of regulating melanoma cell functions, although the mechanism of this regulation is unclear. CD63 is a member of the tetraspan family, traversing the membrane four times and having a major extracellular loop and a small extracellular loop, with short intracellular amino and carboxy termini [1,2,4,5,14,18,23-25]. The intracellular domains of CD63 are small and have no clear motif known to be involved in signal transmission. In this study, the properties of the adhesion of melanoma cells to immobilized CD63 mAb suggested that CD63 was capable of transmitting a signal that requires extracellular calcium. Protein kinase activity, predominantly in the form of serine kinase activity, was found to be associated with CD63 in these cells. The association of serine protein kinase activity with CD63 may play a role in signaling by CD63 in melanoma. Methods Cell Culture Highly metastatic human melanoma cells, A375SM, which were selected by in vivo experimental metastasis assays of parent A375P cells in nude mice [26] were kindly provided by Dr. I. J. Fidler (M.D. Anderson Hospital Cancer Center, Houston, TX). The cells were maintained in Eagle's Minimum Essential Medium (EMEM) supplemented with 10% FBS, vitamin solution, 50 mg/ml gentamycin, and 1 mM sodium pyruvate. Cells were routinely used after fewer than 15 passages from frozen stocks in order to minimize phenotypic drift. Monoclonal antibodies (mAbs) The CD63 mAbs, AHN-16 (IgG2a), AHN-16.1 (IgG1), AHN-16.2 (IgG2b), AHN-16.3 (IgG2b), and AHN-16.5 (IgG1) were previously described [15]. mAb U131 (IgG1) was previously described [27]. mAb 9.2.27 that recognizes the core protein of melanoma associated antigen (MCSP) [28] was a gift of Dr. Ralf Reisfeld (Scripps Clinic, La Jolla, CA). Fab fragments of AHN-16 were prepared by using immobilized papain according to the manufacturer's instruction (Pierce, Rockford, IL), and their purity was confirmed by SDS-PAGE analysis. Cell adhesion assays Cell adhesion assays to immobilized mAbs were performed as described previously [29]. Briefly, 96-well plates were coated with 50 ul of goat anti-mouse IgG Fc (Organon Teknika Corp., Durham, NC) overnight at 37°C and then blocked with PBS containing 2 mg/ml BSA for 3 h at 37°C. The purified mAbs were diluted in PBS, dispersed into each well at 1 ug/ml in 50 ul, incubated for 3 to 4 h at 37°C, and then washed 2 times with PBS. Subconfluent A375SM cells that had been radiolabeled overnight with 3H-thymidine (3H-TdR, specific activity 6.7 Ci/mmol; NEN Research Products, Boston, MA), were harvested by rinsing with PBS containing 1 mM EDTA, washed two times with EMEM/BSA (EMEM containing 2 mg/ml BSA and 15 mM Hepes, pH 7.2), and adjusted to a concentration of 105 cells/ml in the same medium. When cell adhesion assays were carried out in the presence of 10 mM EDTA or EGTA, cells were resuspended in EMEM/BSA containing 10 mM EDTA or EGTA. An aliquot of 100 ul of the cell suspension was dispersed into each well and incubated for 60 min at 37°C. For inhibiting cell adhesion to the immobilized AHN-16, cells were preincubated with Fab fragments of AHN-16 or of normal mouse IgG for 20 min at room temperature. Cell adhesion assays were carried out in the continuous presence of these Fab fragments for 60 min at 37°C. The assays were terminated by aspirating loosely bound and unbound cells from the wells. Bound radioactivity was determined in a liquid scintillation counter (Beckman Model 3801 Liquid Scintillation Counter). Flow cytometry Cells were harvested as described above, and washed 3 times with RPMI 1640 containing 1% heat-inactivated goat serum, 20 mM HEPES, and 0.02% NaN3(FACS buffer). Cells were resuspended in the FACS buffer at a concentration of 105 cells/ml with 10 ug/ml normal goat IgG for 30 min at 4°C and then washed 3 times with the FACS buffer. An aliquot of 1 ml of cell suspension was incubated with 5 ug/ml of mAbs for 30 min at 4°C with mixing every 5 min. Cells were washed 3 times with the FACS buffer and then incubated with FITC conjugated goat anti-mouse IgG (final dilution of 1: 500) at 4°C for 40 min with mixing every 5 min. The binding of the primary and the secondary mAbs were maximal under these experimental conditions. Cells were washed 3 times with the FACS buffer and resuspended in 200 ul of PBS containing 2% formaldehyde. Cells were analyzed on a FACS Star (Becton-Dickinson, Mt. View, CA). Protein kinase assay Immunoprecipitation was performed as previously described with minor modifications [30]. Briefly, 2 × 107 cells were suspended in 1.1 ml of Brij solubilization buffer [20 mM Tris-HCl, pH 8.2, 150 mM NaCl, 1 mM PMSF, 2 mM MgCl2, 0.02% NaN3, and 1.0% Brij 58 (Pierce)], and incubated on ice for 15 min. The suspensions were then centrifuged at 9,800 × g for 15 min at 4°C. The resulting supernatants were used for immunoprecipitation or analyzed by SDS-PAGE directly. For immunoprecipitations, 1 ml of 10% Staphylococcusaureus (Pansorbin A, Calbiochem, La Jolla, CA) was mixed with 100 μl of rabbit anti-mouse IgGH+L (Organon Teknika, Westchester, PA), and the mixture was incubated at 4°C for 1 hr. One ml of buffer A (1 mg/ml BSA, 0.05% NP-40, 20 mM Tris-HCl, pH 7.6, 100 mM NaCl, 1 mM EDTA, 0.05% NaN3) was added and the Pansorbin A-antibody complex was recovered by centrifuging at 2000 × g for 10 min at 4°C and resuspended in 390 μl of buffer A and 40 μl of ascites containing the indicated monoclonal antibody, or normal mouse serum (NMS) was added, and the mixtures were incubated for 2 hr at 4°C. The Pansorbin A-antibody complex was recovered by centrifuging at 2000 × g for 10 min at 4°C, washed once with 1 ml of buffer A, and resuspended in 1 ml of Brij-SA buffer (1 mg/ml BSA, 0.05% Brij 58, 20 mM Tris-HCl, pH 7.6, 100 mM NaCl, 1 mM EDTA, 0.05% NaN3). Cell proteins were then immunoprecipitated in reaction mixtures containing the cell extract, Pansorbin A-bound antibody complex, 20 mM Tris-HCl, pH 8.2, 100 mM NaCl, 0.5% Brij 58, 1 mM EDTA, 0.125 mg/ml gelatin, and 1 mM PMSF in a total volume of approximately 350 μl in 10 × 75 mm glass tubes. After 1 hr at 4°C, the mixture was washed three times with Brij wash buffer (20 mM Tris-HCl, pH 8.2, 150 mM NaCl, 1 mg/ml BSA, 0.5% Brij 58, 2 mM MgCl2, 0.125 mg/ml gelatin, 1 mM PMSF, and 0.02% NaN3) and then once with NaCl-HEPES (145 mM NaCl, 20 mM HEPES, pH 7.3). Immunoprecipitates were suspended in 30 μl of NaCl-HEPES. Thirty ul of labeling buffer (0.1% Brij 58, 6 mM MnCl2, 40 mM MgCl2, 200 uM Na3VO4, 200 uM Na2MoO4, and 10 μCi of [γ-32P]ATP) was then added, and the mixtures were incubated for 10 min at 23°C. The reaction was stopped by adding 4 × Laemmli sample buffer, incubated at 100°C for 2 min, and analyzed by SDS-PAGE [31]. Molecular weight standards were purchased from Sigma. Gel slabs were stained, dried, and subjected to autoradiography by using Dupont Cronex film. Phosphoamino acid analysis Phosphoamino acid analyses were performed as previously described [32,33]. Briefly, radiolabeled proteins resolved by SDS-PAGE were transblotted onto Immobilon-P (PVDF) paper (Millipore Corp, Bedford, MA), localized by autoradiography, and excised. The proteins on the PVDF were then hydrolyzed in vacuo in constantly boiling 6 M HCl for 2 hr at 110°C. The HCl was removed by evaporation using a SpeedVac (Savant Instruments). The dried, partially hydrolyzed samples were then dissolved in water containing phosphoserine (P-S), phosphothreonine (P-T), and phosphotyrosine (P-Y) (Sigma) each at 1 mg/ml. The samples were then spotted on aluminum sheets precoated with silica gel (0.2 mm layer thickness, 11 cm height, Merck Laboratories, EM Science, Gibbstown, NJ) and resolved by one-dimensional thin layer chromatography in ethanol: 25% NH4OH (3.5:1.6). The chromatography cycle was repeated three times to achieve optimal separation. Between cycles, plates were air dried followed by chromatography in the same solvent. The radiolabeled phosphoamino acids were detected by autoradiography using X-Omat AR film. Results Human melanoma cells adhere to immobilized CD63 mAbs To establish a simple assay of CD63 signaling in melanoma cells, we tested several CD63 mAbs including AHN-16, AHN-16.1, AHN-16.2, AHN-16.3, and AHN-16.5 for their ability to promote A375SM human melanoma cell adhesion as previously described [29]. CD63 expression on these cells was confirmed using flow cytometry with five CD63 mAbs as described in the Methods (data not shown). Each of the mAbs was immobilized on plastic microtiter plates as described in the Materials and Methods. Cell adhesion assays were carried out for 30 min. At coating concentrations of 1 ug/ml, each of the CD63 mAbs promoted melanoma cell adhesion (Table 1). In contrast, the control antibodies normal mouse IgG (IgG) and the unclustered mAb U131 did not. As expected, the anti-MCSP mAb 9.2.27 promoted high levels of cell adhesion, consistent with our previous results [29]. AHN-16 promoted cell adhesion in a concentration-dependent manner (Fig 1). As expected [29], an anti-melanoma associated proteoglycan mAb (9.2.27) also promoted cell adhesion in a concentration-dependent manner (Fig 1). Normal mouse IgG failed to promote melanoma cell adhesion (Fig 1). Table 1 Human melanoma cell adhesion to immobilized CD63 mAbs mAba Specificity Cell adhesion (%)b AHN-16 CD63 59 AHN-16.1 CD63 29 AHN-16.2 CD63 55 AHN-16.3 CD63 63 AHN-16.5 CD63 29 U131 not clustered 3 IgGc N/A 4 9.2.27 MCSP 88 a. Each mAb was coated on plates at a concentration of 1 ug/ml. b. Assays of adhesion of A375SM melanoma cells to mAbs immobilized on plastic were performed for 30 min. as described in the text. Cell adhesion is expressed as the percent of the total number of cells added. Values are the means of four replicates that agreed within 10 %. A duplicate experiment gave similar results. c. Normal mouse IgG Figure 1 Concentration dependence of A375SM melanoma cell adhesion to immobilized CD63 mAb. Plates were coated with various concentrations of the CD63 mAb (AHN-16) (open circles), the anti-MCSP mAb 9.2.27 (open squares), or normal mouse IgG (open triangles) as described in Materials and Methods. Cell adhesion assays were performed for 30 min. Results are shown as percent adhesion +/- SEM. A duplicate experiment gave similar results. In order to confirm that the cell adhesion to immobilized AHN-16 was due to reactivity with the CD63 epitope, Fab fragments of AHN-16 and normal mouse IgG were prepared as described in the Materials and Methods, and tested for their ability to inhibit melanoma cell adhesion to immobilized AHN-16. Fab fragments of AHN-16, but not of normal mouse IgG, effectively inhibited cell adhesion to AHN-16 in a concentration dependent manner (Fig 2). At the highest concentration of AHN-16 Fab fragment (20 ug/ml), cell adhesion was inhibited by 90%. These results confirmed that cell adhesion to immobilized AHN-16 was dependent on the antigen binding site. Figure 2 Melanoma cell adhesion to immobilized CD63 mAb was inhibited by Fab fragments. Plates were coated with 0.1 ug/ml of the CD63 mAb AHN-16 as described in Materials and Methods. Cells were preincubated with various concentrations of Fab fragments of AHN-16 (open circles) or of normal mouse IgG (closed circles) for 20 min before adding the cells to the plates for the cell adhesion assays. Cell adhesion assays were performed for 30 min. Results are shown as percent adhesion +/- SEM. A duplicate experiment gave similar results. Human melanoma cell adhesion to immobilized CD63 mAb is inhibited by chelerythrine In order to study the signaling pathways by which CD63 may transmit signals, cell adhesion assays were carried out in the presence of chelerythrine, a potent and specific inhibitor of protein kinase C [34]. Melanoma cells were preincubated with various concentrations of chelerythrine or genistein, a selective tyrosine protein kinase inhibitor, for 20 min at 37°C prior to cell adhesion assays. Chelerythrine inhibited melanoma cell adhesion to immobilized AHN-16 in a concentration dependent manner with an IC50 of approximately 2.5 uM (Fig 3). Chelerythrine did not alter surface expression of CD63 as determined by flow cytometry (not shown). In contrast, genistein did not affect cell adhesion to immobilized AHN-16 under these conditions at concentrations up to 50 ug/ml (not shown). However, cell adhesion to mAb 9.2.27 was totally resistant to both chemicals (not shown). These results suggested that protein kinase C plays a role in cell adhesion to immobilized CD63 mAb. Figure 3 Chelerythrine inhibited cell adhesion to immobilized CD63 mAb. Plates were coated with 0.1 ug/ml of the CD63 mAb AHN-16 or the anti-MCSP mAb 9.2.27 as described in Materials and Methods. Cells were preincubated with various concentrations of chelerythrine for 20 min before adding the cells to the plates for the cell adhesion assays. Cell adhesion assays were performed for 30 min. Results are shown as percent adhesion +/- SEM. A duplicate experiment gave similar results. Human melanoma cell adhesion to immobilized CD63 mAb requires external Ca2+ We next examined the effects of extracellular Ca2+ on melanoma cell adhesion to immobilized CD63 mAbs. Human melanoma cell adhesion to three different immobilized CD63 mAbs was almost totally inhibited by EGTA, indicating that external Ca2+ plays an important role in human melanoma cell adhesion to the mAbs (Fig 4, top). Cells did not adhere to immobilized CD63 mAbs in the absence of calcium, suggesting that this adhesion is an active process. In contrast, cell adhesion to immobilized mAb 9.2.27 was totally resistant to EGTA (Fig 4A, top). EGTA itself was not inhibitory since cell adhesion to AHN-16 was similar to control cell adhesion levels when a normal Ca2+ level was generated by adding 10 mM Ca2+ to the cells in EGTA (Fig 4B, bottom). EGTA did not interfere with the interaction between human melanoma cells and AHN-16, since AHN-16 bound the cell surface normally as assessed by flow cytometery (not shown). Figure 4 EGTA inhibited cell adhesion to immobilized CD63 mAb. Top panel: plates were coated with 0.1 ug/ml of the indicated mAb and cell adhesion assays were carried out in the absence (hatched bars) or presence (solid bars) of 10 mM EGTA for 30 min as described in the Materials and Methods. CD63 mAbs: AHN-16, AHN-16.1, AHN-16.5; unclustered control mAb: U131, anti-MCSP mAb: 9.2.27. Bottom panel; plates were coated with 0.1 ug/ml of the CD63 mAb AHN-16 as described in the Materials and Methods, and cell adhesion assays were performed in either 10 mM EGTA, 10 mM EGTA plus 5 mM Ca2+, or 10 mM EGTA plus 10 mM Ca2+ for 30 min as indicated. Results are shown as percent adhesion +/- SEM. A duplicate experiment gave similar results. In some cells, for example human neutrophils, the binding of a stimulus to its receptor on the cell surface results in the activation of a signal transduction system that is transient, and whose function is dependent on extracellular Ca2+ [35,36]. When calcium is absent during a critical period following ligand-receptor coupling, the subsequent addition of calcium does not result in cell activation, and the cells may be partially desensitized to subsequent stimulation in the presence of Ca2+ [35,36]. The role of extracellular Ca2+ was therefore further evaluated. When calcium (10 mM) was added to the cells in medium with EGTA immediately after adding them to the plate, adhesion was normal. However, if the cells were allowed to interact with the immobilized CD63 mAbs for 20 min before repleting the calcium, adhesion was not restored by the calcium (not shown). Identification of protein kinase activity associated with CD63 Since protein phosphorylation appears to be involved in signaling by CD63 in A375SM cells, we questioned whether protein kinase activity could be associated with CD63 in A375SM cells as has been observed in neutrophils [15]. A375SM melanoma cells were solubilized in a buffer containing Brij 58 as described in the Methods, and the solubilized material was immunoprecipitated by CD63 mAbs. When [γ-32P]ATP was added to the material immunoprecipitated by the CD63 mAb AHN-16 (Fig 5, lane 2) or the CD63 mAb AHN-16.1 (not shown), 32P was reproducibly incorporated into three distinct proteins of ~208 to 236-kD, 53 to 58-kD, and 46 to 50-kD (labeled 1–3), that were not present when material was immunoprecipitated by NMS (Fig 5, lane 1). In contrast, 32P was reproducibly incorporated into a ~66-kD protein (labeled 4) in material immunoprecipitated by mAb 9.2.27 (Fig 5, lane 3). Each of these four individual proteins were examined for phosphoamino acid content as described in the Methods. The majority of radiolabel in each of these four proteins was present on serine residues (Fig 6), demonstrating the presence of serine kinase activity. Figure 5 Co-immunoprecipitation of protein kinase activity with CD63. Panel A, Cells were solubilized in Brij solubilization buffer and immunoprecipitated with NMS (lane 1), the CD63 mAb AHN-16 (lane 2), or the anti-MCSP mAb 9.2.27 (lane 3), and the immunoprecipitates were incubated with [γ-32P]ATP as described in the text. The resulting phosphoproteins were resolved by SDS-PAGE and visualized by autoradiography as described in the text. Three phosphoproteins were reproducibly identified in CD63 mAb immunoprecipitates (labeled 1–3 in lane 2). One phosphoprotein was reproducibly identified in immunoprecipitates using the anti-MCSP mAb 9.2.27 (labeled 4 in lane 3). These four phosphoproteins were not seen in the immunoprecipitate using NMS (lane 1). Proteins used as molecular weight standards were: myosin heavy chain, 200,000; Escherichia coli β-galactosidase, 116,000; bovine serum albumin, 66,000; ovalbumin, 45,000; and carbonic anhydrase, 29,000. Figure 6 Phosphoamino acid analysis of proteins co-precipitated with the CD63 mAb AHN-16 and the anti-MCSP mAb 9.2.27, and labeled in an in vitro kinase assay. Cells were solubilized, immunoprecipitated, labeled with [γ32P-]ATP, and the resulting phosphoproteins were resolved by SDS-PAGE as in Fig 5. Phosphoproteins were transferred to Immobilon by electroblotting, visualized by autoradiography, excised, and subjected to acid hydrolysis. The resultant phosphoamino acids were resolved by thin layer chromatography and visualized by autoradiography as described in the text. The positions of migration of authentic phosphoserine (P-S), phosphothreonine (P-T), and phosphotyrosine (P-Y) are indicated by dotted lines. A duplicate experiment gave similar results. The lane numbers correspond to the phosphoproteins numbered in figure 5. Discussion Immunohistochemistry studies have characterized CD63 as the stage-specific melanoma-associated antigen ME491 [6-8]. Over expression of CD63 partially suppressed malignant phenotypes of H-ras-transformed fibroblasts in vivo[19], and transfection of melanoma cell lines also suggested that CD63 can regulate melanoma cell function [20,21]. The current study also demonstrates that CD63 can transduce signals in melanoma cells and alter melanoma cell behavior. In addition, this study demonstrates that CD63 is capable of transducing signals in melanoma cells that requires extracellular Ca2+ and is inhibited by the protein kinase C inhibitor chelerythrine. Signaling by tetraspans, including the induction of calcium-dependent cell-cell adhesion, has been reported in other systems [37-39]. To examine potential signal transducing pathways through which CD63 may act, an in vitro assay measuring melanoma cell adhesion to immobilized mAbs was used. We previously utilized this method for characterizing molecular mechanisms of signal transduction from proteoglycan as well as integrin receptors in melanoma cells [29,40]. Melanoma cells specifically adhered to immobilized CD63 mAb, and the adhesion was inhibited by chelerythrine but not by genistein, suggesting that signaling pathways involving protein kinase C activity, but not genistein-sensitive tyrosine kinases, play a role in the stimulation by CD63. The inhibitory activity of chelerythrine was not due to cytotoxicity, since the chemicals did not affect cell adhesion to immobilized mAb 9.2.27. The IC50 value of chelerythrine is close to the value that inhibits the activity of protein kinase C in vitro[34], supporting the conclusion that the inhibitory effect of chelerythrine on cell adhesion to immobilized CD63 mAb is of a specific nature. CD63 may interact with different signaling molecules in different cell types. In human neutrophils, tyrosine protein kinase activity has been reported to be associated with CD63; most of this tyrosine kinase activity was found to be due to the src family tyrosine kinases lyn and hck which coimmunoprecipitated with CD63 in neutrophils, suggesting that these kinases may play a role in signaling via CD63 [15]. In the current study, we found serine protein kinase activity, but could not detect tyrosine kinase activity, in immunoprecipitates using CD63 mAbs. Recent studies have reported that CD63 in platelets modulates spreading and tyrosine kinase activity [41]. Thus, these results indicate that the interaction between CD63 and these kinases is dependent upon the cell type. These data also support a role for protein kinase C in CD63 signaling in melanoma cells. CD63 has also been found to associate with the β2 integrin CD18 in neutrophils and the β1 integrin in several other cells, including melanoma cells, and this may also play a role in signaling by CD63 [15,42-45] Melanoma cell adhesion to immobilized CD63 mAb required extracellular calcium, indicating that extracellular Ca2+ plays an important role in signaling by CD63 in these cells. Addition of calcium to cells 20 min after interaction of cells with immobilized CD63 mAb resulted in no adhesion. Thus, the interaction of melanoma cells expressing CD63 to immobilized CD63 mAbs resulted in the generation of a transient signal that required extracellular calcium to effect melanoma cell adhesion. These observations are similar to those reported with human neutrophils where ligation of CD63 results in a transient activation signal that requires extracellular calcium [15]. Conclusion The current study demonstrates that CD63 is capable of transducing signals in melanoma cells that requires extracellular Ca2+ and is inhibited by the protein kinase C inhibitor chelerythrine, and that CD63 is associated with serine protein kinase activity in melanoma cells. This associated protein kinase activity may play a role in the mechanism whereby CD63 exerts its previously reported effects on melanoma cell function [20,21]. Abbreviations Abbreviations used in this paper: EMEM, Eagle's minimum essential medium; NMS, normal mouse serum; FACS buffer, RPMI 1640 containing 1% heat-inactivated goat serum, 20 mM HEPES, and 0.02% NaN3; NaCl-HEPES, 145 mM NaCl, 20 mM HEPES, pH 7.3; Brij solubilization buffer, (20 mM Tris-HCl, pH 8.2, 150 mM NaCl, 1 mM PMSF, 2 mM MgCl2, 0.02% NaN3, and 1.0% Brij 58); Brij wash buffer, (20 mM Tris-HCl, pH 8.2, 150 mM NaCl, 1 mg/ml BSA, 0.5% Brij 58, 2 mM MgCl2, 0.125 mg/ml gelatin, 1 mM PMSF, and 0.02% NaN3); labeling buffer, (0.1% Brij 58, 6 mM MnCl2, 40 mM MgCl2, 200 uM Na3VO4, 200 uM Na2MoO4, and 10 μCi of [γ-32P]ATP); buffer A, (1 mg/ml BSA, 0.05% NP-40, 20 mM Tris-HCl, pH 7.6, 100 mM NaCl, 1 mM EDTA, 0.05% NaN3); buffer B, (PBS, pH 7.4, 0.2% BSA, 0.05% NaN3); Brij SA buffer, (1 mg/ml BSA, 0.05% Brij 58, 20 mM Tris-HCl, pH 7.6, 100 mM NaCl, 1 mM EDTA, 0.05% NaN3, 1 mM Na3VO4, and 1 mM Na2MoO4) Competing interests The author(s) declare that they have no competing interests. Authors' contributions GI helped conceive the study, carried out the adhesion assays and antibody purification, participated in experimental design and data analysis, and helped draft the manuscript APNS helped conceive the study, design experiments, analyze data, and draft the manuscript JBM helped analyze data and draft the manuscript KMS helped conceive the study, design experiments, analyze data, and draft the manuscript Acknowledgements We thank I. Fidler for providing cell lines, K. Campbell for assistance with protein kinase assays, K. Ahmed and A. Davis for assistance with the phosphoamino acid analysis, and R. Reisfeld for providing mAb 9.2.27. Supported in part by the American Heart Association, Minnesota Affiliate, the Office of the Vice President for Research and Dean of the Graduate School of the University of Minnesota, the Minnesota Medical Foundation, and the Masonic Memorial Hospital Fund, Inc. ==== Refs Metzelaar MJ Wijngaard PLJ Peters PJ Sixma JJ Nieuwenhuis HK Clevers HC CD63 Antigen Journal of Biological Chemistry 1991 266 3239 3245 1993697 Azorsa DO Hyman JA Hildreth JEK CD63/Pltgp40: A platelet activation antigen identical to the stage-specific, melanoma-associated antigen ME491 Blood 1991 78 280 284 2070066 Hotta H Ross AH Huebner K Isobe M Wendeborn S Chao MV Ricciardi RP Tsujimoto Y Croce CM Koprowski H Molecular cloning and characterization of an antigen associated with early stages of melanoma tumor progression Cancer Research 1988 48 2955 2962 3365686 Hotta H Miyamoto H Hara I Takahashi N Homma M Genomic structure of the ME491/CD63 antigen gene and functional analysis of the 5'-flanking regulatory sequences Biochemical and Biophysical Research Communications 1992 185 436 442 1599482 10.1016/S0006-291X(05)81004-6 Nishikata H Oliver C Mergenhagen SE Siraganian RP The rat mast cell antigen AD1 (homologue to human CD63 or melanoma antigen ME491) is expressed in other cells in culture Journal of Immunology 1992 149 862 870 Kondoh M Ueda M Ichihashi M Mishima Y Decreased expression of human melanoma-associated antigen ME491 along the progression of melanoma pre-canceroses to invasive and metastatic melanomas Melanoma Res 1993 3 241 245 8219756 Atkinson B Ernst CS Ghrist BF Ross AH Clark WH Herlyn M Herlyn D Maul G Steplewski Z Koprowski H Monoclonal antibody to a highly glycosylated protein reacts in fixed tissue with melanoma and other tumors Hybridoma 1985 4 243 255 4043988 Sikora LK Pinto A Demetrick DJ Dixon WT Urbanski SJ Temple W Jerry LM Characterization of a novel neuroglandular antigen (NGA) expressed on abnormal human melanocytes Int J Cancer 1987 39 138 145 3804488 Metzelaar MJ Schuurman HJ Heijnen HFG Sixma JJ Nieuwenhuis HK Biochemical and immunohistochemical characteristics of CD62 and CD63 monoclonal antibodies Virchows Archiv (Berlin) 1991 61 269 277 Nieuwenhuis HK van Oosterhout JJG Rozemuller E van Iwaarden F Sixma JJ Studies with a monoclonal antibody against activated platelets: Evidence that a secreted 53,000-molecular weight lysosome-like granule protein is exposed on the surface of activated platelets in the circulation Blood 1987 70 838 845 3620703 Knol EF Mul FPJ Jansen H Calafat J Roos D Monitoring human basophil activation via CD63 monoclonal antibody 435 Journal of Allergy and Clinical Immunology 1991 88 328 338 1716273 10.1016/0091-6749(91)90094-5 Peters PJ Borst J Oorschot V Fukuda M Krahenbuhl O Tschopp J Slot JW Geuze HJ Cytotoxic T lymphocyte granules are secretory lysosomes, containing both perforin and granzymes Journal of Experimental Medicine 1991 173 1099 1109 2022921 10.1084/jem.173.5.1099 Kuijpers TW Tool ATJ van der Schoot EC Ginsel LA Onderwater JJM Roos DVAJ Membrane surface antigen expression on neutrophils: A reappraisal of the use of surface markers for neutrophils activation Blood 1991 78 1105 1111 1907873 Horejsí V Vlcek C Novel structurally distinct family of leucocyte surface glycoproteins including CD9, CD37, CD53, and CD63 FEBS Letters 1991 288 1 4 1879540 10.1016/0014-5793(91)80988-F Skubitz KM Campbell KD Iida J Skubitz APN CD63 associates with tyrosine kinase activity and CD11/CD18, and transmits an activation signal in neutrophils Journal of Immunology 1996 157 3617 3626 Vischer UM Wagner DD CD63 is a component of Weibel-Palade bodies of human endothelial cells Blood 1993 82 1184 1191 8353283 Toothill VJ van Mourik JA Niewenhuis HK Metzelaar MJ Pearson JD Characterization of the enhanced adhesion of neutrophil leukocytes to thrombin-stimulated endothelial cells Journal of Immunology 1990 145 283 291 Nishibori M Cham B McNicol A Shalev A Jain N Gerrard JM The protein CD63 is in platelet dense granules, is deficient in a patient with Hermansky-Pudlak syndrome, and appears identical to granulophysin Journal of Clinical Investigation 1993 91 1775 1782 7682577 Hotta H Hara I Miyamoto H Homma M Overexpression of the human melanoma-associated antigen ME491 partially suppresses in vivo malignant phenotypes of H-ras-transformed NIH3T3 cells in athymic nude mice Melanoma Res 1991 1 125 132 1823625 Radford KJ Mallesch J Hersey P Suppression of human melanoma cell growth and metastasis by the melanoma-associated antigen CD63 (ME491) Int J Cancer 1995 62 631 635 7665237 Radford KJ Thorne RF Hersey P Regulation of tumor cell motility and migration by CD63 in a human melanoma cell line J Immunol 1997 158 3353 3358 9120293 Jang HI Lee H A decrease in the expression of CD63 tetraspanin protein elevates invasive potential of human melanoma cells Exp Mol Med 2003 35 317 323 14508073 Azorsa DO Hildreth JEK CD63 cluster workshop report Leucocyte Typing V 1995 , Oxford University Press, Oxford, England 1352 Fukuda M Lysosomal membrane glycoproteins Journal of Biological Chemistry 1991 266 21327 21330 1939168 Imai T Fukudome K Takagi S Nagira M Furuse M Fukuhara N Nishimura M Hinuma Y Yoshie O CD33 antigen recognized by monoclonal antibodies inhibitory to human T cell leukemia virus type 1-induced syncytium formation is a member of a new family of transmembrane proteins including CD9, CD37, CD53, and CD63 Journal of Immunology 1992 149 2879 2886 Fidler IJ Rationale and methods for the use of nude mice to study the biology and therapy of human cancer metastasis Cancer Metastasis Rev 1986 5 29 49 2942306 10.1007/BF00049529 Skubitz KMCKDSAPN S. Schlossman Analysis of the Myeloid Panel of Monoclonal Antibodies by Immunoprecipitation, Immunoblotting, and Flow Cytometry Leukocyte Typing V 1995 Oxford, Oxford University Press 1071 1074 Bumol TF Reisfeld RA Unique glycoprotein-proteoglycan complex defined by monoclonal antibody on human melanoma cells Proc Natl Acad Sci U S A 1982 79 1245 1249 6175965 Iida J Meijne AM Spiro RC Roos E Furcht LT McCarthy JB Spreading and focal contact formation of human melanoma cells in response to the stimulation of both melanoma-associated proteoglycan (NG2) and alpha 4 beta 1 integrin Cancer Res 1995 55 2177 2185 7743521 Skubitz KM Snook RWII Monoclonal antibodies that recognize lacto-N-fucopenatose III (CD15) react with adhesion-promoting glycoprotein family (LFA-1/HMAC-1/GP 150,95) and CR1 on human neutrophils Journal of Immunology 1987 139 1631 1639 Laemmli UK Cleavage of structural proteins during the assembly of the head of bacteriophage T4 Nature 1970 227 680 685 5432063 10.1038/227680a0 Munoz G Marshall SH An alternative method for a fast separation of phosphotyrosine Anal Biochem 1990 190 233 237 1705393 10.1016/0003-2697(90)90185-C Skubitz KM Goueli SA CD31 (PECAM-1), CDw32 (Fc gamma RII), and anti-HLA class I monoclonal antibodies recognize phosphotyrosine-containing proteins on the surface of human neutrophils J Immunol 1994 152 5902 5911 7515918 Herbert JM Augereau JM Gleye J Maffrand JP Chelerythrine is a potent and specific inhibitor of protein kinase C Biochem Biophys Res Commun 1990 172 993 999 2244923 10.1016/0006-291X(90)91544-3 Skubitz KM Hammerschmidt DE Corticosteroids reversibly inhibit chemotactic peptide-receptor binding and granulocyte response, yet allow desensitization and receptor down- regulation Blood 1986 68 830 836 3756350 O'Flaherty JT Kreutzer DL Showell HS Becker EL Ward PA Desensitization of the neutrophil aggregation response to chemotactic factors Am J Pathol 1978 93 693 706 717543 Lagaudriere-Gesbert C Le Naour F Lebel-Binay S Billard M Lemichez E Boquet P Boucheix C Conjeaud H Rubinstein E Functional analysis of four tetraspans, CD9, CD53, CD81, and CD82, suggests a common role in costimulation, cell adhesion, and migration: only CD9 upregulates HB-EGF activity Cell Immunol 1997 182 105 112 9514697 10.1006/cimm.1997.1223 Lazo PA Cuevas L Gutierrez del Arroyo A Orue E Ligation of CD53/OX44, a tetraspan antigen, induces homotypic adhesion mediated by specific cell-cell interactions Cell Immunol 1997 178 132 140 9225004 10.1006/cimm.1997.1139 Masellis-Smith A Jensen GS Seehafer JG Slupsky JR Shaw AR Anti-CD9 monoclonal antibodies induce homotypic adhesion of pre-B cell lines by a novel mechanism J Immunol 1990 144 1607 1613 2307836 Eisenmann KM McCarthy JB Simpson MA Keely PJ Guan JL Tachibana K Lim L Manser E Furcht LT Iida J Melanoma chondroitin sulphate proteoglycan regulates cell spreading through Cdc42, Ack-1 and p130cas Nat Cell Biol 1999 1 507 513 10587647 10.1038/70302 Israels SJ McMillan-Ward EM CD63 modulates spreading and tyrosine phosphorylation of platelets on immobilized fibrinogen Thromb Haemost 2005 93 311 318 15711748 Berditchevski F Bazzoni G Hemler ME Specific association of CD63 with the VLA-3 and VLA-6 integrins Journal of Biological Chemistry 1995 270 17784 17790 7629079 10.1074/jbc.270.30.17784 Berditchevski F Tolias KF Wong K Carpenter CL Hemler ME A novel link between integrins, transmembrane-4 superfamily proteins (CD63 and CD81), and phosphatidylinositol 4-kinase J Biol Chem 1997 272 2595 2598 9006891 10.1074/jbc.272.46.29174 Berditchevski F Odintsova E Characterization of integrin-tetraspanin adhesion complexes: role of tetraspanins in integrin signaling J Cell Biol 1999 146 477 492 10427099 10.1083/jcb.146.2.477 Radford KJ Thorne RF Hersey P CD63 associates with transmembrane 4 superfamily members, CD9 and CD81, and with b1 integrins in human melanoma Biochemical and Biophysical Research Communications 1996 222 13 18 8630057 10.1006/bbrc.1996.0690	16318634	PMC1325047	CC BY	2021-01-04 16:24:26	no	J Transl Med. 2005 Nov 30; 3:42	utf-8	J Transl Med	2,005	10.1186/1479-5876-3-42	oa_comm
==== Front BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1721632976210.1186/1471-2164-6-172Research ArticleThe angiotensin-converting enzyme (ACE) gene family of Anopheles gambiae Burnham Susan [email protected] Judith A [email protected] Alison J [email protected] R Elwyn [email protected] Alan D [email protected] Faculty of Biological Sciences, Miall Building, University of Leeds, Clarendon Way, Leeds LS2 9JT, UK2 Department of Biological Sciences, Lancaster Environment Centre, Lancaster University, Lancaster, LA1 4YQ, UK2005 5 12 2005 6 172 172 22 9 2005 5 12 2005 Copyright © 2005 Burnham et al; licensee BioMed Central Ltd.2005Burnham 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 Members of the M2 family of peptidases, related to mammalian angiotensin converting enzyme (ACE), play important roles in regulating a number of physiological processes. As more invertebrate genomes are sequenced, there is increasing evidence of a variety of M2 peptidase genes, even within a single species. The function of these ACE-like proteins is largely unknown. Sequencing of the A. gambiae genome has revealed a number of ACE-like genes but probable errors in the Ensembl annotation have left the number of ACE-like genes, and their structure, unclear. Results TBLASTN and sequence analysis of cDNAs revealed that the A. gambiae genome contains nine genes (AnoACE genes) which code for proteins with similarity to mammalian ACE. Eight of these genes code for putative single domain enzymes similar to other insect ACEs described so far. AnoACE9, however, has several features in common with mammalian somatic ACE such as a two domain structure and a hydrophobic C terminus. Four of the AnoACE genes (2, 3, 7 and 9) were shown to be expressed at a variety of developmental stages. Expression of AnoACE3, AnoACE7 and AnoACE9 is induced by a blood meal, with AnoACE7 showing the largest (approximately 10-fold) induction. Conclusion Genes coding for two-domain ACEs have arisen several times during the course of evolution suggesting a common selective advantage to having an ACE with two active-sites in tandem in a single protein. AnoACE7 belongs to a sub-group of insect ACEs which are likely to be membrane-bound and which have an unusual, conserved gene structure. ==== Body Background In mammals, angiotensin-converting enzyme (EC 3.4.15.1, ACE, peptidyl-dipeptidase A), an important member of the M2 peptidase family, is found on the surface of endothelial cells and is best known for its role in the biosynthesis of angiotensin II, as well as the degradation of circulating bradykinin and the haemoregulatory peptide, N-acetyl SDKP [1-3]. Mammalian ACE exists as two isoforms, a somatic form (sACE, 150–180 kDa) and a smaller protein (germinal ACE, 90–110 kDa) found exclusively in adult testes. The ACE gene comprises twenty-six exons and two promoters, and has clearly arisen by gene duplication [4]. The somatic promoter drives expression of the larger protein (exons 1–12 and exons 14–26), which consists of two very similar domains connected in tandem by a short inter-domain peptide. Each domain possesses a functional peptidase active site and they are commonly called the N- and C-domains after their relative position to the amino and carboxy termini. The bulk of the protein including the amino terminus is extracellular and is linked by a hydrophobic transmembrane sequence to a short intracellular peptide. The biological rationale for two catalytic units linked in tandem is not known, although there is experimental evidence to suggest that co-operativity exists between the two halves of the protein [5-7]. In developing spermatids, an alternative intragenic promoter utilises exons 13–26 and generates a single domain ACE (gACE) which is largely identical to the C-terminal domain of somatic ACE [4,8]. Male mice lacking gACE are infertile as a result of sperm that are defective in migration in the oviduct and in binding to the zona pellucida [9]. Insect ACE has broad substrate specificity and its presence in the haemolymph of insects raises the possibility that, like mammalian sACE, it is required for extracellular metabolism of peptide hormones [10,11]. Drosophila melanogaster has six ACE-like genes (Ance, Acer, Ance-2, Ance-3, Ance-4 and Ance-5), all coding for single domain proteins. The ANCE protein has been relatively well characterised and has been shown to have several biochemical similarities to the mammalian enzyme. Its activity is developmentally regulated, with highest levels during metamorphosis [12,13]. ANCE is also expressed in spermatocytes and appears to have an important role in spermiogenesis [14]. Of the other Drosophila ACEs, only the Acer gene product has been studied biochemically and it is likely that ANCE-2, ANCE-3, ANCE-4 and ANCE-5 do not function as peptidases since they lack one or more of the residues that are essential for peptidase activity [15]. ACER, like ANCE and human ACE, is a peptidyl dipeptidase, but is generally less efficient than ANCE at cleaving dipeptides from many oligopeptide substrates [16]. Acer is expressed in the embryonic heart [17] and in both the male and female gonads and brain of adult flies (A. Carhan, R. E. Isaac and A. D. Shirras, unpublished results) where it is assumed to have a role in the metabolism of, as yet unidentified, biologically active peptides involved in neuroendocrine signalling and reproduction. ACE activity has been found in the gonads and accessory glands of several insect species in addition to D. melanogaster, suggesting a conserved role for this enzyme in insect reproduction [10,11]. Recently, we have provided evidence for a role for ACE in reproduction in the mosquito, A. stephensi. ACE activity in the adult female mosquito increases by 260% following a blood meal [10,18] and it has been proposed that the induced ACE has a role in regulating peptide signalling in response to a blood meal. We have shown that feeding two different ACE inhibitors to male A. stephensi in their glucose diet, resulted in approx. 80% reduction in the number of eggs laid [10]. In addition, ACE inhibitors introduced in the blood meal, resulted in a dose-dependent effect on brood-size, but had no effect on oocyte development, nor the rate of digestion of the blood. The only observable difference between inhibitor-fed and control insects was that the inhibitor-fed females could not lay eggs, even 8 days after the blood meal, suggesting that the blood-induced ACE is involved in the control of egg-laying [19]. To understand the mechanism of ACE induction in blood-fed mosquitoes and the role of the enzyme in mosquito reproductive physiology, we have characterised the ACE gene family in Anopheles gambiae, a mosquito responsible for the transmission of the human malaria parasite and whose genome has recently been sequenced [20]. The partially annotated A gambiae genome [21] reveals between eight and twelve potential ACE-like genes. We now report the correct number of A. gambiae ACE (AnoACE) genes, confirm their corresponding protein sequences and establish their expression patterns during development and in response to a blood meal. We show that several of the AnoACEs are similar to other insect ACEs, but that AnoACE9 is a novel two-domain ACE whose structure resembles that of human somatic ACE. Results ACE-like genes in the A. gambiae genome The Ensembl annotation of the A. gambiae genome (version 34.2 g) [21] suggests that there are eight genes encoding proteins belonging to the M2 (ACE) family of peptidases. BLAST searching of the genome and an examination of the predicted genes, however, shows that not all ACE-like proteins have been predicted, while several of those that have been identified have missing exons, or are hybrids, composed of exons from more than one ACE-like gene. These problems are particularly acute for a cluster of ACE-like sequences at 35C on chromosome 3R (Fig. 1). In order to confirm the number and organisation of ACE-like genes in the A. gambiae genome a combination of TBLASTN analysis [21], using Drosophila melanogaster ACE proteins as query sequences, and cDNA sequence analysis was carried out. Results are summarised in Tables 2 and 3 and may be viewed as a custom track in Ensembl Mosquito [37]. Figure 1 Annotation of ACE-like genes in the 35C cluster. Ensembl ContigView display of the 43.05–43.07 Mb region of chromosome 3R showing AnoACE custom track annotation (red bars), Ensembl EST transcripts (ENSANGESTT) and Ensembl predicted transcripts (ENSANGT). From left to right the AnoACE genes are numbered 2, 3 (with two alternative 5' exons), 4, 5 and 6. Table 2 ACE-like genes in the A. gambiae genome. Genes were identified by a combination of TBLASTN using Drosophila genes as query sequences and cDNA sequence analysis. Corresponding Ensembl (version 34.2 g) transcripts and ESTs (where known) are indicated. AnoACE ENSANGT ENSANGESTT Chromosome Contig Coordinates 1 0000010993 - 2L, 28C AAAB01008807_41 47,788,338–47,790,069 2 - 00000359639 00000359801 3R, 35C AAAB01008980_249 43,046,775 – 43,049,518 3 00000014438 (part) 0000035981* 00000359997* 00000360042 3R, 35C AAAB01008980_249 43,050,199 – 43,055,041 4 00000014440 (part) - 3R, 35C AAAB01008980_249 43,055,473 – 43,057,570 5 00000014440 (part) - 3R, 35C AAAB01008980_249 43,059,057 – 43,061,130 6 00000014440 (part) - 3R, 35C AAAB01008980_249 43,063,988 – 43,066,391 7 00000022325 00000014398† (part) 00000360081 3R, 35C AAAB01008980_250 AAAB01008980_252† 43,068,658 – 43,072,250 43,136,048 – 43,139,362† 8 00000007115 - 3R, 29C AAAB01008964_349 3,659,796 – 3,661,800 9 00000028617 00000011721 00000365570 00000365630 2R, 19C AAAB01008898_151 59,413,237 – 59,418,650 alternative transcripts with different 5' exons † 3' exons of AnoACE7, 69 kb downstream of the rest of the gene Table 3 Exon coordinates of AnoACE genes. Coordinates are as in Ensembl Mosquito version 34.2 g. Exon annotations may be viewed as a custom track in Ensembl Mosquito [37]. AnoACE, Chr. Exon Start End Length Start codon Stop codon 1, 2L 1 47788338 47790069 1731 47788338 47790069 2, 3R 1 43046775 43046948 173 43047058 43049381 2 43047050 43047327 277 3 43047407 43047771 364 4 43047832 43048053 221 5 43048118 43048261 143 6 43048324 43049032 708 7 43049101 43049274 173 8 43049343 43049518 175 3A, 3R 1 43050199 43050555 356 43050288 43054998 2 43053178 43053542 364 3 43053612 43053976 364 4 43054035 43054916 881 5 43054981 43055041 60 3B, 3R 1 43051447 43051757 310 43051523 43054998 2 43053178 43053542 364 3 43053612 43053976 364 4 43054035 43054916 881 5 43054981 43055041 60 4, 3R 1 43055473 43055722 249 43055473 43057570 2 43055795 43056159 364 3 43056227 43056590 363 4 43056672 43057570 898 5, 3R 1 43059057 43059300 243 43059057 43061130 2 43059368 43059732 364 3 43059801 43060165 364 4 43060238 43061130 892 6, 3R 1 43063988 43064231 243 43063988 43066391 2 43064299 43064663 364 3 43064732 43065096 364 4 43065169 43066391 1222 7, 3R 1 43068658 43069121 463 43068773 43139362 2 43069190 43069805 615 3 43070472 43070631 159 4 43070704 43070827 123 5 43070892 43071025 133 6 43071102 43071687 585 7 43071770 43072014 244 8 43072151 43072250 99 9 43136048 43136189 141 10 43139040 43139362 322 8, 3R 1 3659796 3661191 1395 3659796 3661800 2 3661258 3661542 284 3 3661607 3661800 193 9, 2R 1 59413237 59413683 446 59413677 59418197 2 59414399 59414615 216 3 59414676 59416593 1917 4 59416658 59418650 1992 Start of open reading frame, no start methionine identified TBLASTN analysis revealed 11 possible ACE-like coding sequences. Six of these sequences are arranged in a tightly linked tandem cluster at 35C (Figs. 1 &2). These were numbered (from 5' to 3') AnoACE 2,3,4,5,6 and 7. Several EST clones exist for AnoACE2 and AnoACE3, allowing the exons of these genes, and their coding regions to be unambiguously assigned. AnoACE3 has two alternative transcripts which differ in their first exon. No cDNA clones are available for AnoACEs 4, 5 and 6. The exons of these genes were identified by BLAST similarity and the SNAP exon predictions of Ensembl [21]. The nucleotide sequences of AnoACEs 5 and 6 are 97% identical, including the introns, which raised the possibility that these two genes are the result of a sequence assembly error caused by nucleotide polymorphisms in the PEST strain. This high level of identity is only observed within the putative transcribed region, however. The upstream flanking sequence of AnoACE6 shares only around 75% identity with the AnoACE5 upstream DNA and there is no significant similarity between the downstream flanking sequences of the two genes. An assembly error can therefore be ruled out. The AnoACE7 gene has an unusual organisation; there are eight upstream exons situated at the 3' end of the 35C cluster which encode the bulk of the protein and there are two exons encoding the C terminus of the protein 69 kb downstream (Fig. 2). These exons were shown to belong to the AnoACE7 gene by the sequence of NAP1 cDNA P141-G-08-5 which spans exon 8 and the two downstream exons (Fig. 2., Ensembl annotation). The gene structure was further confirmed by TBLASTN alignment with the Drosophila ANCE-3 protein. The 5' end of AnoACE7 was determined by RT-PCR analysis using nested primers designed against the genomic sequence. A 1.1 kb RT-PCR product was cloned, sequenced and found to contain a long open reading frame coding for a protein which overlapped existing predicted AnoACE7 protein sequence and which had a putative signal peptide. This was assumed to correspond to the N terminus of the AnoACE7 precursor. Figure 2 Annotation of AnoACE7. Ensembl ContigView display of the 43.07–43.13 Mb region of chromosome 3R showing AnoACE custom track annotation (red bar), Ensembl ESTs (purple bars), EST transcripts (ENSANGESTT) and Ensembl predicted transcripts (ENSANGT). Eight upstream exons, encoding the bulk of the protein, are situated at the 3' end of the 35C cluster, downstream of AnoACE6 (Fig. 1). Two exons encoding the C terminus of the protein are situated 69 kb downstream. The sequence of EST clone P141-G-08-5 spans upstream and downstream exons of AnoACE7 and is shown below the gene annotation. The large intron contains several potential genes. Not all introns are shown at this scale. The AnoACE9 gene contains two ACE-like coding regions. The sequence of cDNA 19600449703686 (MRA-468-36) spans both coding regions (Ensembl annotation), indicating that this is a single gene coding for a double-domain protein, rather than two separate genes (Fig. 3). Figure 3 Annotation of AnoACE9. Ensembl ContigView display of the 59.413–59.419 Mb region of chromosome 2R showing AnoACE custom track annotation (red bar), Ensembl EST transcripts (ENSANGESTT) and Ensembl predicted transcripts (ENSANGT). The AnoACE9 gene consists of 4 exons. Exons 3 and 4 both contain ACE active site coding regions and are spanned by cDNA 19600449703686 (ENSANGESTT00000365630) indicating a double-domain enzyme. AnoACE proteins Protein sequences were predicted from cDNAs for AnoACEs 2, 3, 7 and 9. AnoACEs 1, 4, 5, 6 and 8 were predicted from genomic sequence and alignment with Drosophila ACE proteins. Predicted properties of the proteins are summarised in Table 4. Signal peptides were identified for all proteins except AnoACEs 1 and 8. All of the AnoACE genes, except for AnoACE9, code for single domain proteins similar to those previously found in insects. AnoACE9 is a putative double domain enzyme, similar to mammalian somatic ACE. Two of the AnoAces (7 and 9) have hydrophobic C termini which might serve as membrane anchors or be lost as a result of post-translational modification with a GPI anchor [22,23]. AnoACE7 has an extended N terminus, compared to the other proteins. All proteins have a full complement of required active site amino acids in the HExxH...EAV/I/L motif (Fig. 4). Predicted protein sequences are available as additional data file 1. Table 4 Predicted AnoACE proteins. Conceptual proteins were translated from cDNA sequences for AnoACEs 2, 3, 7 and 9 and from genomic sequence for AnoACEs 1, 4, 5, 6, and 8. AnoACE Length (aa) Active sites Hydrophobic C terminus Predicted GPI anchor 1‡ 576 1 No No 2 638 1 No No 3A 630 1 No No 3B* 619 1 No No 4 625 1 No No 5† 619 1 No No 6† 730 1 No No 7 917 1 Yes No/marginal 8‡ 623 1 No No 9 1225 2 Yes Yes *AnoACE3A and AnoACE3B are derived from alternative transcripts and differ at their N termini †AnoACE5 and AnoACE6 are 97% identical, except for an extended C terminus in AnoACE6 ‡N-terminus probably missing, no signal peptide Figure 4 Active site region of ACE-like proteins. The active site regions of AnoACEs, the N and C domains of human ACE, D. melanogaster ANCE, ACER and ANCE-3, and Apis 3929 were aligned using CLUSTALX [36]. Amino acids shown to be required for activity in mammalian ACE are highlighted. The histidines in the HExxH motif are required for zinc coordination, whereas the glutamate is required for catalysis. The downstream glutamate is required for zinc coordination. This is replaced by glutamine in ANCE-3. Comparative genomics of ACE-like genes A phylogram showing the relationships of the AnoACEs to other insect and chordate ACEs is shown in Fig. 5. AnoACEs 2, 3, 4, 5 and 6 cluster with Drosophila ANCE and ACER, and ACE-like proteins from Apis mellifera, Haematobia irritans, Lutzomyia longipalpis and Bombyx mori. These proteins are all single domain proteins which lack a hydrophobic C terminus. A second cluster of insect proteins contains AnoACEs 7 and 8, the two remaining Apis ACEs and Drosophila ANCE-3. AnoACE7 is a one-to-one orthologue of ANCE-3 and of Apis 3929. AnoACE8 is a one-to-one orthologue of Apis 10228. There appears to be no Drosophila orthologue of AnoACE8. The similarity of the AnoACE7 group extends to the genomic organisation of the genes. In each case the bulk of the protein is encoded by a set of upstream exons and these are separated from the downstream exons, encoding the C terminus of the protein by a large intron. In the case of Anopheles the intron is 69 kb in length, in Drosophila 35 kb and in Apis 21 kb. The proteins in this group are characterised by an extended N terminus, rich in proline and charged amino acids. AnoACE7 and Drosophila ANCE-3 have a hydrophobic C terminus but this is absent from the predicted Apis protein. GPI anchor prediction software [23] predicts no GPI anchor for AnoACE7 whereas Drosophila ANCE-3 has a strong GPI anchor prediction. Figure 5 Evolutionary relationships of invertebrate and chordate ACE-like proteins. Core sequences (corresponding to amino acids161 to 569 of the AnoACE2 pre-protein) of each protein were aligned and a neighbour-joining tree calculated using CLUSTALX [36]. TreeView was used to produce a radial tree. Sequences used were AnoACEs (Ag)1, 2, 3, 4, 5, 6, 7, 8, and 9 (N and C domains); D. melanogaster ANCE, ACER, ANCE-2, -3, -4 and -5; B. mori ACE (accession no. BAA97657); H. irritans ACE (accession no. Q10715); C. elegans ACN-1; N and C domains of Human ACE, Fugu rubripes (Ensembl SINFRUP00000174092) ACE; A. mellifera Ensembl proteins ENSAPMP00000005043, ENSAPMP00000003929 and ENSAPMP000000010228; Sandfly (Lutzomyia longipalpis) ACE (accession no. AAS16911); Human and F. rubripes (Ensembl SINFRUP00000161972) ACE2. N and C domains of Ciona intestinalis ACE, were assembled from genomic sequence in contig AABS01000302.1. Scale bar shows number of amino acid substitutions per site. AnoACEs 1 and 9 are placed in a cluster with a diverse group of ACE-like proteins that includes Drosophila ANCEs-2, -4 and -5, the C. elegans ACE-like protein ACN-1 and vertebrate ACE2. The latter protein, while having sequence similarity to ACEs, has a different enzymatic activity and acts as a carboxypeptidase rather than a dipeptidase. Several members of this group (Drosophila ANCEs-2, -4 and -5, ACN-1) lack one or more essential active site amino acids and are therefore unlikely to be active peptidases. AnoACE9 has several structural similarities with chordate ACEs. It has two active site domains and a hydrophobic C terminus with a strong GPI anchor prediction (Fig. 6). The N and C domains of this protein are more similar to each other than to other ACEs, unlike the vertebrate N and C domains which cluster together. Figure 6 Diagrammatic alignment of human somatic ACE (1306 amino acids) and AnoACE9 (1225 amino acids). The hatched boxes represent the N and C domain active site regions. Filled boxes are the C-terminal hydrophobic potential membrane anchor regions. AnoACE9 has a very short potential cytoplasmic region and a sequence gap relative to the human enzyme on the N terminal side of the putative transmembrane region. Arrows indicate potential secretase recognition sequences. Expression of AnoACE genes Primers for all AnoACE genes were validated by PCR amplification of genomic DNA or cDNA. Each primer pair produced a product of the expected size (Fig. 7). Figure 7 RT-PCR analysis of AnoACE gene expression. RNA was extracted from glucose fed adult females, blood fed adult females, eggs, larvae, pupae and adult males. Control templates for amplification were genomic DNA except for AnoACE2 primers when the template was DNA from cDNA clone 19600449622752 (MRA-467-39). Amplification products were observed for AnoACEs 2, 3, 7 and 9 only. For semi-quantitative RT-PCR analysis RNA was extracted and cDNA produced from different developmental stages of mosquitoes, namely, eggs, larvae, pupae, adult males, glucose-fed females and blood-fed females. Figure 7 shows that AnoACEs 2, 3, 7 and 9 are expressed at all developmental stages tested, whereas no PCR product was observed for AnoACEs 1, 4, 5, 6 and 8. The authenticity of the products from AnoACE2, 3, 7 and 9 was confirmed by cloning into the pGEM T-Easy® vector and sequencing. Gene expression for all the AnoACE genes was analysed in glucose and blood-fed females using quantitative real-time RT-PCR. Other than AnoACEs 2,3,7 and 9, expression levels were as for no template controls. Transcription of AnoACE3, AnoACE7 and AnoACE9 was shown to be up-regulated 48 hours following a blood meal (Fig. 8). The increase in transcription was consistent for three separate cohorts of female mosquitoes. Figure 8 Relative expression of AnoACE genes in glucose and blood-fed females. RNA was extracted from three different cohorts of mosquitoes, either after continuous glucose feeding or 48 hours after a blood meal, and analysed by quantitative RT-PCR. The mean increase in gene expression was 11.6 fold for AnoACE3, 16 fold for AnoACE7 and 2.4 fold for AnoACE9. Discussion Our analysis of the A. gambiae genome has revealed nine ACE-like genes. This is the largest number of ACE genes for any species with a complete genome sequence. Drosophila melanogaster has six ACE-like genes [15], but only two of these (Ance and Acer) are known to code for active peptidase proteins. As well as having the ACE-like genes found in D. melanogaster, D. pseudoobscura has an additional Ance-like gene which is likely to code for an active peptidase, since all the amino acids that are necessary for catalysis and coordination with the active site zinc are conserved (A.D. Shirras, unpublished work). Apis mellifera has three ACE-like genes, each with a full complement of the critical active site amino acids [24]. At present we have evidence that at least four AnoACE genes (2, 3, 7 and 9) are expressed, but have no evidence for expression of AnoACEs 1, 4, 5, 6 and 8. All of the AnoACE genes, however, have complete open reading frames and the corresponding proteins have a full complement of essential active site amino acids, so we cannot rule out the expression of the other genes at low levels, in a small number of cells or at stages that we have not examined. Some of the AnoACE genes have clearly arisen by recent gene duplications. The cluster at 35C contains five genes that are closely related in sequence. Two of these (AnoACE5 and AnoACE6) are almost identical. Similar clusters of other duplicated genes exist within the A. gambiae genome; for example there is a cluster of M14 carboxypeptidase genes at 29D which contains adjacent genes with greater than 90% identity. In both these clusters the high levels of identity are restricted to the transcribed regions with flanking DNA showing lower levels of similarity. This suggests that selective pressures are maintaining the nucleotide sequences of these genes and that they are expressed at some stage. Comparison of ACE protein sequences (Fig. 5) suggests that there are three functional groupings in insects. The first contains Drosophila ANCE, AnoACEs 2, 3, 4, 5 and 6 and several other insect ACEs. These are single domain, secreted enzymes. This class of ACE proteins is similar to the chordate N and C domains, both in terms of sequence and enzyme activity. The second group includes AnoACE7, Drosophila ANCE-3 and Apis 3929. These proteins are single domain, but have extended N termini. The Drosophila and mosquito proteins have a hydrophobic C terminus suggesting that they may be membrane-tethered. The active sites of these three proteins are characterised by an alanine in place of glycine at position 4 of the HExxH motif (Fig. 4). AnoACE8 and Apis 10228 form diverged members of this group. The third group of insect ACEs, containing AnoACE1 and Drosophila ANCE-2, -4 and -5, is highly divergent. These appear to be secreted proteins and, whereas AnoACE1 may be enzymically active, the Drosophila examples lack essential active site amino acids. C. elegans ACN-1 and vertebrate ACE2 proteins form highly diverged members of this family. This result is to be expected as they have distinctive and highly evolved roles in these animals [25,11]. AnoACE9 is highly unusual. Phylogenetic analyses using several different tree construction methods (Neighbour-Joining, Maximum Parsimony, Minimum Evolution, Unweighted Pair Group Method with Arithmetic Mean) do not place it convincingly in any of the insect groups described above. In terms of overall structure it is more similar to mammalian somatic ACE than to any invertebrate ACE described so far (Fig. 6). It has two active site domains and has a hydrophobic C terminus. This similarity with vertebrate ACEs is, as far as we know, unique amongst non-chordate invertebrates. It is likely that AnoACE9 is membrane bound, either by virtue of its hydrophobic C-terminus or via a potential GPI anchor. In mammals, membrane sACE can be released from the plasma membrane by unidentified metalloproteases, known as secretases or sheddases, that cleave the Arg-Ser bond in the Ala-Arg-Ser-Glu motif of the juxtamembrane stalk [26,27]. Interestingly, a similar peptide sequence (Leu-Arg-Ser-Asp) is found in the same position in the proposed juxtamembrane stalk of AnoACE9 (Fig. 6), which might serve as a cleavage site for a mosquito ACE secretase. In mammals, the two active sites of sACE are involved in the conversion of angiotensin I to angiotensin II and the hydrolysis of bradykinin, but only the N-domain is involved in N-acetyl SDKP metabolism [2]. Thus, duplication appears to have provided an opportunity for the N-domain to acquire distinctive substrate specificity. However, this does not explain why mammalian sACE is a single protein with two catalytic domains joined by a peptide linker, rather than single genes encoding single domain proteins as occurs in D. melanogaster. Our phylogenetic analysis indicates that the gene duplication event leading to the two-domain vertebrate somatic ACE, occurred around 450 million years ago, before the divergence of amphibians and fishes, and that two other very similar events occurred independently to give rise to a two-domain enzyme: one in the lineage leading to A. gambiae and one in the urochordate lineage leading to Ciona intestinalis (Fig. 5). That is, two-domain ACE proteins have been selected for on three separate occasions, suggesting that there is a distinct functional advantage to having two active sites organised in this manner over two single-domain proteins. Support for this hypothesis comes from several studies on mammalian sACE that suggest that the two active sites can have cooperative effects on ACE activity [5-7]. Recent studies indicate that the N-domain of human somatic ACE also has a negative effect on the shedding of the protein from the cell surface by cleavage within the stalk region that separates the C-domain from the membrane anchor [28]. More detailed biochemical studies on AnoACE9 are now required to determine whether similar cooperative phenomena occur in mosquitoes. The expression of AnoACEs 3, 7 and 9 is induced by a blood meal. Two of these genes (AnoACE 3 and 7) are induced more than ten-fold. Neuropeptides are known to be important for mediating several behavioural and physiological responses to a blood meal in mosquitoes [29-34]. Our previous work showing that ACE inhibitors block egg-laying, but not egg development, in blood fed A. stephensi [19] led us to propose that the elevated peptidase activity might be important for regulating a peptide signal involved in egg-laying. The up-regulation of AnoACEs 3, 7 and 9 following a blood meal in A. gambiae, raises the possibility that ACE may also be involved in regulating post feeding oviposition in this species. A role for at least some AnoACEs in the immune response is suggested by recent microarray data [35]. AnoACE7 (ENSANGT00000022325) was shown to be 2.4-fold upregulated by challenge with Salmonella typhimurium. AnoACE1 (ENSANGT0000010993) was 3.5-fold upregulated by Staphylococcus aureus infection, whereas the expression of AnoACE9 (ENSANGT00000028617) was reduced 2.4-fold by Beauveria bassiana. We were unable to detect AnoACE1 expression in a variety of developmental stages. However, the expression of this gene suggested by the infection microarray may mean that this gene plays a specific role as a component of the immune response. Conclusion The Anopheles gambiae genome contains 9 genes coding for proteins with similarity to mammalian ACE. Six of these genes are found in a tandem cluster situated at 35C on chromosome 3R. Five of the members of this cluster code for proteins which are similar to previously described insect ACEs: they contain a single active site domain and lack a C-terminal hydrophobic region. AnoACE7, situated at the 3' end of the 35C cluster, is an orthologue of Drosophila Ance-3 and of Apis gene ENSAPMG00000014390. In all three species, there is a similar gene structure with the bulk of the coding exons separated from the C-terminal exons by a large intron which contains other genes. The AnoACE7 protein, like Drosophila ANCE-3, has a hydrophobic C-terminus and is likely to be membrane bound. AnoACEs 8 and 10 are also predicted to be single domain proteins but AnoACE9 is unique amongst insect ACEs described so far in that it contains two active site domains and has a hydrophobic C-terminus, a structure similar to that of mammalian somatic ACE. Expression of only 4 of the Anopheles ACE genes (2, 3, 7 and 9) could be detected by RT-PCR in the stages tested. Expression of AnoACEs 3, 7 and 9 increased following a blood meal. Methods Insects Male and female Anopheles gambiae were maintained at a temperature of 26°C and at a relative humidity of 80 %. Newly hatched adult mosquitoes were fed for seven days on sterile 5% (w/v) glucose/0.5% (w/v) p-aminobenzoic acid solution [20]. On the seventh day female mosquitoes were harvested and divided into two cohorts. Mosquitoes in cohort A were maintained on the glucose/p-aminobenzoic acid solution, whilst mosquitoes in cohort B were fed mouse blood. After the blood meal any unfed female mosquitoes (from cohort B) were discarded. Mosquitoes from cohort B were collected 48 hours after the blood meal. Insects from cohort A were collected 48 hours after cohort B was given the blood meal. Extraction of nucleic acids Nucleic acids were extracted using either the Qiagen DNeasy® or RNeasy® kits as per manufacturer's directions. During RNA extraction the optional RNase-free DNase I from Qiagen® was used to digest DNA. After quantification, extracted DNA was stored at 4°C and RNA at -80°C until required. PCR Primers were designed for each possible mosquito ACE gene, with each pair crossing exon boundaries, where this was possible (see Table 1 for primer sequences). Table 1 Primers for PCR of AnoACE genes Gene Forward Primer sequence (5'-3') Reverse Primer Sequence (5'-3') Predicted Product Size From Genomic DNA(bp) AnoACE1 GGAATGCATCGTGTGTGTTC TACTCGAAGGGAATGGTTGG 943 AnoACE2 GGTACCGATCAACCAGTGCT AAATCGCATCGACAAGCTCT 440 AnoACE3 CCCTGTGCAAGTGCTGATAA ATACGGAACGCCATTGGATA 214 AnoACE4 CAGCCGAGATGCTGAAGAGT GAACTCGCGCAACCTATACC 328 AnoACE5 TTGCTGCAGCGAATGCTAT CACACCAGCCCAATCTTGT 754 AnoACE6 CAGCGAATGCTGTGTTCCT CAATGTTGCCCCACTTCTGT 748 AnoACE7 CGCAGTGGAACTTTGAGACA GTCTATTCGGCGCATTTGAT 782 AnoACE8 GGTCTGGGCGATACCAAGTA CGCGTCAAATCGGAGTTAAT 220 AnoACE9 GCACACGTATCGGAAAATGA GCGCTTCTCGATGAACTACC 803 The primers were also designed so that the product size from cDNA would be approximately 100 bp as appropriate for use with QPCR, except for AnoACE1 where the product was 943 bp. Amplification reactions were prepared in 20 μl volumes and contained; 100 ng of template, 2 μl of PCR buffer (670 mM Tris-HCl, pH 8.8; 166 mM [NH4]2SO4; 67 mM MgCl2), 200 μM of dNTPs, 330 ng of BSA, 0.14 μl of 5% (v/v) β-mercaptoethanol, 1 unit of Promega® Taq polymerase, 20 pmol of forward primer, 20 pmol of reverse primer. The following cycling conditions were used: one incubation at 94°C for 5 min; 35 cycles of 94°C for 30 s; Tm for 30 s; 74°C for 1 min: followed by one incubation at 74°C for 5 min (where the Tm is appropriate for each set of primers). RT-PCR To amplify the 5' end of AnoACE7, cDNA was prepared using an Invitrogen Cloned AMV kit as per the manufacturer's instructions. Using 2 μg total RNA from blood fed A. gambiae and 10 mM gene specific primer (ACE7 REV: 5'-GTCTATTCGGCGCATTTGAT-3'), first strand synthesis was performed at 50°C for 1 hour. PCR was then carried out using 1 μl of the first strand reaction, 2 mM dNTPs, 0.5 mM each primer (ACE7 REV and ACE7 FOR3: 5'-AGTGAAAACGTAACCATTGCAGAA-3') and 2.5 units Proofstart enzyme (Qiagen). PCR conditions were as follows: initial denaturation at 95°C, 5 min. then 35 cycles of 95°C for 30s, 54°C for 30s, 72°C for 3 min., and a final extension at 72°C for 5 min. A 1.1 kb product was purified using the Qiagen® gel purification kit as per the manufacturer's instructions, cloned into pGEM®T-Easy and sequenced using M13 forward and reverse, and internal primers. The accession number of this sequence is [EMBL:AM085517]. For semi-quantitative RT-PCR, first strand synthesis of cDNA took place in a 20 μl reaction volume using 1 μg of RNA and Promega ImPromII RT® kit as per the manufacturer's instructions and either random or gene-specific primers. PCR was carried out using the method described above with 2 μl of cDNA, primers as in Table 1, and the following cycling conditions: 94°C for 5 min. then 35 cycles of 94°C for 15 s, Tm for 15 s, 74°C for 30 s followed by incubation at 74°C for 5 min. PCR products were separated by agarose gel electrophoresis. To confirm identity of the PCR products, bands were purified using the Qiagen® gel purification kit, cloned into pGEM® T-Easy and sequenced using M13 forward and reverse primers. Quantitative RT-PCR RNA (3 μg) was reverse transcribed using the Promega ImPromII® kit and the resulting cDNA quantified using a Biorad iCycler instrument. cDNA (5 μl) was diluted in water to a total of 200 μl. Three-fold serial dilutions were made from this mixture as templates for the calibration curve. The remaining 15 μl of each experimental cDNA was made up to 400 μl with water. Reactions were performed in duplicate using 10 μl of template, 12.5 μl of ABsolute™ QPCR SYBR® green and 2.5 μl of primers at a concentration determined to be optimum for each primer pair. Reactions were subjected to the following amplification cycles; 95°C for 15 min, once; 95°C for 15 s; 60°C for 30 s for 40 cycles. Melt curves were also performed to confirm that a single PCR product and no primer dimers were produced. The melt curves were performed using the following conditions; one incubation at 95°C for 30 s, 55°C for 30 s, once; followed by incubation at 55°C for 10 s 82 times, with a temperature increase of 0.5°C every cycle after cycle 2. Transcript levels of the AnoACE genes were normalised against the levels of mRNA for the ribosomal protein RPS17, amplified using primers 5'-TTGACCATGGATTTCGACAC-3' and 5'-TGATGGAAATACCACGCACT-3', forward and reverse, respectively. List of abbreviations ACE: angiotensin I-converting enzyme, AnoACE: Anopheles ACE Authors' contributions SB and AJL identified AnoACE genes and carried out expression studies. SB produced the initial draft of the manuscript. JAS cloned and sequenced the 5' end of AnoACE7. REI helped draft the manuscript. ADS refined the AnoACE annotation, carried out sequence comparisons and phylogenetic analysis, and re-drafted the manuscript. REI and ADS jointly planned the study and supervised experiments. Supplementary Material Additional File 1 AnoACE protein sequences in FASTA format. Click here for file Acknowledgements This work was supported by the Wellcome Trust (Grant GR071921). We thank Debbie Evans and Martin Looker for providing mosquitoes. The Malaria Research and Reference Reagents Resource Center (MR4) supplied cDNA clones deposited by R. A. Holt. NAP1 cDNA clones were supplied by George Christophedes, European Molecular Biology Laboratory. We also acknowledge valuable discussion with Ed Parkin of the School of Biochemistry and Molecular Biology, University of Leeds, about ACE secretase. ==== Refs Erdos EG Angiotensin I-converting enzyme and the changes in our concepts through the years Hypertension 1990 16 363 370 2170273 Corvol P Williams TA Soubrier F Peptidyl dipeptidase A: Angiotensin I-converting enzyme Methods In Enzymology 1995 248 283 305 7674927 Corvol P Eyries M Soubrier F Barrett AJ, Rawlings ND, Woessner JF Peptidyl-dipeptidase A/angiotensin I-converting enzyme Handbook of Proteolytic Enzymes 2004 1 2 Amsterdam: Elsevier Academic Press 332 346 Hubert C Houot AM Corvol P Soubrier F Structure of the angiotensin I-converting enzyme gene – 2 alternate promoters correspond to evolutionary steps of a duplicated gene J Biol Chem 1991 266 15377 15383 1651327 Binevski PV Sizova EA Pozdnev VF Kost OA Evidence for the negative cooperativity of the two active sites within bovine somatic angiotensin-converting enzyme FEBS Lett 2003 550 84 88 12935891 10.1016/S0014-5793(03)00825-1 Georgiadis D Beau F Czarny B Cotton J Yiotakis A Dive V Roles of the two active sites of somatic angiotensin-converting enzyme in the cleavage of angiotensin I and bradykinin: insights from selective inhibitors Circ Res 2003 93 148 154 12805239 10.1161/01.RES.0000081593.33848.FC Andújar-Sánchez M Cámara-Artigas A Jara-Pérez V A calorimetric study of the binding of lisinopril, enalaprilat and captopril to angiotensin-converting enzyme Biophys Chem 2004 111 183 189 15381315 10.1016/j.bpc.2004.05.011 Langford KG Shai SY Howard TE Kovac MJ Overbeek PA Bernstein KE Transgenic mice demonstrate a testis-specific promoter for angiotensin-converting enzyme J Biol Chem 1991 266 15559 15562 1651914 Hagaman JR Moyer JS Bachman ES Sibony M Magyar PL Welch JE Smithies O Krege JH O'Brien DA Angiotensin-converting enzyme and male fertility Proc Natl Acad Sci USA 1998 95 2552 2557 9482924 10.1073/pnas.95.5.2552 Isaac RE Ekbote U Coates D Shirras AD Insect angiotensin-converting enzyme. A processing enzyme with broad substrate specificity and a role in reproduction Ann NY Acad Sci 1999 897 342 347 10676461 Macours N Hens K Zinc-metalloproteases in insects: ACE and ECE Insect Biochem Mol Biol 2004 34 501 510 15147752 10.1016/j.ibmb.2004.03.007 Houard X Williams TA Michaud A Dani P Isaac RE Shirras AD Coates D Corvol P The Drosophila melanogaster-related angiotensin-I-converting enzymes Acer and Ance: Distinct enzymic characteristics and alternative expression during pupal development Eur J Biochem 1998 257 599 606 9839949 10.1046/j.1432-1327.1998.2570599.x Siviter RJ Taylor CAM Cottam DM Denton A Dani MP Milner MJ Shirras AD Isaac RE Ance, a Drosophila angiotensin-converting enzyme homologue, is expressed in imaginal cellsduring metamorphosis and is regulated by the steroid, 20-hydroxyecdysone Biochem J 2002 367 187 193 12093364 10.1042/BJ20020567 Hurst D Rylett CM Isaac RE Shirras AD The Drosophila angiotensin-converting enzyme homologue Ance is required for spermiogenesis Dev Biol 2003 254 238 247 12591244 10.1016/S0012-1606(02)00082-9 Coates D Isaac RE Cotton J Siviter R Williams TA Shirras A Corvol P Dive V Functional conservation of the active sites of human and Drosophila angiotensin I-converting enzyme Biochemistry 2000 39 8963 8969 10913309 10.1021/bi000593q Siviter RJ Nachman RJ Dani MP Keen JN Shirras AD Isaac RE Peptidyl dipeptidases (Ance and Acer) of Drosophila melanogaster : major differences in the substrate specificity of two homologs of human angiotensin I-converting enzyme Peptides 2002 23 2025 2034 12431741 10.1016/S0196-9781(02)00190-0 Taylor CAM Coates D Shirras AD The Acer gene of Drosophila codes for an angiotensin-converting enzyme homologue Gene 1996 181 191 197 8973330 10.1016/S0378-1119(96)00503-3 Ekbote U Coates D Isaac RE A mosquito (Anopheles stephensi) angiotensin I-converting enzyme (ACE) is induced by a blood meal and accumulates in the developing ovary FEBS Lett 1999 455 219 222 10437776 10.1016/S0014-5793(99)00870-4 Ekbote U Looker M Isaac RE ACE inhibitors reduce fecundity in the mosquito, Anopheles stephensi Comp Biochem Physiol B Biochem Mol Biol 2003 134 593 598 12670786 10.1016/S1096-4959(03)00019-8 Holt RA Subramanian GM Halpern A Sutton GG Charlab R Nusskern DR Wincker P Clark AG Ribeiro JM Wides R Salzberg SL Loftus B Yandell M Majoros WH Rusch DB Lai Z Kraft CL Abril JF Anthouard V Arensburger P Atkinson PW Baden H de Berardinis V Baldwin D Benes V Biedler J Blass C Bolanos R Boscus D Barnstead M Cai S Center A Chaturverdi K Christophides GK Chrystal MA Clamp M Cravchik A Curwen V Dana A Delcher A Dew I Evans CA Flanigan M Grundschober-Freimoser A Friedli L Gu Z Guan P Guigo R Hillenmeyer ME Hladun SL Hogan JR Hong YS Hoover J Jaillon O Ke Z Kodira C Kokoza E Koutsos A Letunic I Levitsky A Liang Y Lin JJ Lobo NF Lopez JR Malek JA McIntosh TC Meister S Miller J Mobarry C Mongin E Murphy SD O'Brochta DA Pfannkoch C Qi R Regier MA Remington K Shao H Sharakhova MV Sitter CD Shetty J Smith TJ Strong R Sun J Thomasova D Ton LQ Topalis P Tu Z Unger MF Walenz B Wang A Wang J Wang M Wang X Woodford KJ Wortman JR Wu M Yao A Zdobnov EM Zhang H Zhao Q Zhao S Zhu SC Zhimulev I Coluzzi M della Torre A Roth CW Louis C Kalush F Mural RJ Myers EW Adams MD Smith HO Broder S Gardner MJ Fraser CM Birney E Bork P Brey PT Venter JC Weissenbach J Kafatos FC Collins FH Hoffman SL The genome sequence of the malaria mosquito Anopheles gambiae Science 2002 298 129 149 12364791 10.1126/science.1076181 EnsemblMosquito Eisenhaber B Bork P Eisenhaber F Prediction of potential GPI-modification sites in proprotein sequences J Mol Biol 1999 292 741 758 10497036 10.1006/jmbi.1999.3069 GPI Lipid Anchor Project Ensembl Honeybee 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 10.1074/jbc.M308858200 Ramchandran R Sen I Cleavage processing of angiotensin-converting enzyme by a membrane-associated metalloprotease Biochemistry 1995 34 12645 12652 7548015 10.1021/bi00039a021 Woodman ZL Oppong SY Cook S Hooper NM Schwager SL Brandt WF Ehlers MR Sturrock ED Shedding of somatic angiotensin-converting enzyme (ACE) is inefficient compared with testis ACE despite cleavage at identical stalk sites Biochem J 2000 347 711 718 10769174 10.1042/0264-6021:3470711 Woodman ZL Schwager SL Redelinghuys P Carmona AK Ehlers MR Sturrock ED The N domain of somatic angiotensin-converting enzyme negatively regulates ectodomain shedding and catalytic activity Biochem J 2005 389 739 744 15813703 10.1042/BJ20050187 Clements AN The biology of mosquitoes 1992 London: Chapman & Hall Stracker TH Thompson S Grossman GL Riehle MA Brown MR Characterization of the AeaHP gene and its expression in the mosquito Aedes aegypti (Diptera: Culicidae) J Med Entomol 2002 39 331 342 11931033 Brown MR Graf R Swiderek KM Fendley D Stracker TH Champagne DE Lea AO Identification of a steroidogenic neurohormone in female mosquitoes J Biol Chem 1998 273 3967 3971 9461584 10.1074/jbc.273.7.3967 Hayes TK Holman GM Pannabecker TL Wright MS Strey AA Nachman RJ Hoel DF Olson JK Beyenbach KW Culekinin depolarizing peptide: A mosquito leucokinin-like peptide that influences insect malpighian tubule ion transport Regul Pept 1994 52 235 248 7800855 10.1016/0167-0115(94)90058-2 Veenstra JA Isolation and identification of three leucokinins from the mosquito Aedes aegypti Biochem Biophys Res Commun 1994 202 715 719 8048942 10.1006/bbrc.1994.1989 Veenstra JA Costes L Isolation and identification of a peptide and its cDNA from the mosquito Aedes aegypti related to Manduca sexta allatotropin Peptides 1999 20 1145 1151 10573285 10.1016/S0196-9781(99)00117-5 Aguilar R Jedlicka AE Mintz M Mahairaki V Scott AL Dimopoulos G Global gene expression analysis of Anopheles gambiae responses to microbial challenge Insect Biochem Mol Biol 2005 35 709 719 15894188 10.1016/j.ibmb.2005.02.019 Thompson JD Gibson TJ Plewniak F Jeanmougin F Higgins DG The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools Nucleic Acids Res 1997 24 4876 4882 9396791 10.1093/nar/25.24.4876 Ensembl Mosquito AnoACE annotation	16329762	PMC1325048	CC BY	2021-01-04 16:32:47	no	BMC Genomics. 2005 Dec 5; 6:172	utf-8	BMC Genomics	2,005	10.1186/1471-2164-6-172	oa_comm
==== Front Health Qual Life OutcomesHealth and Quality of Life Outcomes1477-7525BioMed Central London 1477-7525-3-721628865010.1186/1477-7525-3-72ReviewQuality of life in bipolar disorder: A review of the literature Michalak Erin E [email protected] Lakshmi N [email protected] Raymond W [email protected] Department of Psychiatry, University of British Columbia, Vancouver, Canada2005 15 11 2005 3 72 72 4 8 2005 15 11 2005 Copyright © 2005 Michalak et al; licensee BioMed Central Ltd.2005Michalak et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. A sizable body of research has now examined the complex relationship between quality of life (QoL) and depressive disorder. Uptake of QoL research in relation to bipolar disorder (BD) has been comparatively slow, although increasing numbers of QoL studies are now being conducted in bipolar populations. We aimed to perform a review of studies addressing the assessment of generic and health-related QoL in patients with bipolar disorder. A literature search was conducted in a comprehensive selection of databases including MEDLINE up to November 2004. Key words included: bipolar disorder or manic-depression, mania, bipolar depression, bipolar spectrum and variants AND quality of life, health-related QoL, functional status, well-being and variants. Articles were included if they were published in English and reported on an assessment of generic or health-related QoL in patients with BD. Articles were not included if they had assessed fewer than 10 patients with BD, were only published in abstract form or only assessed single dimensions of functioning. The literature search initially yielded 790 articles or abstracts. Of these, 762 did not meet our inclusion criteria, leaving a final total of 28 articles. These were sub-divided into four categories (assessment of QoL in patients with BD at different stages of the disorder, comparisons of QoL in Patients with BD with that of other patient populations, QoL instrument evaluation in patients with BD and treatment studies using QoL instruments to assess outcome in Patients with BD) and described in detail. The review indicated that there is growing interest in QoL research in bipolar populations. Although the scientific quality of the research identified was variable, increasing numbers of studies of good design are being conducted. The majority of the studies we identified indicated that QoL is markedly impaired in patients with BD, even when they are considered to be clinically euthymic. We identified several important avenues for future research, including a need for more assessment of QoL in hypo/manic patients, more longitudinal research and the development of a disease-specific measure of QoL for patients with BD. bipolar disorderquality of lifeliterature review ==== Body Review Good quality of life (QoL) encompasses more than just good health. At a basic level, it can represent the sum of a person's physical, emotional, social, occupational and spiritual well-being, the study of which is complicated by the fact that there is no consensus as to what constitutes QoL. The World Health Organization has described QoL as "individuals' perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns" [1]. This broad, generic conceptualization of QoL can be distinguished from the more specific concept of 'health-related quality of life' (HRQOL), which refers to those aspects of an individual's life that impact directly upon their health [2] and the more economically-derived 'cost-utility' models of QoL. This area of research is further complicated by the understanding that QoL can be highly subjective, potentially fluid and open to distortion, making it challenging to measure reliably and accurately. Yet, there is a growing body of evidence to suggest that QoL is an important indicator of well-being, and one that we should be attempting to capture when assessing the patient health. The assessment of QoL in medical settings may be of value in several ways. QoL instruments can provide levels of information not always supplied by traditional outcome measures. For example, some instruments such as the Schedule for the Evaluation of Individualized Quality of Life (SEIQoL) [3] and the Patient Generated Index [4] allow patients to prioritize which life domains are most important to them. While the reduction of symptoms may be the primary goal of the clinician, it may be that the patient places more emphasis upon restoring family relationships, or being able to engage in leisure activities. These 'individualized' measures, although sometimes difficult to administer and interpret, put the patient at the centre rather than at the periphery of assessing the effectiveness of treatment interventions. QoL assessments can also help determine patient preference, allow comparisons of well-being between different conditions and detect subtle differences in response to treatment that may be missed by traditional outcome measures. While a host of studies have now examined QoL in patients with major depressive disorder (MDD) (for example, [5-8] until recently few had specifically focused upon QoL in patients with bipolar disorder (BD). The slow uptake of QoL research in BD may have occurred in part because of the absence of a 'disease-specific' measure of QoL for bipolar populations, or because of reservations about the ability of patients with BD to reliably and accurately complete self-report measures, particularly when in a manic phase. Two reviews of previous research addressing health-related QoL (HRQOL) in BD have been conducted [9,10]. In the first of these, Namjoshi and colleagues (1999) assessed all relevant English-language articles published prior to 1999, identifying 10 studies for inclusion. The studies proved to be quite heterogeneous, and used a variety of generic and depression-specific instruments to assess different aspects of HRQOL. They also tended to be relatively small (only one study had a sample size in excess of 100 patients), were conducted in depressed or euthymic (rather than hypo/manic) patients, and rarely included descriptions of the psychometric properties of the instruments they utilized. The authors of the review made a number of suggestions for future research, including the development of a disease-targeted measure of QoL for BD, more assessments in acutely manic patients, and more longitudinal research. The second review conducted by Dean and colleagues (2004) examined studies that had assessed HRQOL, work-impairment or healthcare costs and utilization in patients with BD published prior to November 2002. The review applied a very broad definition of HRQOL, including in this category studies that had assessed social or physical functioning in isolation (for example, the Global Assessment of Functioning or GAF scale was included as a measure of HRQOL). Using this broad definition, the review identified 65 HRQOL articles. The authors concluded that deficits in HRQOL in patients with BD are similar to those observed in patients with unipolar depression and equal or lower than levels of HRQOL observed in patients with other chronic medical conditions. Given the recent upsurge of interest in describing QoL in BD, the present study aimed to provide an updated literature review of studies that have assessed both generic and HRQOL in patients with bipolar disorder. Materials and methods A comprehensive literature search (supplemented by hand searching where appropriate) was conducted in the following databases up to November 2004: MEDLINE (1966–2004) EMBASE (the Excerptra Medica database) (1988–2004) PubMed (1967–2004) PsychINFO (1967–2004) CINAHL (Cumulated Index to Nursing and Allied Health Literature) (1982–2004) American College of Physicians Journal Club (1991–2004) CDSR (Cochrane Database of Systematic Reviews) (-2004) CCTR (Cochrane Controlled Trials Register) (-2004) DARE (Database of Abstracts of Reviews of Effectiveness) IPA (International Pharmaceutical Abstracts) (1965–2004) Key words used for the search included: bipolar disorder or manic-depression, mania, bipolar depression, bipolar spectrum and variants AND quality of life, health-related QoL, functional status, well-being and variants. Articles were included if they were published in the English language, and reported on the assessment of generic or HRQOL in patients with BD. Our definition of QoL was not overly-inclusive; we required that studies had used a QoL or HRQOL scale that assessed several domains of functioning. Studies using scales that examined single domains of QoL (for example, those assessing solely social or occupational functioning, or single-item scales such as the GAF) were excluded. We omitted studies that included fewer than 10 patients with BD, but did not reject reports for other scientific limitations (for example, convenience sampling or cross-sectional designs). Studies that were underway but were not completed were excluded, as were conference abstracts, dissertations or reports on QoL in BD that were not published in peer-reviewed journals. We also excluded studies that reported assessments in groups of patients with heterogeneous diagnoses where results for patients with BD were not reported separately, and where individual results this population could not be provided by the authors after personal communication (for example, [11-25]. Results The results of the literature search are summarized QUOROM-style in Figure 1. The final 28 included articles are summarized in Table 1. Figure 1 Flowchart of review results. Table 1 Summary of studies assessing quality of life in patients with bipolar disorder Study Location Population(s) QoL instrument(s) Main findings and limitations Arnold et al., (2002) US 44 BD patients (38 type I, 5 type II, I NOS) 30 back pain patients 2474 general population SF-36 HRQOL impaired in BD patients compared to non-clinical sample. Chronic back pain patients more impaired in all SF-36 domains except role limitation (emotional) and mental health. Limitation – disparate sample sizes. Atkinson et al., (1997) Canada 37 BD patients 69 patients with schizophrenia 35 MDD patients QoL index BD and MDD patients subjectively reported lower QoL than patients with schizophrenia, but schizophrenia group had poorer objectively measured QoL. Limitation – relatively small BD and MDD sample sizes. Bond et al., (2000) US 149 patients with SMI (21 with BD) QOLI Mean overall life satisfaction QOLI scores showed mid-range impairment. Limitation – small sample of patients with BD. Chand et al., (2004) India 50 BD patients in remission 20 patients with schizophrenia 20 control subjects Q-LES-Q, WHO-QOL-BREF Patients with BD generally reported better QoL than patients with schizophrenia, and equivalent QoL to control group subjects. Limitation – incomplete matching between groups; unusually low Q-LES-Q scores in control group Cooke et al., (1996)* Canada 68 euthymic BD patients (55 type I, 13 type II) SF-20 SF-20 scores comparable to those reported for patients with MDD. BD type II patients reported poorer HRQOL that BD type I. Limitation – shortcomings of SF-20 compared to SF-36. Dogan et al., (2003) Turkey 26 outpatients with BD stabilized on lithium WHO-QOL-BREF Significant improvement in general health, physical functioning and social functioning 3 months after a psychoeducation intervention. Limitation – small sample size. Kusznir et al., (2000) Canada 61 euthymic BD patients (47 type I, 14 type II) OPQ One third of sample did not meet criteria for adequate community functioning. Limitation – cross-sectional research design. Leidy et al., (1998) US 62 BD patients, type I (34 euthymic, 28 depressed) SF-36, QLDS, MHI-17 and CFS Psychometric properties of instruments generally in acceptable ranges. Marked impairment in SF-36 scores apparent and QLDS scores lower than reported elsewhere for patients with unipolar MDD. Limitation – test-retest reliability was measured over an unusually long period. MacQueen et al., (1997) Canada 62 euthymic BD patients, type I SF-20 No significant differences in SF-20 scores between psychotic and non-psychotic patients. Limitation – small sample of patients with psychotic symptoms. MacQueen et al., (2000) Canada 64 euthymic BD patients, type I SF-20 Number of previous depressive episodes a stronger determinant of HRQOL than number of previous manic episodes. Limitation – number of previous episodes determined retrospectively. Namjoshi et al., (2002) US 139 BD patients, type I SF-36 Acute treatment with olanzapine resulted in improved SF-36 physical functioning scores; improvement in vitality, pain, general health and social functioning domains apparent in open-label phase. Limitation – adjunctive use of lithium and fluoxetine during open-label phase. Namjoshi et al., (2004) US 224 BD patients, type I QOLI Olanzapine cotherapy associated with better outcome in several QOLI domains compared to monotherapy with lithium or valproate. Limitation – only acute QoL outcome data available. Olusina et al., (2003) Nigeria 25 outpatients with BD type I or II WHO-QOL-BREF-TR Majority of sample report 'fair/average' QoL. Small sample of patients with BD, little clinical information for sample. Ozer et al., (2002) Turkey 100 interepisode BD patients Q-LES-Q Depression scores on SADS interview significantly predicted lower Q-LES-Q scores. Limitation – cross-sectional nature of research. Patelis-Siotis et al., (2001) Canada 49 BD mildly depressed or euthymic patients SF-36 SF-36 vitality and role (emotional) scores significantly improved after CBT. Limitation – Open study, and SF-36 scores only available for a sub-set of patients. Perlis et al., (2004) US 983 patients with BD type I, II or NOS Q-LES-Q Younger age of onset of BD predicts Q-LES-Q scores. Revicki et al., (1997) US 28 outpatients diagnosed with DSM-III-R BD SF-36 Onset of BD determined retrospectively. No significant differences in SF-36 domain scores according to mode of administration (in-person vs. telephone). Limitation – small sample size. Revicki et al., (2003) US 120 BD type I patients hospitalized for acute mania Q-LES-Q No differential effects of treatment with divalproex sodium vs. olanzapine on QoL Limitation – only 43% of randomized patients completed Q-LES-Q Ritsner et al., (2002) Israel 17 BD patients (9 manic, 4 depressed, 4 mixed) Q-LES-Q and LQOLP Q-LES-Q scores poorest in depressed patients, highest in manic. Limitation – small sample of patients diagnosed with BD. Robb et al., (1997)* Canada 68 euthymic BD patients (55 type I, 13 type II) IIRS Greater illness intrusiveness associated with higher Ham-D scores, recent depression and BD type II. Limitation – IIRS not validated for use in BD populations. Robb et al., (1998)* Canada 69 euthymic BD patients (54 type I, 15 type II) SF-20 Women possessed significantly lower SF-20 scores in the domains of pain and physical health. Limitation – shortcomings of SF-20 as a HRQOL measure. Russo et al., (1997) US 241 BD inpatients (138 depressed, 103 manic) QOLI Manic BD patients reported better QoL than BD depressed patients. Limitation – lower response rate in acutely manic group. Ruggeri et al., (2002) Italy 22 BD patients LQOLP LQOLP mean scores similar to those observed in larger mixed sample of psychiatric patients. Limitation – small sample of bipolar patients. Salyers et al., (2000) US 164 BD patients SF-12 Mental health scores significantly lower in patients with unipolar depression. Limitation – brief nature of SF-12. Shi et al., (2002) Europe US, South America South Africa 453 BD patients, type I SF-36 Olanzapine superior to haloperidol in improving HRQOL during acute and continuation treatment in most SF-36 domains. Limitation – relatively high drop-out rates during acute treatment phase. Shi et al., (2004) 7 countries 573 BD in/outpatients, type I, most recent episode depressed SF-36, QLDS Olanzapine-fluoxetine combination associated with grater improvement in HRQOL. Limitation – high drop-out rate for an 8-week trial (55%). ten Have et al., (2002) Netherlands 136 BD patients (93 type I, 43 NOS) SF-36 BD sample generally showed greater impairment in SF-36 scores than patients with other psychiatric diagnoses. Limitation – accuracy of CIDI diagnosis of BD NOS in question. Tsevat et al., (2000) US 53 BD patients SF-36, TTO and SG TTO (0.61) and SG (0.70) scores for mental health comparable to those reported for other psychiatric conditions. Limitation – cognitive complexity of TTO and SG tasks. Vojta et al., (2001) US 86 BD patients (16 manic/hypomanic, 26 MDD, 14 mixed, 30 euthymic) SF-12 and EuroQoL SF-12 mental health scores significantly lower in manic group than in euthymic group. MDD/mixed group SF-12 scores significantly poorer than in manic/euthymic groups. Limitation – small sub-samples, brief nature of the SF-12. Wells et al., (1999) US 331 BD patients 944 double depression 3479 MDD 151 dysthymia 987 depressive symptoms SF-12, TTO and SG BD group had lower health utility than MDD, dysthymia and depressive symptoms groups. Limitation – cognitive complexity of TTO and SG tasks. Yatham et al., (2004) 15 countries 920 BD type I patients (currently depressed/experienced episode of depression in previous 60 days) SF-36 SF-36 scores markedly impaired compared to general population norms and consistently lower than sub-scale scores for patients with unipolar MDD. Limitation – depression severity not controlled for. * counted as one study for purposes of review EuroQoL visual analog scale Illness Intrusiveness Rating Scale Lancashire Quality of Life Profile Lehman Quality of Life Interview Longitudinal Interval Follow-up Evaluation Mental Health Index 17 MOS Cognitive Function Scale MOS Short Form 12 MOS Short Form 20 MOS Short Form 36 Occupational Performance Questionnaire Quality of Life Enjoyment and Satisfaction Questionnaire Quality of Life in Depression Scale Quality of Life Index Quality of Life Interview Severe Mental Illness Standard gamble Time tradeoff World Health Organization Quality of Life Assessment Review of studies This section will review the 28 studies we identified. For ease of interpretation they are classified into the following four categories, although several studies met criteria for more than one category. i) Assessment of QoL in patients with BD at different stages of the disorder ii) Comparisons of QoL in patients with BD with that of other patient populations iii) QoL instrument evaluation in patients with BD iv) Treatment studies using QoL instruments to assess outcome in patients with BD i) Assessment of QoL in patients with BD at different stages of the disorder We identified ten studies of QoL in patients with BD at different stages of the disorder. Four of these were generated by a research group in Canada, and will be dealt with in unison. Following this, six other studies (one comparing QoL in patients with during different phases of the disorder, a recent study assessing QoL in bipolar depression, one performed in a Turkish sample of interepisode patients, one conducted in a sample of patients attending a mental health service in Italy, one in recently discharged patients in Nigeria and a report on patients enrolled in the STEP-BD Program) will be described. A research group in Toronto, Canada has generated a series of interrelated reports on QoL in BD. Three of the series [26-28] describe various aspects of QoL in a single sample of outpatients (N = ~68) with BD type I (with manic episodes) or II (with hypomanic episodes) who had been clinically euthymic for at least one month (these have been counted as one study for the purposes of this review). Three of the series report on QoL in other patient populations [29-31]. Cooke and colleagues [26] examined levels of HRQOL using the MOS SF-20, [32] a self-report questionnaire designed to assess perceived well-being in six domains (physical, social and role functioning, mental health status, health perceptions and bodily pain). Mean scores on the SF-20 domains in study patients were comparable to those reported for patients with MDD by Wells and colleagues in the large RAND Corporation MOS Study [33]. Analysis of SF-20 scores by type of BD showed that patients with BD type II reported significantly poorer HRQOL than BD type I in the areas of social functioning and mental health. In another paper, Robb and colleagues [27] reported on functioning in the context of the 'Illness Intrusiveness Model' in patients with BD [34,35]. The model addresses the impact a disorder and/or its treatment has upon an individual's activities across 13 life domains: health, diet, active/passive recreation, work/financial status, self expression/improvement, family relations, relations with spouse, sex life, other relationships, religious expression and community involvement. The Illness Intrusiveness Rating Scale (IIRS) is used to yield a 'total illness intrusiveness' (TII) score. Illness intrusiveness occurred in several areas of functioning, with TII being associated with higher Hamilton Depression Rating Scale (Ham-D) scores, patients having experienced a recent episode of depression and having type II BD. Robb and colleagues [28] specifically focused upon gender differences in SF-20 scores, finding that women possessed numerically lower scores in all of the questionnaire's domains except for mental health, with significant differences in the domains of pain and physical health. Interestingly, objective measures of functioning (clinician rated Global Assessment of Functioning or GAF scores) were not significantly different by gender. MacQueen and colleagues [29] examined SF-20 scores in euthymic BD type I patients (N = 62) with or without psychotic symptoms during an index episode of mania. No significant differences in SF-20 scores were apparent between patients with or without psychosis, although the sample identified with psychosis may have been too small (N = 16) to detect statistically significant differences between sub-groups. Kusznir and colleagues [30] assessed levels of community functioning via the Occupational Performance Questionnaire (OPQ) in a similar population, finding that one-third of patients did not meet criteria for adequate functioning on the 'Community Functioning Scale' component of the questionnaire. Finally, MacQueen and colleagues [31] focused upon the effect of number of manic and depressive episodes on SF-20 and GAF scores in euthymic patients (N = 64), finding that number of past episodes of depression was a stronger determinant of HRQOL than number of previous manic episodes. Good correlation between the subjectively rated SF-20 and objectively rated GAF scores provided some evidence that euthymic patients with BD are capable of providing accurate descriptions of their HRQOL. A potential advantage of this series of studies is that the majority of them were conducted in euthymic outpatients; interepisode patients are likely to be less prone to the effects of cognitive distortion than are symptomatic patients. However, euthymic patients are not necessarily asymptomatic as many have mild sub-syndromal symptoms, and several studies in this review will demonstrate that even residual depressive symptoms can be strongly associated with impaired QoL. The relationship between QoL and hypo/mania is less well understood. Both mania and hypomania can be associated with substantial depressive symptomatology, either in the form of 'dysphoric mania/hypomania' or when the patient experiences a mixed episode. This understanding led Vojta and colleagues to hypothesize that patients with manic symptoms would report significantly lower QoL than would patients who were euthymic [36]. To test this theory, the authors administered two brief self-report measures (the SF-12 and the EuroQoL visual analog scale) in bipolar patients with mania/hypomania (N = 16), MDD (N = 26), mixed mania/hypomania and depression (N = 14) or who were euthymic (N = 30). In keeping with their hypothesis, patients with mania/hypomania did show significantly lower SF-12 mental health scores than euthymic patients, with depressed or mixed patients showing significantly poorer HRQOL again. Mean EuroQoL scores ran in the same direction, although the difference between euthymic and manic/hypomanic patients was not significant. In the largest study to date of QoL in bipolar depression, Yatham and colleagues have reported on SF-36 scores in BD type I patients (N = 920) who were either currently depressed, or had experienced a recent episode of depression [37]. SF-36 scores were remarkably low in the role-physical, vitality, social functioning, role-emotional and mental health sub-scales (see Table 2). The authors went on to compare these scores with those derived from seven large (>100 outpatients) studies of HRQOL in unipolar depression that had also administered the SF-36. Sub-scale scores tended to be lower in the bipolar sample than in the unipolar sample, with the exception of the bodily pain sub-scale, where unipolar depressives tended to exhibit higher scores. Mean SF-36 scores were significantly (weakly: range -0.1 to -0.3) negatively correlated with HAM-D scores, providing some evidence for the construct validity of the instrument in this population. Whilst this study is robust in terms of its large sample size and well-described clinical population, it did not control for depression severity or demographic variables in between-group comparisons. Furthermore, diagnosis of bipolar disorder was made by careful clinical interview, whereas unipolar depression was diagnosed via a number of subjective and objective methods. Table 2 Summary of studies using the SF-36 to assess quality of life in patients with bipolar disorder1 Study Patient population Physical Social Role physical Role emotional Pain Mental health General health Vitality Arnold (2000) 44 BD outpatients 78.8 ± 22.4 57.9 ± 27.7 63.1 ± 41.6 38.6 ± 43.1 64.9 ± 25.7 55.3 ± 23.8 61.9 ± 25.4 43.6 ± 24.3 Have (2002)2 93 BD type I 43 BD NOS 89.6 91.2 73.6 80.8 77.6 81.7 69.5 80.6 74.1 82.5 62.3 68.7 62.6 68.2 58.0 62.0 Leidy (1998) 34 euthymic 28 depressed 84.4 ± 20.2 72.2 ± 28.3 73.2 ± 18.2 29.3 ± 20.0 86.2 ± 28.0 32.3 ± 38.6 76.2 ± 31.2 8.3 ± 20.3 59.6 ± 29.0 54.7 ± 25.3 69.2 ± 17.9 33.4 ± 16.5 70.9 ± 20.7 58.0 ± 21.2 52.0 ± 16.2 20.4 ± 17.5 Namjoshi (2002) 122 BD type I (manic/mixed) 65 olanzapine 57 placebo 86.8 ± 16.8 84.5 ± 21.9 47.1 ± 28.3 46.0 ± 31.8 70.4 ± 40.2 65.4 ± 40.3 37.4 ± 42.3 36.3 ± 43.3 68.4 ± 26.4 61.7 ± 25.0 59.9 ± 22.6 58.5 ± 19.8 69.0 ± 22.7 65.2 ± 24.3 63.3 ± 24.0 66.6 ± 20.0 Patelis-Siotis (2001) 34 BD CBT completers 8 BD CBT non-completers 80.4 ± 19.3 63.8 ± 30.6 58.1 ± 25.0 46.9 ± 28.1 41.2 ± 39.8 40.6 ± 44.2 17.6 ± 33.1 29.2 ± 41.5 68.5 ± 23.7 63.4 ± 27.0 52.4 ± 18.0 44.0 ± 22.0 66.6 ± 21.7 46.4 ± 29.6 39.4 ± 19.3 28.1 ± 21.4 Revicki (1997) 14 BD patients (in-person) 14 BD patients (by telephone) 78.4 ± 25.2 77.0 ± 29.3 53.6 ± 30.2 57.1 ± 29.9 65.2 ± 38.7 59.8 ± 41.0 40.5 ± 42.9 33.3 ± 40.6 68.0 ± 31.8 69.3 ± 28.2 53.4 ± 22.8 53.9 ± 20.0 59.8 ± 22.8 57.5 ± 22.7 41.4 ± 18.7 41.4 ± 20.5 Tsevat (2000) 53 BD patients 78.7 ± 23.4 58.7 ± 27.9 63.2 ± 40.9 38.9 ± 42.3 65.3 ± 26.0 56.2 ± 23.7 62.1 ± 24.3 45.4 ± 24.4 Shi (2002) 453 BD type I 234 olanzapine 219 haloperidol 85.2 ± 23.2 90.5 ± 15.7 61.1 ± 31.8 61.2 ± 29.1 66.1 ± 39.6 72.8 ± 36.3 53.3 ± 43.1 50.1 ± 43.7 79.8 ± 26.2 81.2 ± 26.1 71.0 ± 20.4 72.8 ± 16.5 73.6 ± 21.8 75.1 ± 19.2 75.8 ± 19.1 80.0 ± 14.9 Shi (2004) 573 BD type I (currently depressed) 250 olanzapine 58 olanzapine/fluoxetine combination 265 placebo 65.8 ± 27.6 68.8 ± 25.0 66.6 ± 26.2 29.1 ± 20.9 30.6 ± 20.8 32.5 ± 21.4 47.8 ± 44.0 44.8 ± 41.8 46.4 ± 42.3 12.9 ± 25.4 9.8 ± 23.4 14.6 ± 28.7 60.6 ± 27.1 60.8 ± 25.6 57.8 ± 26.1 30.0 ± 16.1 31.0 ± 17.3 31.3 ± 15.7 51.1 ± 22.3 52.3 ± 20.7 48.6 ± 22.6 25.5 ± 17.5 25.3 ± 19.0 25.6 ± 17.6 Yatham (2004) 920 BD type I (currently depressed/ depressive episode in previous 60 days) 70.2 ± 26.2 29.9 ± 22.8 36.7 ± 40.9 11.4 ± 23.5 62.2 ± 27.1 31.0 ± 17.3 47.5 ± 23.3 22.4 ± 17.7 1 Data presented as mean ± SD 2 SDs not available A study by Ozer and colleagues [38] assessed 100 interepisode patients with BD in Turkey with the aim of examining the impact of 'history of illness' and 'present symptomatology' factors upon a variety of outcome measures including the Schedule for Affective Disorder and Schizophrenia (SADS) and Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q) [39]. The Q-LES-Q is a 93-item self-report measure of the degree of enjoyment and satisfaction in various areas of daily living. The questionnaire was developed and validated for use in depressed outpatients and has eight summary scales that reflect major areas of functioning: physical health, mood, leisure time activities, social relationships, general activities, work, household duties and school/coursework. Mean Q-LES-Q scores can be derived from the eight summary scales and range from 0–100, where higher scores indicate better QoL. Using multivariate analysis, Ozer and colleagues found that none of the historical variables (including age at first episode, number of previous depressive/manic episodes, duration of illness, number of hospitalizations, age at first hospitalization, or number of symptoms during first episode) were predictive of mean Q-LES-Q scores. Of the current symptoms assessed, only the depression subscale of the SADS interview significantly predicted lower Q-LES-Q scores, accounting for only 13% of the observed variance. When the patient population was subdivided into three groups (low, moderate and high) according to severity of SADS depression scores, mean Q-LES-Q scores were 39%, 38% and 35%, respectively. In comparison, mean Q-LES-Q scores have been reported to be 42% in hospitalized psychiatric inpatients [40], 42% in outpatients with MDD [41], 44% in patients with seasonal affective disorder (SAD) [41], 53% in patients with chronic MDD [42], and 83% in the general population (Rapaport, personal communication). Ruggeri and colleagues [43] investigated the relationship between QoL and a variety of clinical and demographic variables in a community-based sample of patients (N = 268) with mixed psychiatric diagnoses, 22 of whom were bipolar. QoL was assessed via the Lancashire Quality of Life Profile (LQOLP), which assesses perceived well-being and functioning in 9 major life domains on a 7-point Likert scale (where higher scores indicate better QoL). We extracted LQOLP results for the bipolar sample from data provided by the authors, finding that mean satisfaction scores for the 9 domains were 4.4 ± 1.0, a score similar to that reported for the entire patient sample. In another study of recently discharged Nigerian outpatients (N = 25) with BD type I or II, World Health Organization Quality of Life Assessment (WHO-QOL-BREF-TR) scores were reported to be 'good' in 5 (20%) of patients, 'fair/average' in 14 (56%) and 'poor' in 6 (24%) of patients (data by WHOQOL-Bref domain also provided by authors during personal communication) [44]. Finally, Perlis and colleagues have recently provided an analysis of 'early onset' in 983 patients (BD type I, II or NOS) enrolled in the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) [45] in which QoL was assessed. The multicentre STEP-BD program, a large prospective, naturalistic study than combines several randomized-controlled trials, has selected to use the Q-LES-Q to assess QoL and the GAF and 'Range of Impaired Functioning Tool' (LIFE-RIFT) to measure functional status. Perlis and colleagues provide the first report on QoL from the project, having looked specifically at the effect of age of onset (grouped into 'very early age, <13 years', 'early age, 13–18 years' and 'adult, > 18 years') of mood symptoms in BD upon outcome. Younger age of onset was found to be a significant predictor of Q-LES-Q scores at study entry (where treatment and clinical status would have varied widely between patients), but not of functioning as measured by the GAF or LIFE-RIFT. These results represent early data from a study that has the potential to address several important questions surrounding QoL in BD. ii) Comparisons of QoL in patients with BD with that of other patient populations We identified five studies comparing QoL between patients with BD and patients with other conditions. Two of these used the SF-36, one used the 'Quality of Life Index', the Q-LES-Q and the WHO-QOL-BREF and one applied a 'health utilities' model. The SF-36 [46] is currently the most widely used measure of HRQOL [47]. The self-report questionnaire contains eight sub-scales assessing physical functioning, social functioning, role limitations (physical), role limitations (emotional), pain, mental health, general health and vitality. These yield an overall domain score on a 0–100 scale, where 0 represents worst possible health and 100 best possible health. Arnold and colleagues [48] compared SF-36 scores between patients with BD (N = 44) and chronic back pain (N = 30) with norms previously reported for a general population sample (N = 2,474) [49]. The results of the study indicated that HRQOL was compromised in all SF-36 domains except physical functioning in patients with BD compared with the general population sample (see Table 2). The BD group fared better than the back pain group in the physical, role limitation (physical), pain and social domains, although no significant differences were observed in terms of role limitation (emotional) or mental health domains. While the study provides a useful initial comparison of HRQOL between BD and other conditions, its findings should be interpreted with some caution owing to the disparate sample sizes involved. It also utilized previously published norms for the SF-36 that had been derived by different data collection methods. The Netherlands Mental Health Survey and Incidence Study (NEMESIS) has examined the epidemiology of psychiatric disorders in a large general population sample [50]. Using the Composite International Diagnostic Interview (CIDI), 136 adults were identified with DSM-III-R lifetime BD (93 with BD type I and 43 with BD NOS) and administered the SF-36. Participants with BD showed significantly more impairment in most of the questionnaire's domains compared with NEMESIS subjects diagnosed with other psychiatric disorders (SF-36 scores for the BD group are presented in Table 2). For example, in the domain of mental health, participants with BD type I experienced significantly lower mean scores (62.3) than people with other mood (75.2), anxiety (74.0), substance use (80.2) or no psychiatric disorders (85.8). BD type I subjects also experienced significantly lower SF-36 scores than patients with BD NOS in the domains of mental health, role limitations (emotional), social functioning and pain. However, there remains some controversy about the accuracy with which the CIDI detects BD NOS, limiting somewhat the inferences that can be made on the basis of these sub-group results. A later analysis of a sub-set (N = 40) of the original NEMESIS sample administered the EuroQol: 5 Dimensions (EQ-5D) scale, which can be used to provide health utility values [51]. Mean utility values (see below) for the sample were reported to be 0.82 ± 0.20, comparable to those observed in the general population of the Netherlands. Atkinson and colleagues [52] used a different measure, the 'Quality of Life Index' [53], to assess QoL in patients with BD (N = 37), MDD (N = 35) or schizophrenia (N = 69). The authors found that subjectively reported QoL was lower in patients with BD and MDD than in those with schizophrenia. Interestingly, this trend was reversed for objectively assessed QoL, which included measures such as medical history, health risk behaviors, educational and financial levels and social functioning. These findings led the authors to speculate about the validity of subjective measures of QoL, particularly in people with affective disorders. These results were not replicated in Indian by Chand and colleagues, who compared the QoL of patients with BD (in remission and stabilized on lithium prophylaxis, N = 50) with patients with schizophrenia (N = 20) and healthy controls (N = 20) [54]. Using the Q-LES-Q and the WHOQOL-BREF, the authors found that the bipolar group reported significantly better QoL than the schizophrenia group in all domains of the Q-LES-Q, and in general well-being, physical health and psychological health on the WHO scale. Surprisingly, the authors also observed that perceived QoL was equivalent between patients with BD and healthy controls, with the exception of the Q-LES-Q leisure domain, where the patient group actually reported better functioning. Having said this, mean Q-LES-Q scores for this particular control group were unusually low (approximately 47%, where general population norms for the United States are around 83%, Rapaport, personal communication). Although a growing number of studies have now evaluated the 'health utilities' and 'health preferences' of patients with physical conditions, relatively few have examined these values in patients with mental illnesses, including BD. The concept of health utility refers to an individual's preferences for different health states under conditions of uncertainty. Health preferences are values that reflect an individual's level of subjective satisfaction, distress or desirability associated with various health conditions. Health utility and preferences are frequently assessed by the 'time tradeoff' (TTO) and 'standard gamble' (SG) approaches [55]. TTO refers to the years of life a person is willing to exchange for perfect health. For example, patients might be asked to imagine that a treatment exists that would allow them to live in perfect physical and mental health, but reduces their life expectancy. They might then be asked to indicate how much time they would give up for a treatment that would permit them to live in perfect health, if they had ten years to live. SG refers to the required chance for successful outcome to accept a treatment that could result in either immediate death or perfect health. For example, patients might be asked to imagine that they had ten years to live in their current state of health, and that a treatment existed that could either give them perfect health, or kill them immediately. Patients might then be asked to indicate what chance of success the treatment would have to have before they would accept it. Health utility and preference values are frequently expressed as a score of 0 to 1, with higher values representing better health. We identified one study comparing health utility in patients with BD with other patient populations. Wells and colleagues (1999) [56] assessed functioning and utility in patients with depression or chronic medical conditions within seven managed care organizations in the United States. HRQOL was assessed via the global mental and physical scales of the SF-12 and utility was measured via TTO and SG. Patients with depression were categorized as those with BD (N = 331), 12-month MDD (N = 3479), 12-month double depression (N = 944), 12-month dysthymia (N = 151) or brief subthreshold depressive symptoms (N = 987). In terms of HRQOL, the bipolar group showed levels of impairment second only to patients with double depression. Utility was also lower in the bipolar group compared with patients with MDD, dysthymia or brief depressive symptoms, although not double depression. In terms of health utility, bipolar patients were willing to give up on average 17% of their life expectancy in return for perfect health, and would accept on average an 11% risk of death in exchange for perfect health. In comparison, patients with MDD were willing to give up 11% of their life expectancy, and accept a 6% risk of death. iii) QoL instrument evaluation in patients with BD We identified five studies that had evaluated different QoL instruments in BD populations and one study that examined the effects of mode of questionnaire administration. The first of these examined the application of the aforementioned health utility approach. The second assessed the psychometric properties of the Lehman Qualify of Life Interview (QOLI) in a heterogeneous sample of psychiatric inpatients. The third evaluated four QoL scales in a smaller sample of Patients with BD, while the fourth assessed the MOS SF-12 in a large population of patients with severe mental illness. The fifth study evaluated the properties of the Q-LES-Q and the LQOLP in a sample of Israeli patients with severe mental disorders. The final study we identified examined telephone versus in-person health status assessment in outpatients with BD. Tsevat and colleagues (2000) [57] examined functional status and health utility in 53 outpatients with BD recruited from one site of the multicenter Stanley Foundation Bipolar Network study. The authors aimed to assess how patients with BD rated their current overall health versus their current mental health, and to determine the extent to which health utility correlated with disease state. TTO scores for current overall health were 0.71, but were significantly higher than scores for current mental health, which averaged 0.61. In other words, patients with BD were willing to give up on average 39% of their life expectancy in return for perfect mental health. These values are similar to TTO values obtained in the Beaver Dam Health Outcomes Study in patients with depression (0.70) or anxiety (0.77). SG scores were not significantly different for overall health (0.77) and mental health (0.70). SF-36 scores for the study are presented in Table 2. Certain SF-36 domains (general health, vitality and role-emotional) were significantly correlated with mental health TTO and SG scores, but levels of mania were not correlated with utilities for either overall health or mental health. The authors concluded that health utilities may be related to certain health status attributes and to level of depression, but may not be related to level of mania in patients with BD. One advantage of the health utility/preference approach to QoL assessment is that it allows the calculation of quality-adjusted life years (QALYs). QALYs are a commonly used outcome measure in cost-effectiveness studies, but our literature search did not find any studies that had calculated QALYs for BD populations. Russo and colleagues [58] performed a rigorous psychometric evaluation of the QOLI [59] in a large sample (N = 981) of acutely ill hospitalized psychiatric inpatients. Of these, 138 were diagnosed according to DSM-III-R criteria with bipolar depression, 103 with acute mania and the remainder with unipolar depression, schizophrenia, or 'other' diagnoses. The QOLI contains 44 items and 7 satisfaction scales, a global satisfaction item and 14 functional items, with all satisfaction scores ranging from 1 (terrible) to 7 (delighted). Patients were administered the instrument using a structured interview procedure within 48 hours of admission and discharge. While the QOLI was successfully completed by 90% of patients overall, rates did vary according to patient diagnoses with non-completion rates being lowest in patients with bipolar depression (12%) and highest in manic patients (31%). Reasons given for non-completion of the measure varied, the most common being 'inadequate staff time' (39%), 'patient too psychotic, demented, or confused' (13%), or 'too agitated or sleepy' (12%). The QOLI showed good psychometric properties overall, although there was some concern about an apparent lack of construct consistency (low correlations between satisfaction and functional measures) in patients with mania. Analysis of QOLI sub-scales showed that, broadly speaking, manic patients reported the highest levels of satisfaction and function, with bipolar and unipolar depressed patients reporting the lowest levels. Leidy and colleagues [60] examined the psychometric properties of four QoL measures in 62 BD type I patients (34 euthymic, 28 depressed). Patients completed the SF-36, the Quality of Life in Depression Scale (QLDS), the Mental Health Index 17 (MHI-17) and the MOS Cognitive Function Scale (CFS). The study provided further evidence that both euthymic and depressed patients with BD are capable of providing subjective reports of their HRQOL. Baseline SF-36 scores were markedly impaired in the depressed sub-group, with the vitality, social and role limitation (emotional) domains all falling below the 25th percentile (see Table 2). QoL as measured by the QLDS was poorer than has been reported elsewhere for patients with unipolar depression. Cronbach's alpha scores for the QLDS, MHI-17, CFS and four of the eight SF-36 sub-scales (physical functioning, role physical, vitality and metal health) all fell above the generally accepted level of 0.80. Test-retest reliability for the scales were modest (intraclass correlations ranged between 0.18 on the SF-36 role emotional scale and 0.80 for physical functioning), although the reliability of the scales was assessed over an unusually long time period (8 weeks). Scores on the QLDS, MHI-17 and CFS were significantly correlated with patients' Ham-D scores, as were several of the SF-36 sub-scales, thus confirming the construct validity of the scales in patients with BD. Finally, the MHI-17, CFS, QLDS and SF-36 vitality, role emotional and mental health sub-scales were shown to be responsive to change in depression status over time; the QLDS has recently been successfully used as an outcome measure in a large pharmaceutical treatment trial (Shi et al., 2004, see section IV). Salyers and colleagues [61] conducted a psychometric evaluation of another MOS instrument, the SF-12 [62], in a sample of 946 patients with severe mental illness, 164 of whom were diagnosed with BD. Mean (± SD) SF-12 physical functioning and mental functioning scores for the bipolar group were 46.1 ± 11.5 and 39.6 ± 12.7 respectively, although mental health functioning scores were significantly lower (31.8 ± 13.4) in patients with unipolar MDD. The instrument showed acceptable levels of reliability and validity in the entire sample, although it is worth noting that is was administered by trained interviewers, rather than as a self-report measure. Ritsner and colleagues [63,64] have compared responses on the Q-LES-Q and the LQOLP in a sample of 175 non-clinical controls and 199 Israeli patients with severe mental illness (SMI), 17 of whom were diagnosed with BD. In personal communication with the authors, we were informed that mean Q-LES-Q scores for the manic, depressed and mixed sub-groups of Patients with BD in the study were 40%, 25% and 33% respectively. Both instruments showed generally acceptable levels of internal consistency, test-retest reliability and criterion validity (in the entire patient population) but notably low levels of convergent validity between the instruments' domains, particularly in the control group. Finally, Revicki and colleagues (1997) [65] examined the effects of administering the SF-36 either in person or by telephone in 28 patients with BD (see Table 2). SF-36 domain scores were not significantly affected by mode of administration. iv) Studies using QoL instruments to assess outcome in patients with BD We identified eight studies that had used a QoL measure to assess outcome in BD populations: five clinical trials that examined pharmacological interventions for the disorder and three studies that assessed non-pharmacological interventions. Namjoshi and colleagues from a Lilly research group have conducted a series of studies examining the impact of treatment with olanzapine upon QoL [66-70]. In the first, Namjoshi et al., (2002) evaluated the impact of acute (3-week) treatment with olanzapine or placebo and long-term (49-week open label) treatment of BD type I (manic/mixed). Baseline SF-36 scores for the olanzapine and placebo group are shown in Table 2. During acute-phase treatment, a significant improvement was observed in the physical functioning domain of the SF-36 in the olanzapine group. During the open label treatment period, however, the SF-36 bodily pain, vitality, general health and social functioning domains showed significant improvements over time. This may indicate that olanzapine has a relatively rapid effect in terms of improving physical functioning in patients with acute mania, but that treatment may be required for longer periods for functioning to improve in other QoL domains. Shi and colleagues have also compared the treatment effects of olanzapine and haloperidol in patients with acute mania (N = 453) [67,68] weeks of acute-phase treatment, significantly greater improvement in five of the SF-36 domains (general health, physical functioning, role limitations – physical, social functioning and vitality) was apparent in the olanzapine group. Superiority of olanzapine over haloperidol persisted over the study's 6-week continuation phase, with concomitant improvements in work and household functioning. Baseline SF-36 scores for the olanzapine and haloperidol groups are shown in Table 2. A further study examined the effects of adding olanzapine to lithium or valproate in patients with BD (N = 224) [69]. Olanzapine cotherapy was associated with better outcome in several QOLI domains compared to monotherapy with lithium or valproate alone. The SF-36 and QLDS have been used in a study comparing the benefits of olanzapine alone versus an olanzapine-fluoxetine combination or placebo [70]. Compared with placebo, patients who received olanzapine showed greater improvement at 8 weeks in SF-36 mental health summary scores, and in mental health, role-emotional and social functioning domain scores (SF-36 scores summarized in Table 2). The combination group fared significantly better in terms of HRQOL improvement than the olanzapine-alone group, showing improvement in 5 of the SF-36 domain scores and in QLDS total score. The authors also performed a psychometric evaluation of the QLDS (see section III). Finally, the Q-LES-Q has been administered at baseline (hospital discharge), 6 and 12 weeks in a comparison of divalproex sodium and olanzapine in the treatment of acute mania [71]. No significant treatment effects were detected in Q-LES-Q scores in the study, although only 52 (43%) of the 120 patients randomized to either divalproex or olanzapine completed the QoL instrument. Interestingly, the authors reported an association between weight gain being reported as an adverse event and poorer change scores in the physical, leisure, and general activities domains of the Q-LES-Q at 6 weeks (but not at 12 weeks). Negative correlations were reported between increased weight (at 6 weeks) and overall life satisfaction, physical health, mood, general activities and satisfaction with medication on the Q-LES-Q. Although current recommendations favor the use of pharmacological treatments such as lithium and mood stabilizers in the initial treatment and symptom control of BD, there is increasing recognition of the role of psychotherapy in the management of the disorder. We identified one study that had used a QoL tool to assess outcome following a psychotherapy intervention for BD. Patelis-Siotis and colleagues [72] used the SF-36 in a feasibility study of group cognitive behavior therapy (CBT) in patients with BD. Although baseline SF-36 data was available for 42 patients (see Table 2), pre and post intervention data was only available for a proportion of participants (N = 22) as completion of the QoL questionnaires was optional. Nevertheless, SF-36 vitality and role emotional scores were significantly improved following CBT, with an accompanying trend towards improved social functioning. Another study we identified specifically examined the effects of vocational rehabilitation upon outcome in 149 patients with SMI, 21 of whom were diagnosed with bipolar disorder [73]. In personal communication with the authors, we learned that mean (± SD, range 1–7 where higher scores indicate better QoL) baseline QOLI 'overall life satisfaction' scores were 4.7 ± 1.1, with 'general satisfaction' domain scores of 4.9 ± 1.3. Although outcome data was not available specifically for the bipolar group, better QoL outcomes were associated with 'competitive work activity' in the overall sample compared to other reduced forms of work activity. Finally, a recent study has examined the effects of providing 3 sessions of psychoeducation about lithium treatment to patients (N = 26) with BD [74]. In addition to assessing the effects of psychoeducation upon medication adherence, the authors examined the impact of education upon QoL, as measured by the WHO-QOL-BREF. Following psychoeducation, patients in the intervention arm of the study showed significant improvement in 2 of the WHOQOL BREF's 4 domains (physical health and social functioning) and in overall perceived health. Patients in the control arm of the study, in comparison, showed no significant changes in their perceived QoL. The results of the study indicate that it may be possible to alter patients' perceptions of their QoL even with relatively brief psychological interventions. Discussion Prior to 1999, only 10 studies had systematically addressed the measurement of HRQOL in patients with bipolar disorder [10]. A second review of studies that had examined HRQOL in BD prior to 2004 identified 65 studies [9]. We conducted a subsequent review of studies examining generic and HRQOL in bipolar disorder that had been published prior to November 2004. Our literature search identified 28 studies in total, 7 (25%) of which were published before 1999 (there is discrepancy in the number of studies identified prior to 1999 between the two reviews due to differing inclusion criteria). The remaining 21 (75%) were published between 2000 and 2004, indicating that there is developing interest in this field of research. The studies we identified were quite heterogeneous in nature. Several undertook to assess QoL during different phases of the disorder, for example, cross-sectional research that compared perceived QoL in euthymic, manic or depressed patients with BD. Other studies compared QoL in bipolar samples to that of other patient populations, both with other psychiatric disorders and with chronic physical conditions. Another vein of research examined the psychometric properties of a variety of generic and HRQOL instruments in BD populations. Finally, we identified several studies that had used a QoL instrument to assess outcome in trials of pharmacological and non-pharmacological treatment inventions for the condition. The studies were also of variable scientific quality. Methodological shortcomings included small sample sizes, cross-sectional designs, idiosyncratic diagnostic methods or undifferentiated diagnostic groups, and use of inappropriate or poorly validated QoL instruments. This being said, the overall scientific quality of research in this field does appear to be improving. Of the 10 studies identified in the review by Namjoshi and colleagues, only one possessed a sample of size of more than 100 patients with BD. In comparison, we identified eleven studies that had enrolled more than 100 patients. It was particularly encouraging to see that some of the large pharmacological trials of treatment interventions for BD are now using QoL measures as secondary outcome measures. Clinical trials in bipolar populations have traditionally used objectively rated measures such as rates of relapse, hospitalization or symptom reduction to assess patient outcome. However, the concomitant use of QoL indices does appear to pay dividends. For example, Namjoshi and colleagues [66] found that the timing of response to treatment with olanzapine differed in terms of symptomatic and QoL outcome. Symptom reduction in the olanzapine group occurred relatively quickly, with patients showing a 10-point decrease in Young Mania Rating Scale (YMRS) scores during the study's 3-week acute treatment phase. Improvements in SF-36 scores, however, occurred more slowly. Only the domain of physical functioning showed significant improvement by the end of the acute treatment phase, whereas it was the domains of social functioning, general health, vitality and bodily pain that were significantly improved at the end of the 49-week maintenance phase. These findings are in accordance with other research showing that 98% of first episode mania patients achieve syndromal recovery after 24 months, but only 38% achieve functional recovery [75]. Sole reliance on symptomatic outcome measures may not detect these more subtle changes in well-being, functioning and QoL. Although there appears to be increasing use of QoL measures in pharmacological research in bipolar populations, we identified surprisingly few studies of psychological interventions that had incorporated a QoL assessment. In a review of psychosocial interventions for BD, Huxley and colleagues (2000) [76] identified 32 peer-reviewed reports examining group, couple/family or individual psychotherapy in BD, none of which systematically assessed QoL. Our own review only identified one relevant publication, a feasibility study of group CBT which incorporated the SF-36 [72]. As Huxley and colleagues note, future research in this area should employ much broader measures of outcome, such as the assessment of QoL, which may be less amenable to pharmacological treatment in isolation. An important general conclusion from this review is that the measurement of QoL is feasible in some patients with BD. However, there remains a paucity of information about the ability of patients in the hypo/manic phase of their illness to reliably and accurately complete QoL measures. One of the more rigorous studies to date was performed by Russo and colleagues (1997) [58], in which nurses administered the QOLI via a structured interview procedure to 103 patients with acute mania. Completion rates for the questionnaire were 69% in acutely manic patients, compared to 88% in bipolar depressed patients. In the Namjoshi et al., [66] study of olanzapine in patients with acute mania, SF-36 scores were successfully obtained for 122 of 139 (88%) of patients who entered the study's randomization phase. In the smaller study by Vojta and colleagues [36], two brief QoL instruments were successfully completed by 16 patients with mania/hypomania. More research is needed to ascertain how feasible it is to administer self-report measures of QoL in patients with hypo/mania, although other research in bipolar populations has indicated that patients with mild to moderate manic symptoms can provide reliable descriptions of their symptoms [77]. Although there is controversy about the validity of the technique, [78] proxy measures of QoL can be obtained from family members or clinicians, and may offer one solution to this problem. Additional research is also needed to determine the longitudinal course of QoL in patients with BD. The majority of the studies we identified were cross-sectional in nature. Useful future research would longitudinally follow the course of QoL in a cohort of patients with BD as they experienced different phases of the illness. Research is also required in first-episode mania patients to help elucidate the relationship between length of illness, episode frequency and QoL. The studies we identified were also heterogeneous in terms of the QoL instruments they incorporated. By far the most frequently utilized were the MOS range of HRQOL measures; 16 of the 28 studies we identified (57%) utilized the SF-12, SF-20 or the SF-36. The results of studies using the SF-36 were amalgamated in Table 2. Inspection of these data indicates that, in general, physical functioning appears to be relatively good in patients with BD (range 63.8 – 91.2). Mental health scores are unsurprisingly much lower (range 30.0 – 72.8). In comparison, SF-36 mental health functioning scores have been reported to be 40.0 (± 17.5) in primary care patients (N = 536) initiating treatment for depression [79]. However, it remains difficult to make any broad generalizations on the basis of this grouped data owing to differences in patient populations, recruitment methods and sample sizes that result in wide ranges of scores in some domains. Given the breadth of existing data for the SF-36 in bipolar populations, the scale's acceptable psychometric properties and detailed normative data, we recommend this scale for the measurement of health-related QoL in patients with BD. The WHO-QOL-BREF is an alternative that has undergone rigorous international development and is available in a wide variety of languages. A number of other QoL instruments have been utilized in bipolar populations, including the Illness Intrusiveness Rating Scale (IIRS), the Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q), the Lehman Quality of Life Interview (QOLI), the WHO-QOL-BREF and the health utility time tradeoff (TTO) and standard gamble (SG) approaches. However, only a small number of the studies we identified reported on the psychometric properties of these instruments, and it remains the case that few measures of QoL have been appropriately evaluated for use in BD populations. Some of these instruments are semi disease-targeted in the sense that they have been developed in and for depressed populations (i.e. the Q-LES-Q and QLDS). Both the Q-LES-Q and QLDS appear to possess acceptable psychometric properties and, importantly, are responsive to change in response to both psychological and pharmacological treatment interventions and can be recommended here for use in bipolar populations. There is at present no QoL measure specifically designed for use in bipolar populations. Although existing QoL instruments are likely to capture key aspects of QoL, they may be insensitive to some of the unique problems posed by this complex psychiatric condition. For example, few QoL instruments designed for use in psychiatric populations assess routine, independence, spirituality or stigma, which have been sown to have particular bearing upon QoL in patients with BD [80]. Bipolar disorder is also unique in that it can be characterized by a variety of mood states, including hypo/mania, depression or mixed states. The understanding that mania can also be associated with depressive symptoms, and that patients will experience periods when they are relatively euthymic, complicates the assessment of QoL in this population. We suggest that an important step forward in this field of research would be made with the development of a disease-specific QoL instrument for BD. We believe that such an instrument would need to have a number of qualities. It would have to work effectively in the depressed, hypo/manic, mixed and euthymic phases of BD. It would need to be concise enough not to put overdue response burden on the patient, but detailed enough to tap into the major areas of well-being affected by the disorder. The relevance of the scale would need to be ensured by thorough consultation with patients, their families and their clinicians. This process should involve individual qualitative interviews and focus group work with patients with BD and their family members at different stages of the disorder, and consultation with psychiatrists, mental health workers, and public-sector organizations. It would be useful if the instrument was available in self-report, interviewer-administered and proxy-respondent formats to provide alternative methods of administration in acutely manic populations. Finally, the psychometric properties of the instrument would need to be carefully evaluated in terms of reliability, validity, responsiveness and other standard psychometric assessments. Conclusion In recent years, major developments in the pharmacological control of bipolar disorder have occurred. One result of these improvements has been that some patients will BD now experience fewer side effects and less physical symptomatology, allowing the focus to shift to other concerns, including improving inter-episode functioning and perceived quality of life. Our review found that there is growing interest in characterizing QoL in bipolar disorder populations, and determining the impact of treatment interventions upon life quality. The scientific quality of research in this field has been variable, but increasing numbers of studies of good design are now being conducted. We highlighted several important avenues for future research, including the need for assessments of QoL in first episode and hypo/manic patients, more well-designed longitudinal research and research into the impact of psychosocial interventions upon QoL, and the development of a disease-specific measure for use in bipolar populations. Bipolar disorder creates a major health concern, both for the individual and for society, and more information is still needed about the impact of the condition upon QoL. Such information may then bring us one step closer towards developing treatment regimens that maximize both symptom reduction and quality of life for patients with this complex condition. ==== Refs The WHOQOL Group The World Health Organization Quality of Life Assessment (WHOQOL): Position Papaer From the World Health Organization. Social Science Medicine 1995 10 1403 1409 Patrick DL Erickson P Walker SR and Rosser RM Assessing health-related quality of life for clinical decision making. Quality of life assessment Key issues in the 1990s 1993 Lancaster, Kluwer 11 64 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 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 Angermeyer MC Holzinger A Matschinger H Stengler-Wenzke K Depression and quality of life: results of a follow-up study Int J Soc Psychiatry 2002 48 189 199 12413247 10.1177/002076402128783235 Doraiswamy PM Khan ZM Donahue RM Richard NE Quality of life in geriatric depression: a comparison of remitters, partial responders, and nonresponders Am J Geriatr Psychiatry 2001 9 423 428 11739069 10.1176/appi.ajgp.9.4.423 Kennedy SH Eisfeld BS Cooke RG Quality of life: an important dimension in assessing the treatment of depression? J Psychiatry Neurosci 2001 26 Suppl S23 S28 11590966 Revicki DA Simon GE Chan K Katon W Heiligenstein J Depression, health-related quality of life, and medical cost outcomes of receiving recommended levels of antidepressant treatment J Fam Pract 1998 47 446 452 9866670 Dean BB Gerner D Gerner RH A systematic review evaluating health-related quality of life, work impairment, and healthcare costs and utilization in bipolar disorder Curr Med Res Opin 2004 20 139 154 15006007 Namjoshi MA Buesching DP A review of the health-related quality of life literature in bipolar disorder Qual Life Res 2001 10 105 115 11642680 10.1023/A:1016662018075 Evenson RC Vieweg BW Using a quality of life measure to investigate outcome in outpatient treatment of severely impaired psychiatric clients Compr Psychiatry 1998 39 57 62 9515189 10.1016/S0010-440X(98)90079-7 Bishop SL Walling DP Dott SG Folkes CC Bucy J Refining quality of life: validating a multidimensional factor measure in the severe mentally ill Qual Life Res 1999 8 151 160 10457748 10.1023/A:1026489331009 Corrigan PW Faber D Rashid F Leary M The construct validity of empowerment among consumers of mental health services Schizophr Res 1999 38 77 84 10427613 10.1016/S0920-9964(98)00180-7 O'Donnell M Parker G Proberts M Matthews R Fisher D Johnson B Hadzi-Pavlovic D A study of client-focused case management and consumer advocacy: the Community and Consumer Service Project Aust N Z J Psychiatry 1999 33 684 693 10544992 10.1046/j.1440-1614.1999.00629.x Patterson TL Kaplan RM Grant I Semple SJ Moscona S Koch WL Harris MJ Jeste DV Quality of well-being in late-life psychosis Psychiatry Res 1996 63 169 181 8878313 10.1016/0165-1781(96)02797-7 Drake RE McHugo GJ Clark RE Teague GB Xie H Miles K Ackerson TH Assertive community treatment for patients with co-occurring severe mental illness and substance use disorder: a clinical trial Am J Orthopsychiatry 1998 68 201 215 9589759 Bonicatto SC Dew MA Zaratiegui R Lorenzo L Pecina P Adult outpatients with depression: worse quality of life than in other chronic medical diseases in Argentina Soc Sci Med 2001 52 911 919 11234864 10.1016/S0277-9536(00)00192-1 Yanos PT Rosenfield S Horwitz AV Negative and supportive social interactions and quality of life among persons diagnosed with severe mental illness Community Ment Health J 2001 37 405 419 11419518 10.1023/A:1017528029127 Kuehner C Subjective quality of life: validity issues with depressed patients Acta Psychiatr Scand 2002 106 62 70 12100349 10.1034/j.1600-0447.2002.02204.x Welham J Haire M Mercer D Stedman T A gap approach to exploring quality of life in mental health Qual Life Res 2001 10 421 429 11763204 10.1023/A:1012549622363 Hirschfeld RM Allen MH McEvoy JP Keck PE JrRussell JM Safety and tolerability of oral loading divalproex sodium in acutely manic bipolar patients J Clin Psychiatry 1999 60 815 818 10665626 Dazord A Astolfl F Guisti P Rebetez MC Mino A Terra JL Brochier C Quality of life assessment in psychiatry: the Subjective Quality of Life Profile (SQLP)--first results of a new instrument Community Ment Health J 1998 34 525 535 9793742 10.1023/A:1018746614552 Atkinson MJ Caldwell L The differential effects of mood on patients' ratings of life quality and satisfaction with their care J Affect Disord 1997 44 169 175 9241577 10.1016/S0165-0327(97)00041-4 Thunedborg K Black CH Bech P Beyond the Hamilton depression scores in long-term treatment of manic-melancholic patients: prediction of recurrence of depression by quality of life measurements Psychother Psychosom 1995 64 131 140 8657843 Berlim MT Mattevi BS Pavanello DP Caldieraro MA Fleck MP Suicidal ideation and quality of life among adult Brazilian outpatients with depressive disorders J Nerv Ment Dis 2003 191 193 197 12637847 10.1097/00005053-200303000-00010 Cooke RG Robb JC Young LT Joffe RT Well-being and functioning in patients with bipolar disorder assessed using the MOS 20-ITEM short form (SF-20) J Affect Disord 1996 39 93 97 8827417 10.1016/0165-0327(96)00016-X Robb JC Cooke RG Devins GM Young LT Joffe RT Quality of life and lifestyle disruption in euthymic bipolar disorder J Psychiatr Res 1997 31 509 517 9368193 10.1016/S0022-3956(97)00030-7 Robb JC Young LT Cooke RG Joffe RT Gender differences in patients with bipolar disorder influence outcome in the medical outcomes survey (SF-20) subscale scores J Affect Disord 1998 49 189 193 9629948 10.1016/S0165-0327(98)00003-2 MacQueen GM Young LT Robb JC Cooke RG Joffe RT Levels of functioning and well-being in recovered psychotic versus nonpsychotic mania J Affect Disord 1997 46 69 72 9387088 10.1016/S0165-0327(97)00083-9 Kusznir A Cooke RG Young LT The correlates of community functioning in patients with bipolar disorder J Affect Disord 2000 61 81 85 11099744 10.1016/S0165-0327(99)00193-7 MacQueen GM Young LT Robb JC Marriott M Cooke RG Joffe RT Effect of number of episodes on wellbeing and functioning of patients with bipolar disorder Acta Psychiatr Scand 2000 101 374 381 10823297 10.1034/j.1600-0447.2000.101005374.x Stewart AL Hays RD Ware JE Jr The MOS short-form general health survey. Reliability and validity in a patient population Med Care 1988 26 724 735 3393032 Wells KB Stewart A Hays RD Burnam MA Rogers W Daniels M Berry S Greenfield S Ware J The functioning and well-being of depressed patients. Results from the Medical Outcomes Study JAMA 1989 262 914 919 2754791 10.1001/jama.262.7.914 Flanagan JC A research approach to improving our quality of life American Psychology 1978 33 138 147 10.1037//0003-066X.33.2.138 Devins GM Edworthy SM Seland TP Klein GM Paul LC Mandin H Differences in illness intrusiveness across rheumatoid arthritis, end-stage renal disease, and multiple sclerosis J Nerv Ment Dis 1993 181 377 381 8501459 Vojta C Kinosian B Glick H Altshuler L Bauer MS Self-reported quality of life across mood states in bipolar disorder Compr Psychiatry 2001 42 190 195 11349236 10.1053/comp.2001.23143 Yatham LN Lecrubier Y Fieve RR Davis KH Harris SD Krishnan AA Quality of life in patients with bipolar I depression: data from 920 patients Bipolar Disord 2004 6 379 385 15383130 10.1111/j.1399-5618.2004.00134.x Ozer S Ulusahin A Batur S Kabakci E Saka MC Outcome measures of interepisode bipolar patients in a Turkish sample Social Psychiatry and Psychiatric Epidemiology 2002 37 31 37 11924747 10.1007/s127-002-8211-z Endicott J Nee J Harrison W Blumenthal R Quality of Life Enjoyment and Satisfaction Questionnaire: a new measure Psychopharmacol Bull 1993 29 321 326 8290681 Rapaport MH Clary CM Judd LL The impact of depression and its treatment. Presented at the 154th annual meeting of the American Psychiatric Association, New Orleans, La 2001 Michalak EE Tam EM Manjunath CV Solomons K Levitt AJ Levitan R Enns M Morehouse R Yatham LN Lam RW Generic and health-related quality of life in patients with seasonal and nonseasonal depression Psychiatry Res 2004 128 245 251 15541781 10.1016/j.psychres.2004.01.012 Miller IW Keitner GI Schatzberg AF Klein DN Thase ME Rush AJ Markowitz JC Schlager DS Kornstein SG Davis SM Harrison WM Keller MB The treatment of chronic depression, part 3: psychosocial functioning before and after treatment with sertraline or imipramine J Clin Psychiatry 1998 59 608 619 9862607 Ruggeri M Gater R Bisoffi G Barbui C Tansella M Determinants of subjective quality of life in patients attending community-based mental health services. The South-Verona Outcome Project 5 Acta Psychiatr Scand 2002 105 131 140 11939963 10.1034/j.1600-0447.2002.01293.x Olusina AK Ohaeri JU Subjective quality of life of recently discharged Nigerian psychiatric patients Soc Psychiatry Psychiatr Epidemiol 2003 38 707 714 14689175 10.1007/s00127-003-0691-7 Perlis RH Miyahara S Marangell LB Wisniewski SR Ostacher M DelBello MP Bowden CL Sachs GS Nierenberg AA (STEP-BD) Long-Term implications of early onset in bipolar disorder: data from the first 1000 participants in the systematic treatment enhancement program for bipolar disorder Biol Psychiatry 2004 55 875 881 15110730 10.1016/j.biopsych.2004.01.022 Stewart AL Hays RD Ware JE Jr The MOS short-form general health survey. Reliability and validity in a patient population. Med Care 1998 26 724 735 3393032 Garratt A Schmidt L Mackintosh A Fitzpatrick R Quality of life measurement: bibliographic study of patient assessed health outcome measures BMJ 2002 324 1417 12065262 10.1136/bmj.324.7351.1417 Arnold LM Witzeman KA Swank ML McElroy SL Keck PE Jr Health-related quality of life using the SF-36 in patients with bipolar disorder compared with patients with chronic back pain and the general population J Affect Disord 2000 57 235 239 10708837 10.1016/S0165-0327(99)00042-7 McHorney CA Kosinski M Ware JE Jr Comparisons of the costs and quality of norms for the SF-36 health survey collected by mail versus telephone interview: results from a national survey Med Care 1994 32 551 567 8189774 ten Have M Vollebergh W Bijl R Nolen WA Bipolar disorder in the general population in The Netherlands (prevalence, consequences and care utilisation): results from The Netherlands Mental Health Survey and Incidence Study (NEMESIS) J Affect Disord 2002 68 203 213 12063148 10.1016/S0165-0327(00)00310-4 Hakkaart-van Roijen L Hoeijenbos MB Regeer EJ ten Have M Nolen WA Veraart CP Rutten FF The societal costs and quality of life of patients suffering from bipolar disorder in the Netherlands Acta Psychiatr Scand 2004 110 383 392 15458562 10.1111/j.1600-0447.2004.00403.x Atkinson M Zibin S Chuang H Characterizing quality of life among patients with chronic mental illness: a critical examination of the self-report methodology Am J Psychiatry 1997 154 99 105 8988966 Ferrans CE Powers MJ Psychometric assessment of the Quality of Life Index Res Nurs Health 1992 15 29 38 1579648 Chand PK Mattoo SK Sharan P Quality of life and its correlates in patients with bipolar disorder stabilized on lithium prophylaxis Psychiatry Clin Neurosci 2004 58 311 318 15149299 10.1111/j.1440-1819.2004.01237.x Torrance GW Measurement of health state utilities for economic appraisal J Health Econ 1986 5 1 30 10311607 10.1016/0167-6296(86)90020-2 Wells KB Sherbourne CD Functioning and utility for current health of patients with depression or chronic medical conditions in managed, primary care practices Arch Gen Psychiatry 1999 56 897 904 10530631 10.1001/archpsyc.56.10.897 Tsevat J Keck PE Hornung RW McElroy SL Health values of patients with bipolar disorder Qual Life Res 2000 9 579 586 11190012 10.1023/A:1008979825704 Russo J Roy-Byrne P Reeder D Alexander M Dwyer-O'Connor E Dagadakis C Ries R Patrick PD Longitudinal assessment of quality of life in acute psychiatric inpatients: reliability and validity J Nerv Ment Dis 1997 185 166 175 9091598 10.1097/00005053-199703000-00006 Lehman AF A quality of life interview for the chronically mentally ill Evaluation and Program Planning 1988 51 62 10.1016/0149-7189(88)90033-X Leidy NK Palmer C Murray M Robb J Revicki DA Health-related quality of life assessment in euthymic and depressed patients with bipolar disorder. Psychometric performance of four self-report measures J Affect Disord 1998 48 207 214 9543211 10.1016/S0165-0327(97)00147-X Salyers MP Bosworth HB Swanson JW Lamb-Pagone J Osher FC Reliability and validity of the SF-12 health survey among people with severe mental illness Med Care 2000 38 1141 1150 11078054 10.1097/00005650-200011000-00008 Ware J JrKosinski M Keller SD A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity Med Care 1996 34 220 233 8628042 10.1097/00005650-199603000-00003 Ritsner M Modai I Endicott J Rivkin O Nechamkin Y Barak P Goldin V Ponizovsky A Differences in quality of life domains and psychopathologic and psychosocial factors in psychiatric patients J Clin Psychiatry 2000 61 880 889 11105747 Ritsner M Kurs R Kostizky H Ponizovsky A Modai I Subjective quality of life in severely mentally ill patients: a comparison of two instruments Qual Life Res 2002 11 553 561 12206576 10.1023/A:1016323009671 Revicki DA Tohen M Gyulai L Thompson C Pike S Davis-Vogel A Zarate C Telephone versus in-person clinical and health status assessment interviews in patients with bipolar disorder Harv Rev Psychiatry 1997 5 75 81 9385024 Namjoshi MA Rajamannar G Jacobs T Sanger TM Risser R Tohen MF Breier A Keck PE Jr Economic, clinical, and quality-of-life outcomes associated with olanzapine treatment in mania. Results from a randomized controlled J Affect Disord 2002 69 109 118 12103458 10.1016/S0165-0327(01)00310-X Shi L Namjoshi MA Zhang F Gandhi G Edgell ET Tohen M Breier A Haro JM Olanzapine versus haloperidol in the treatment of acute mania: clinical outcomes, health-related quality of life and work status Int Clin Psychopharmacol 2002 17 227 237 12177585 10.1097/00004850-200209000-00003 Tohen M Goldberg JF Gonzalez-Pinto Arrillaga AM Azorin JM Vieta E Hardy-Bayle MC Lawson WB Emsley RA Zhang F Baker RW Risser RC Namjoshi MA Evans AR Breier A A 12-week, double-blind comparison of olanzapine vs haloperidol in the treatment of acute mania Arch Gen Psychiatry 2003 60 1218 1226 14662554 10.1001/archpsyc.60.12.1218 Namjoshi MA Risser R Shi L Tohen M Breier A Quality of life assessment in patients with bipolar disorder treated with olanzapine added to lithium or valproic acid J Affect Disord 2004 81 223 229 15337326 10.1016/j.jad.2003.07.005 Shi L Namjoshi MA Swindle R Yu X Risser R Baker RW Tohen M Effects of olanzapine alone and olanzapine/fluoxetine combination on health-related quality of life in patients with bipolar depression: secondary analyses of a double-blind, placebo-controlled, randomized clinical trial Clin Ther 2004 26 125 134 14996525 10.1016/S0149-2918(04)90013-6 Revicki DA Paramore LC Sommerville KW Swann AC Zajecka JM Divalproex sodium versus olanzapine in the treatment of acute mania in bipolar disorder: health-related quality of life and medical cost outcomes J Clin Psychiatry 2003 64 288 294 12716270 Patelis-Siotis I Young LT Robb JC Marriott M Bieling PJ Cox LC Joffe RT Group cognitive behavioral therapy for bipolar disorder: a feasibility and effectiveness study J Affect Disord 2001 65 145 153 11356238 10.1016/S0165-0327(00)00277-9 Bond GR Resnick SG Drake RE Xie H McHugo GJ Bebout RR Does competitive employment improve nonvocational outcomes for people with severe mental illness? J Consult Clin Psychol 2001 69 489 501 11495178 10.1037/0022-006X.69.3.489 Dogan S Sabanciogullari S The effects of patient education in lithium therapy on quality of life and compliance Arch Psychiatr Nurs 2003 17 270 275 14685951 10.1053/j.apnu.2003.10.001 Tohen M Hennen J Zarate CM JrBaldessarini RJ Strakowski SM Stoll AL Faedda GL Suppes T Gebre-Medhin P Cohen BM Two-year syndromal and functional recovery in 219 cases of first-episode major affective disorder with psychotic features Am J Psychiatry 2000 157 220 228 10671390 10.1176/appi.ajp.157.2.220 Huxley NA Parikh SV Baldessarini RJ Effectiveness of psychosocial treatments in bipolar disorder: state of the evidence Harv Rev Psychiatry 2000 8 126 140 10973937 10.1093/hrp/8.3.126 Altman E Rating scales for mania: is self-rating reliable? J Affect Disord 1998 50 283 286 9858088 10.1016/S0165-0327(98)00018-4 Wilson KA Dowling AJ Tannock IF Perception of quality of life by patients, partners and treating physicians Qual Life Res 2000 9 1041 1052 11332225 10.1023/A:1016647407161 Simon GE Revicki DA Grothaus L VonKorff M SF-36 summary scores: are physical and mental health truly distinct? Med Care 1998 36 567 572 9544596 10.1097/00005650-199804000-00012 Michalak EE Yatham LN Kolesar S Lam RW Biolar disorder and quality of life: A patient-centered perspective Qual Life Res 2005 9544596 10.1097/00005650-199804000-00012	16288650	PMC1325049	CC BY	2021-01-04 16:23:37	no	Health Qual Life Outcomes. 2005 Nov 15; 3:72	utf-8	Health Qual Life Outcomes	2,005	10.1186/1477-7525-3-72	oa_comm
==== Front Theor Biol Med ModelTheoretical Biology & Medical Modelling1742-4682BioMed Central London 1742-4682-2-451628393910.1186/1742-4682-2-45ResearchSupply-demand balance in outward-directed networks and Kleiber's law Painter Page R [email protected] Office of Environmental Health Hazard Assessment, California Environmental Protection, Agency, P.O. Box 4010, Sacramento CA 95812, USA2005 10 11 2005 2 45 45 3 5 2005 10 11 2005 Copyright © 2005 Painter; licensee BioMed Central Ltd.2005Painter; 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 theories have attempted to derive the value of the exponent α in the allometric formula for scaling of basal metabolic rate from the properties of distribution network models for arteries and capillaries. It has recently been stated that a basic theorem relating the sum of nutrient currents to the specific nutrient uptake rate, together with a relationship claimed to be required in order to match nutrient supply to nutrient demand in 3-dimensional outward-directed networks, leads to Kleiber's law (b = 3/4). Methods The validity of the supply-demand matching principle and the assumptions required to prove the basic theorem are assessed. The supply-demand principle is evaluated by examining the supply term and the demand term in outward-directed lattice models of nutrient and water distribution systems and by applying the principle to fractal-like models of mammalian arterial systems. Results Application of the supply-demand principle to bifurcating fractal-like networks that are outward-directed does not predict 3/4-power scaling, and evaluation of water distribution system models shows that the matching principle does not match supply to demand in such systems. Furthermore, proof of the basic theorem is shown to require that the covariance of nutrient uptake and current path length is 0, an assumption unlikely to be true in mammalian arterial systems. Conclusion The supply-demand matching principle does not lead to a satisfactory explanation for the approximately 3/4-power scaling of mammalian basal metabolic rate. nutrient supply networksallometric scalingmetabolism ==== Body Introduction Regression analyses of measurements of a physiological or structural variable R (e.g. cardiac output or pulmonary alveolar surface area) in mammals of different mass M have shown in many cases that the variable is closely approximated by a function of the form R = R1Mb, which is often termed an allometric relationship [1,2]. A prominent example is Kleiber's law for scaling the basal metabolic rate, B, in mammals [3,4], B = B1M3/4, which is equivalent to scaling the specific basal metabolic rate, B/M, proportionally to M-1/4. The search for a theory to explain Kleiber's law has recently focused on the nutrient distribution network formed by arteries and capillaries. Banavar et al. [5-7] argue that the law follows from basic properties of an outward-directed network (ODN). In the initial description of an ODN [5,6], Banavar, Maritan and Rinaldo (BMR) assume that a network consists of sites for nutrient uptake that are connected to a single source (e.g. the heart). An uptake site is located at each network branching point and at each terminal network point. Network distance Ly along a path from the nutrient source O to a site Y is defined as the number of uptake sites on the path. The rate of uptake of nutrient at site Y is denoted By. A network segment that goes from a site X to an adjacent site Y is termed the link XY, and the rate at which nutrient enters the link is termed the current and is denoted Ixy. For a link that carries nutrient current from a site X to a site Y, the level of the link XY and the level of the site Y is defined as the network distance Ly to the site Y. In an ODN, direction of flow is away from O on each link. The authors denote the sum of currents on all links ΣIxy, termed total network current, by F, which is shown to be defined by the equation F = ΣByLy. (1) The initial ODN theory is completed by the introduction of the relation: F = nE(By)E(Ly), (2) where n is the number of uptake sites and E(By) and E(Ly) denote average values. In the first attempt to derive the law using Relation (2), total network current is assumed to be proportional to blood volume in the mammalian systemic arterial and capillary system, and this blood volume is assumed to be proportional to body mass. With the additional assumption that E(Ly) ∝ Lp, where Lp is the linear dimension of the region supplied by the network, it follows that total uptake rate, B, scales as and that blood volume and body mass scale as . Consequently, B scales as M3/4. However, body tissue density, M/V, is predicted in this model to scale as Lp= V1/3 [8,9]. If density were to scale as Lp, the density of hippopotamus tissue would be more than ten times the density of mouse tissue and would far exceed the density of granite. An additional problem in the BMR theory is that total network current, an abstract property of the arterial system, is not necessarily proportional to blood volume [10]. Furthermore, Relation (2) is not true in examples of ODNs where uptake occurs only at terminal sites [10]. In a second attempt to derive Kleiber's law using the concept of total network current, Banavar, Damuth, Maritan and Rinaldo (BDMR) add the assumption that networks are embedded in spatial regions "such that mass and volume scale isometrically" [7]. Cubic and square regions are examples of such isometric bodies. They also assume that body mass scales as body volume, , where D is the dimension of the region representing the body. Citing the previous attempt to derive Kleiber's law, they write F ∝ (Lp/u)B, (3) and claim that "Eq. 3 has been proven as a mathematical theorem" (u is the average physical distance between connected uptake sites). Next, they define the function r1= F/S, (4) where S denotes the system's size (measured as area for a 2-dimensional system and volume for a 3-dimensional system). They define the "service volume" by the relationship , and they consider the scaling of r2= ls/u. Clearly, r2, which is described as the rate "with which the metabolites are taken in at the level of the tissue," is defined by the relation r2∝ (B/S)-1/D/u. (5) Next, BDMR state: "Maintaining a match between these two rates across body size would require that both rates scale with body mass in the same manner, i.e., if r1 ∝ and r2 ∝ , then s1 = s2. If this were not true, under changes of body mass either the supply of the metabolite would exceed the demand or vice versa." Based on this reasoning, they assert their supply-demand matching principle: r1∝ r2. (6) Combining Relations (3) and (6) leads to Kleiber's law. BDMR state that their "conclusions are based on general arguments incorporating the minimum of biological detail and should therefore apply to the widest range of organisms" [7]. They support their theory with two lattice models of isometric ODNs, a 2 × 2 lattice and a 3 × 3 lattice. In their examples, the uptake rates are identical and the physical link lengths are identical throughout a lattice. Lattices with these properties are termed simple lattice networks. Figure 1a illustrates a simple 3 × 3 lattice ODN. Figure 1b illustrates an 8 × 8 simple lattice ODN with four embedded 3 × 3 lattice ODNs. The lattice in Figure 1a and the embedded lattices in Figure 1b are formally equivalent to the example provided by BDMR in their Figure 1b[7]. BDMR do not test their model using ODNs that are not square simple lattices. Figure 1 Simple lattice ODN models. a. A 3 × 3 ODN where current or nutrient is supplied by the link from the origin (large open circle) to the lower left corner. Uptake sites are denoted by closed circles. b. An 8 × 8 ODN where current or nutrient is supplied by the link from the origin (large open circle) to the lower left corner. Uptake sites are denoted by closed circles. Note that there are four identical embedded 3 × 3 outward-directed networks (e.g. the network in the upper right corner) within the 8 × 8 network. Results While BDMR repeatedly state that they proved Relation (3) in their original publication on ODNs [7,11], they could not have proved this result. This can be demonstrated by considering ODNs in non-isometric solid bodies. (The networks considered in their original publication were not assumed to be isometric.) Consider two lattice ODNs that have identical spacing u between adjacent uptake sites and identical uptake rate By at each uptake site. These two networks differ in their total network current (denoted, respectively, by F1 and F2), in their linear dimension (denoted Lp1 and Lp2) and in their number of uptake sites (denoted n1 and n2). If Relation (3) is correct, we can write F1/F2= (Lp1n1)/(Lp2n2). However, this equation is a false statement whenever Lp1/Lp2 is an irrational number because both F1/F2 and n1/n2 are rational numbers. For example, when network 1 is a 2 × 3 × 4 lattice and network 2 is a 3 × 4 × 5 lattice, Lp1/Lp2 is (5/2)1/3 and the above equation must be false. For the isometric networks considered by BDMR, the ratio Lp1/Lp2 is a rational number, and Relation (3) is correct for some, but not all, families of isometric ODNs. In the remainder of this section, the theory of BDMR is evaluated in three ways. The first is an evaluation of its predictions for an ODN that is not a simple lattice. The second is an evaluation of whether Relation (6) is correct for simple outward-directed current network models, and the third is the identification of mathematical conditions required for the validity of the critical mathematical relationships, Relations (2), (3) and (6). If the logic used by BDMR to derive Relation (3) is correct for all ODNs that supply isometric regions, the assumption of Relation (6) should lead to the conclusion that Kleiber's law holds for outward-directed models of the arterial system that differ from the lattice models presented by BDMR. To see if this is correct, we apply this assumption to the well-known outward-branching "fractal-like" model studied by West et al. [12]. Figure 2 and Figure 3 illustrate how an outward-bifurcating network can be folded inside a square or cube with side or edge length equal to 2ilt, where i is a positive integer and lt is the linear dimension of the region supplied by terminal uptake sites. The supply network for a square starts with an H-shaped network of linear dimension Lp/2 that is connected to the nutrient source (Figure 2a). The network is extended by iteratively connecting each terminal site to an H-shaped structure that is one-half the size (in terms of linear dimension) of the structures added in the previous step (Figure 2b). For a network that supplies a cube, we start with two parallel H-shaped structures of linear dimension Lp/2 that are connected by a conduit of length Lp/2. This structure, termed an H-H structure, is illustrated in Figure 3a. This network is extended by iterative additions of H-H structures of one-half the dimension of the previously added H-H structure (Figure 3b). Each added structure is connected at its midpoint to a terminus. Iterative addition of smaller and smaller H-shaped structures gives the fractal lung model of Mandelbrot [13], and iterative addition of H-H structures gives a 3-dimensional fractal model. An infinite sequence of additions gives an area-filling network of fractal dimension 2 for the 2-dimensional network and a space-filling network of fractal dimension 3 for the 3-dimensional network. These networks have the topological structure of a Cayley tree. Consequently, the claim [6] that Cayley-tree networks are not plausible models of the mammalian arterial network because a Cayley tree "for large enough size, cannot exist in any finite-dimensional space" is incorrect. Figure 2 A 2-dimensional fractal-like, branching network model for an arterial tree. Blood enters the network through the structure represented as a thick horizontal line. Terminal arteries are represented by thin horizontal lines. a. A network that uniformly supplies a 2 × 2 area where the unit distance is the spacing between adjacent termini of small arteries. b. A network that uniformly supplies a 4 × 4 area. Figure 3 A 3-dimensional fractal-like, branching network model for an arterial tree. Blood enters the network through the structure represented as a thick horizontal line. Terminal arteries are represented by thin horizontal lines. a. A network that uniformly supplies a 2 × 2 × 2 volume where the unit distance is the spacing between adjacent termini of small arteries. b. A network that uniformly supplies a 4 × 4 × 4 volume. Because these outward-bifurcating networks are folded inside a square or a cube, their scaling behavior can be directly compared with the scaling behavior of isometric lattices embedded in regions of identical shape and size. The networks shown in Figure 2 and Figure 3 start with a single link and bifurcate at each branch point until a terminal uptake site is reached at path length (number of links) k. The uptake rates at terminal sites and branch points are denoted Ba and Bb, respectively. The number of terminal uptake sites 2k-1 is equal to (Lp/ lt)D, where Lp is the length of the side or edge. Total network uptake B is Ba2k-1+ Bb[2k-1 - 1], and total network current F is Bak2k-1+ Bb[(k-2)2k-1 - 1]. First, we assume that Bb is negligible compared to Ba. The biological justification for this simplification is that Bb represents nutrient uptake by endothelial cells in arteries and by smooth muscle cells in small arteries and arterioles, and this uptake may be very small compared to the nutrient uptake from capillaries represented by Ba. Application of the scaling assumption in Relation [6] to the formulas for B and F in this example gives the relation (B/S)-1/D/u ∝ (B/S)k, which is equivalent to B/S ∝ (uk)-D/(D+1). This expression would be equivalent to Kleiber's law if k scaled as S1/D. However, for the networks in Figure 2 and Figure 3, the length of a link between neighboring sites is a constant (denoted u), and k scales as Dln(S1/D/u)/ln(2)+1. If it is assumed that Bb = Ba, the approximations 2k-1 ≈ 2k-1 - 1 and (k-2)2k-1 ≈ (k-2)2k-1 - 1 lead to the relationship B/S ∝ [u(k-1)]-D/(D+1, which is again very different from Kleiber's law. While the above example shows that Kleiber's law cannot be derived from general properties of ODNs using Relation (6), the possibility remains that the derivation of BDMR is correct for outward-directed lattices and that the arterial system is more accurately modeled as a simple lattice ODN (where Relation (3) is true) than as a Cayley tree. If this is true, the validity of a claim that Kleiber's law is correct for lattice-like arterial supply-demand models depends on the validity of the supply-demand matching principle in Relation (6). To see if this principle is correct in general for lattice ODNs that supply metabolites, we consider an example where a lattice network of pipes supplies liquid nutrient to nearly identical mature animals (e.g. inbred adult laboratory rats). A single animal is located in a cage at each vertex and at each terminal site. The length of a link connecting neighboring sites is a constant (denoted u), and each animal takes up nutrient through a valve that provides liquid to the animal only when it sucks and swallows all the liquid provided. Uptake by a caged animal is measured as the amount of nutrient or water ingested at the site per day and is denoted By. This model is analyzed because the overall uptake rate is determined by demand, as is nutrient uptake in the "Allometric Cascade" model for basal metabolic rate scaling [14,15]. In such a biological example, supply is exactly matched to demand, and the logic used by BDMR to justify Relation (6), if correct, should predict the scaling of the system in the following cases: In case 1, the number of uptake sites, n, of the lattice is increased while u and By remain constant. In case 2, u is increased while n and By remain constant. In case 3, By increases while n and u remain constant. Case 3 can be achieved by replacing adult animals in a nutrient-supply lattice with young growing animals that increase their uptake rate as they grow. These three cases are easily translated into equivalent examples where the lattice supplies electrical power to residences located at each lattice junction. The scaling behavior of r1 and r2 in these three examples is listed in Table 1 along with the scaling relations for each of these supply-demand lattices. In each case, the network maintains a match between supply and demand. In none of the three cases does the network do this by "maintaining a match" between the rates r1 and r2 "across body size." Furthermore, none of these cases has the scaling of Kleiber's law. In case 1 and case 2, where the size of the system is increased, r2 is clearly an intensive property of the system while r1 depends on system size. In case 3, supply is matched to demand by balancing an increase in r1 with a decrease in r2. Total network current and r1 increase directly in proportion to By. On the other hand, r2 decreases in proportion to . Clearly, r2 is not the rate of "the demand for delivered metabolites" which increases in proportion to By, nor are the units of r2 "inverse time units" as claimed by BDMR. Table 1 Scaling of r1 and r2 in three cases of parameter variation in supply-demand lattice ODNs with uptake determined by demand. Parameter variation Scaling of: Scaling of B r1 r2 n ↑ ∝ S1/D ∝ S0 ∝ S u ↑ ∝ S-1 ∝ S0 ∝ S0 By ↑ ∝ By ∝ By-1/D ∝ By Another peculiarity of the ODN theory becomes apparent when it is applied to ODN lattices embedded in a larger ODN lattice. Figure 1b illustrates four 3 × 3 lattices supplied at a corner and embedded in an 8 × 8 lattice supplied at a corner. If Relationship (6) is correct for the ODNs in the figure, then the metabolic rate for each 3 × 3 lattice, denoted B3, should be a(32)2/3, where a denotes the constant of proportionality. Similarly, the metabolic rate of the 8 × 8 lattice, denoted B8, is a(82)2/3. From conservation of energy, it is clear that 4B3 must be less than or equal to B8, i. e., 4(32)2/3 must be less than or equal to (82)2/3. However, 4(32)2/3 is an irrational number between 17 and 18, while (82)2/3 is equal to 16. A similar argument applied to eight 4 × 4 × 4 cubic lattices embedded in a 10 × 10 × 10 cubic lattice leads to a contradiction if it is assumed that the same 3/4-power scaling relationship applies to the entire lattice and to the embedded lattices. This argument can be generalized to show that an allometric scaling law, B = B1Sα for isometric lattices with invariant u cannot have an exponent less than 1. The above network examples show that the relations derived by BDMR are not true for all ODNs with supply-demand balance. We now identify assumptions that are not stated by BDMR but that guarantee the validity of these relations. First we define the conditions required for the validity of Relation (2). To do this, we apply the following basic theorem from statistical theory: For random variables X and Y, the well-known formula for E(XY), the average value of the product of random variables, is E(XY) = E(X)E(Y) + Covariance(X,Y). Application of this theorem to Relation (1) gives F = nE(By)E(Ly)+ nCovariance(By, Ly) which simplifies to F = E(Ly)B + nCovariance(By, Ly) Therefore, Relation (2) is correct for an ODN if and only if Covariance(By, Ly) is 0, and the covariance is 0 if By and Ly are independent. If By is invariant, independence is assured. Now assume that Relation (2) is true for an ODN. We denote the physical length of a network link from site X to site Y and carrying current toward site Y by uxy. Next, define a path to a site Y as a sequence of connected links carrying outward-directed current from the source to site Y. The physical length of this path is the sum of the lengths of the links that form the path. To derive Relation (3), we assume that all paths to a site Y have the same path length (denoted dy) and that the length of all links carrying current to site Y is the same (denoted uy). These assumptions are true for simple lattice ODNs and for fractal-like ODNs. The number of paths that pass through or terminate at site Y is denoted by νy. We define the average path length as E(dy) and assume that E(dy) is proportional to Lp. In the sum that defines the numerator of E(dy), the sum of the values of uy is νyuy. Therefore, E(dx) = (Σνyuy)/n which is equivalent to E(dy) = E(νy)E(uy) + Covariance(νy, uy) In computing E(νy), we note that the sum of the values of νy for all level 1 sites is equal to the number of level 1 links on all paths. In general, the sum of the values of νy for all level j sites is the number of level j links on all paths. Therefore, E(νy) is the sum of all links on all paths divided by the number of paths, i.e. E(νy) = E(Ly). Consequently, E(dy)/u = E(Ly) + Covariance(νy, uy)/u which shows that, when Lp∝ E(dy) and the two additional assumptions on physical path length are true, Relation (3) follows from Relation (2) if and only if Covariance(νy, uy)/u is 0 or is proportional to E(Lp). In isometric lattice models with constant spacing between uptake sites, this covariance is 0, and E(dy)/u and E(Ly) are proportional to E(Ly). However, in the models of Figure 2 and Figure 3, this covariance is not 0 because both νy and uy decrease from level 1 links to level k links. Furthermore, for these bifurcating ODNs, E(Ly) is approximately proportional to the logarithm of E(Ly) [8]. Consequently, Relation (3) is not true for these ODNs. Finally, we show that when Relation (3) is true for an isometric 3-dimensional ODN, assuming that Relation (6) is true is equivalent to assuming that Kleiber's law is true. If total current in a family of networks is described by Relation (3), if mass and volume scale proportionally to and if metabolic rate is described by Kleiber's law, then B-4/3∝ M-1, and B-1/3∝ B/M. Multiplying both sides by Lp/u gives , and substituting F for (Lp/u)B gives Relation (6). The steps of this argument can be reversed to show that if Relation (3) and Relation (6) are true, then Kleiber's law is true. Therefore, for isometric networks where Relation (3) is true, the supply-demand principle in Relation (6) and Kleiber's law are equivalent statements. Discussion and conclusion The incorrect prediction of the BMR model that body tissue density scales as Lp is not a prediction of the BDMR model, which contains the assumption that body mass scales as body volume. However, the related current model of Dreyer and Puzio does predict that the mass of blood in a body scales as Lp [16,17]. One issue in evaluating the model of BDMR is the validity of Relation (2) and Relation (3). BDMR state that they proved these relations as theorems [7,11]. However, a counterexample to their "theorem" of Relation (2) has been published [10], and the above results show that when uptake rates are not independent of path length, there is no reason to believe that Relation (2) or Relation (3) is true. A second issue in evaluating the model of BDMR is whether a network of cubic lattices or the bifurcating ODN model of West et al. [12] more closely resembles the mammalian system of arteries and capillaries. This is a critical question because the basic assumptions of BDMR, Relation (2) and Relation (3), are true for simple lattices but not for the bifurcating ODNs of Figure 2 and Figure 3. The arterial system is clearly more similar to the West et al. model than to a simple lattice [18]. The principle claimed to be required to match supply to demand in ODNs is not correct for plausible conceptual models where supply must be matched to demand, and it does not lead to Kleiber's law for "fractal-like" ODNs. Therefore, the supply-demand matching principle does not lead to a satisfactory explanation for the approximately 3/4-power scaling of mammalian basal metabolic rate. The supply-demand principle of BDMR has also been investigated by Makarieva et al. [19]. They, too, conclude that r2 is not the rate of "the demand for delivered metabolites" which increases in proportion to By, nor are the units of r2 "inverse time units" as claimed by BDMR. A final issue in the evaluation of the model of BDMR and other models that predict 3/4-power scaling of the basal metabolic rate is that experimental support for Kleiber's Law is rapidly eroding. As reviewed by Heusner [19] and Dodds et al. [8], the slope of the allometric scaling expression is less than 3/4. Furthermore, these investigators showed that the slope for mammals weighing less than 10 kg is approximately 2/3 while the slope for mammals weighing more than 10 kg is approximately 3/4. More recently, White and Seymour [21] showed that, following a correction for the effect of body temperature on metabolic rate, the slope is 0.67 for a very large collection of data (619 mammalian species). Statistical analysis of these data yields a slope that is less than 2/3 for animals smaller than 1 kg and a slope greater than 3/4 for animals larger than 50 kg [22]. The erosion of support for Kleiber's law should not result in a loss of interest in explanations for the scaling of metabolic rate. To the contrary, large collections of metabolic data that exhibit upward curvature support models based on physiological and anatomical considerations [14,15,22] but do not support Kleiber's law. Such models may focus attention on relationships at the heart of metabolic scaling issues, the physiological relationships between tissue blood flow and tissue metabolic rate. ==== Refs Schmidt-Nielsen K Scaling: Why is Animal Size so Important? 1984 New York: Cambridge University Press Calder WA III Size, Function and Life History 1984 Cambridge. MA: Harvard University Press Kleiber M Body size and metabolism Hilgardia 1932 6 315 353 Kleiber M Body size and metabolic rate Physiol Rev 1947 27 511 541 Banavar JR Maritan A Rinaldo A Size and form in efficient transportation networks Nature 1999 399 130 132 10335841 Banavar JR Maritan A Rinaldo A Scaling. Rivers, blood and transportation networks Nature 2000 408 160 Banavar JR Damuth J Maritan A Rinaldo A Supply-demand balance and metabolic scaling Proc Natl Acad Sci USA 2002 99 10506 10509 12149461 10.1073/pnas.162216899 Dodds PS Rothman DH Weitz JS Re-examination of the "3/4-law" of metabolism J Theor Biol 2001 209 9 27 11237567 10.1006/jtbi.2000.2238 Makarieva AM Gorshkov VG Li BL. Revising the distributive networks models of West, Brown and Enquist (1997) and Banavar, Maritan and Rinaldo (1999): Metabolic inequity of living tissues provides clues for the observed allometric scaling rules J Theor Biol 2005 237 291 301 16023678 Painter PR Scaling. Rivers, blood and transportation networks Nature 2000 408 159 11089961 10.1038/35041631 Banavar JR Damuth J Maritan A Rinaldo A Comment on "Revising the distributive networks models of West, Brown and Enquist (1997) and Banavar, Maritan and Rinaldo (1999): Metabolic inequity of living tissues provides clues for the observed allometric scaling rules" by Makarieva, Gorshkov and Li J Theor Biol 2005 16197962 West GB Brown JH Enquist BJ A general model for the origin of allometric scaling laws in biology Science 1997 276 122 126 9082983 10.1126/science.276.5309.122 Mandelbrot BB The Fractal Geometry of Nature 1983 New York: Freeman Darveau C-A Suarez RK Andrews RD Hochachka PW Allometric cascade as a unifying principle of body mass effects on metabolism Nature 2002 147 166 170 12000958 10.1038/417166a Hochachka PW Darveau C-A Andrews RD Suarez RK Allometric cascade: a model for resolving body mass effects on metabolism Comp Biochem Physiol Part A 2003 134 675 691 10.1016/S1095-6433(02)00364-1 Dreyer O Puzio R Allometric scaling in animals and plants J Math Biol 2001 43 144 156 11570589 10.1007/s002850170001 Dreyer O Allometric scaling and central source systems Phys Rev Lett 2001 87 38101 38103 10.1103/PhysRevLett.87.038101 Agur AM Ming JL Grant JC 3 Grant's Atlas of Anatomy 1999 Baltimore: The Williams and Wilkins Co Makarieva AM Gorshkov VG Li BL. Distributive network model of Banavar, Damuth, Maritan and Rinaldo (2002): Critique and perspective J Theor Biol 2005 Heusner AA Size and power in mammals J Exp Biol 1991 160 25 54 1960515 White CR Seymour RS Mammalian basal metabolic rate is proportional to body mass2/3 Proc Natl Acad Sci USA 2003 100 4046 4049 12637681 10.1073/pnas.0436428100 Painter PR Data from necropsy studies and in vitro tissue studies lead to a model for allometric scaling of basal metabolic rate Theor Biol Med Model 2005 2 39 16188039 10.1186/1742-4682-2-39	16283939	PMC1325050	CC BY	2021-01-04 16:24:57	no	Theor Biol Med Model. 2005 Nov 10; 2:45	utf-8	Theor Biol Med Model	2,005	10.1186/1742-4682-2-45	oa_comm
==== Front BMC Med GenetBMC Medical Genetics1471-2350BioMed Central London 1471-2350-6-421633663710.1186/1471-2350-6-42Research ArticleGenetic analysis of the GLUT10 glucose transporter (SLC2A10) polymorphisms in Caucasian American type 2 diabetes Bento Jennifer L [email protected] Donald W [email protected] Josyf C [email protected] Shohei [email protected] Stephen S [email protected] Barry I [email protected] Fernando [email protected] Department of Biochemistry, Wake Forest University School of Medicine, Winston-Salem, North Carolina, 27157, USA2 Department of Internal Medicine, Wake Forest University School of Medicine, Winston-Salem, North Carolina, 27157, USA3 Department of Physiology and Pharmacology, Wake Forest University School of Medicine, Winston-Salem, North Carolina, 27157, USA4 Department of Public Health Sciences, Wake Forest University School of Medicine, Winston-Salem, North Carolina, 27157, USA5 Center for Human Genomics, Wake Forest University School of Medicine, Winston-Salem, North Carolina, 27157, USA2005 7 12 2005 6 42 42 12 7 2005 7 12 2005 Copyright © 2005 Bento 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 GLUT10 (gene symbol SLC2A10) is a facilitative glucose transporter within the type 2 diabetes (T2DM)-linked region on chromosome 20q12-13.1. Therefore, we evaluated GLUT10 as a positional candidate gene for T2DM in Caucasian Americans. Methods Twenty SNPs including 4 coding, 10 intronic and 6 5' and 3' to the coding sequence were genotyped across a 100 kb region containing the SLC2A10 gene in DNAs from 300 T2DM cases and 310 controls using the Sequenom MassArray Genotyping System. Allelic association was evaluated, and linkage disequilibrium (LD) and haplotype structure of SLC2A10 were also determined to assess whether any specific haplotypes were associated with T2DM. Results Of these variants, fifteen had heterozygosities greater than 0.80 and were analyzed further for association with T2DM. No evidence of significant association was observed for any variant with T2DM (all P ≥ 0.05), including Ala206Thr (rs2235491) which was previously reported to be associated with fasting insulin. Linkage disequilibrium analysis suggests that the SLC2A10 gene is contained in a single haplotype block of 14 kb. Haplotype association analysis with T2DM did not reveal any significant differences between haplotype frequencies in T2DM cases and controls. Conclusion From our findings, we can conclude that sequence variants in or near GLUT10 are unlikely to contribute significantly to T2DM in Caucasian Americans. ==== Body Background Multiple genetic studies have been carried out that link human chromosome 20q13.1-13.2 to type 2 diabetes (T2DM) [1-5]. This linkage evidence has led investigators to search for T2DM susceptibility genes in this genomic region. Our laboratory has carried out analysis of specific genes [6-8] and developed high resolution physical maps of the region [9-11]. In an association analysis of genetic markers Price et al. [12] identified three regions of T2DM susceptibility. Among the genes mapped to the linkage disequilibrium regions, a novel facilitative glucose transporter (GLUT) was identified and designated GLUT10 (gene symbol SLC2A10) [6,13]. The gene spans 28 kb of genomic sequence, is split into 5 exons and 4 introns [6,13] and is expressed mainly in heart, liver, lung, skeletal muscle, pancreas, placenta, thyroid, and adipose tissue [6,13,14]. In Xenopus oocytes, human GLUT10 exhibited a high affinity, saturable 2-deoxy-D-glucose transport activity [6]. Glucose transport plays a central role in metabolism. Defects in glucose transport have been implicated in the reduced insulin sensitivity and in the increased insulin resistance observed in type 2 diabetes [14,15]. The genes coding for the classical GLUT1-5 proteins have been extensively analyzed for mutations contributing to type 2 diabetes, but no common causative mutation has been identified [16,17]. Both the function and genomic location of the novel SLC2A10 gene are consistent with a role in T2DM. In addition, Andersen et al. recently observed evidence for association of an Ala206Thr polymorphism in SLC2A10 with lower fasting insulin levels although they observed no evidence for association with T2DM [18]. We have carried out a systematic genetic evaluation of SLC2A10 to assess association with T2DM. Methods Subjects The individual samples evaluated in this study consisted of a collection of 300 unrelated Caucasian T2DM patients with end-stage renal disease (ESRD), with a corresponding collection of 310 randomly-ascertained unrelated Caucasian subjects without known diabetes. Both cases and controls were recruited simultaneously. This group will be referred to as "T2DM-ESRD"; their ascertainment and recruitment have been previously described in detail [19]. The T2DM-ESRD subjects have a mean age at diagnosis of diabetes of 46.5 ± 12.8 years, mean BMI at recruitment of 28.5 ± 7.0, and mean maximum reported BMI of 36.1 ± 8.3, mean duration of diabetes > 15 years, and mean HbA1c of 8.6%. SNP selection and genotyping Twenty SNPs used in this study were selected from the dbSNP public database (rs1004571, rs6012006, rs4810542, rs4810544, rs2425897, rs4428069, rs2425902, rs6094438, rs2235491, rs2076293, rs2425906, rs2425907, rs6094440, rs998422, rs2425911, rs3091904, rs6018008, rs707507, rs1059217, rs6090547). SNPs with frequency information were preferentially selected. The majority of the SNPs had minor allele frequencies greater than 0.2. Genotyping was performed on a Sequenom MassArray Genotyping System using methods previously described [19]. Discordance between blind duplicate samples included in the genotyping was <0.9% and the call rate for each assay was set at ≥ 90%. Average call rate was 98.1% (range 95.6% to 100%). Statistical analysis Pearson's test of homogeneity of proportions was applied to analyze allele frequency differences between diabetic and nondiabetic subjects SNPs were tested for Hardy-Weinberg equilibrium and pairwise linkage disequilibrium statistics D' calculated using the SNP-Analysis software package [20]. HAPLO.SCORE [21] was used to test for association of haplotypes within the case-control populations. HAPLO.SCORE uses posterior population-based frequency weighting of haplogenotypes that have prior consistency with observed heterozygous diplotypes. Variance estimates are inflated to account for uncertainty in specific haplotype assignment. We performed a global test of association between GLUT10 haplotypes and T2DM status, followed by individual tests of haplotype risk, in a collapsed 2 × 2 analysis. The skip.haplo option was set to 0.01. DANDELION [22], another expectation-maximization (EM) algorithm-based program for haplotype analyses was used to calculate haplotype frequencies in specific subject groups. A P < 0.05 was considered statistically significant. Results To examine the association of SLC2A10 variants with T2DM, 20 SNPs spanning the genomic region were evaluated. The average SNP density was 1 SNP/5.2 kb with inter-SNP distances ranging from 0.078 kb to 26.856 kb. The location of the SNPs relative to the exon-intron structure of SLC2A10 is shown in Figure 1. The 20 SNPs were genotyped in a collection of Caucasian T2DM-ESRD cases (n = 300) and controls (n = 310) and single SNP association analysis was performed on those SNPs with minor allele frequencies greater than 0.20. Five of the SNPs are located in coding regions (rs6094438, rs2235491, rs6018008, rs707507, rs1059217), but only rs1059217 had a minor allele frequency greater than 0.20. The coding variant rs2235491 (Ala206Thr) had a low minor allele frequency of 0.04, but was included since it has been previously described as associated with reduced fasting insulin [18]. The remaining coding variants, rs6094438, rs6018008, rs707507, are so rare that they are very poorly informative with minor allele frequencies ≤ 0.01. In addition two SNPs in the noncoding regions, rs4428069 and rs609440, were also found to be nonpolymorphic. In the control population, rs2076293 (P = <0.0001) and rs6090547 (P = 0.00043) did not conform to Hardy-Weinberg equilibrium. The allelic association analysis is summarized in Table 1, which shows the SNP identifier, the relevant alleles for each SNP, frequencies in cases and controls, and P-values. This analysis revealed moderate evidence of association between SNP rs2076293 and T2DM (P = 0.025). In addition, genotypic association analysis showed moderate evidence of association of this variant with T2DM under an additive model (P = 0.039). However, these results should be viewed with caution since the proportions of the genotypes for the variant are significantly out of Hardy-Weinberg equilibrium in the control population. Linkage disequilibrium (LD) and haplotype structure of SLC2A10 were determined to assess whether any specific haplotypes were associated with T2DM. Inter-SNP D' statistics were calculated for the gene region as shown in Figure 2 using LD Viewer (unpublished). The upper portion of the matrix displays the calculated inter-SNP D' values, where a D' ≥ 0.70 is defined as high LD. The lower portion of the matrix shows the calculated P-values for each D' value, where a P-value ≤ 0.05 is significant (e.g. the calculated D' value is significant). Using these criteria (D' ≥ 0.70, P ≤ 0.05), SLC2A10 SNPs appear to be one LD block covering a 14 kb region which begins with variant rs2425902 in intron 1 and ends beyond the 3'-UTR with variant rs1059217. Haplotype frequencies were estimated and association analyses performed using the programs Dandelion [22] and HAPLO.SCORE [21]. Shown in Table 2 are the results from the analyses evaluating 7 SNP haplotypes within the single major LD block of the SLC2A10 gene. Three common haplotypes are estimated and no significant evidence of haplotype association was observed with T2DM. It is important to note that although the variants rs2235491 and rs2235491 were within the LD block, they were removed from the analysis due to their low minor allele frequencies (rs2235491) or due to departure from Hardy-Weinberg equilibrium (rs2235491). Discussion SNPs spanning the SLC2A10 genomic sequence have been genotyped in a Caucasian American T2DM case-control population. Most of the coding SNPs validated in this gene region were so rare in this study sample that they were uninformative for our analysis. The remaining SNPs did not show significant evidence for association to T2DM. This included SNP rs2235491 (Ala206Thr) which was previously reported to be associated with fasting insulin (but without evidence of associaton with T2DM) [18]. Linkage disequilibrium analysis suggests that there is one LD block covering a 14 kb region beginning with variant rs2425902 in intron 1 and ending with variant rs1059217 in the 3'UTR region. Haplotype analysis based on the LD structure did not reveal evidence for association with T2DM. However, the study is not powered (with the criterion MAF > 0.1) or designed to address less common SNPs or haplotypes. We have carried out a detailed power analysis for this case-control study group elsewhere (Bento et al, submitted). Briefly, assuming a T2DM disease prevalence of 10% and modest multiplicative genotype risk ratios of 1.5, the power to detect association at a nominal SNP significance level of 5% is 74.7% under an additive trend test with risk allele frequency of 10%, and >95% for risk allele frequency 20%. Conclusion From these observations, we can conclude that variants in or near GLUT10 do not contribute substantially to T2DM in this sample of Caucasian Americans. This does not exclude the possibility that evidence for association could be observed in a larger case-control study or a study with cases ascertained in a different manner. However, another study has confirmed these negative association results. When examining SNPs in SLC2A10 using a Finnish population, no significant evidence for T2DM association with any SNP was observed [23]. In addition, this same study and another performed in a Danish population could not confirm the published association of Ala206Thr and fasting insulin [18, 24]. It is noteworthy that we have observed evidence for association to T2DM with the nearby PTPN1 gene, and to a lesser extent, with the HNF4A gene which is also located within 20q12-13.1. Abbreviations SNP- single nucleotide polymorphism, T2DM- Type 2 diabetes, LD- linkage disequilibrium, ESRD- end stage renal disease, BMI- body mass index, GLUT10- glucose transporter 10, SLC2A10- (gene name for GLUT10) solute carrier family 2 member 10, PTPN1- protein tyrosine phosphatase non-receptor type 1, HNF4A- hepatocyte nuclear factor 4-alpha Competing interests The author(s) declare that they have no competing interests. Authors' contributions Dr. Bento carried out the majority of the genotyping and data analysis. The data analysis plan was designed and supervised by Drs. Mychaleckyj and Rich. Dr. Hirakawa carried out the initial laboratory studies identifying and genotyping SNPs in the coding sequence of GLUT10. Dr. Freedman oversaw recruitment, diagnosis, and clinical characterization of the subjects used in this study. Drs. Bowden and Segade conceived the overall study design and supervised the performance of the study. All authors participated in writing and editing the manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements This work was supported by NIH grants R01 DK56289 to DWB. Figures and Tables Figure 1 Genomic map of the SLC2A10 gene with the locations of the 20 genotyped SNPs. The shaded regions are exons, numbered 1–5. The ruler along the bottom represents the relative location and spacing of SNPs in kilobases within the 40 kb region containing SLC2A10. Note that this does not have a uniform scale. Figure 2 Marker-to-marker D' plot for the SLC2A10 SNPs. Inter-SNP D' values are graphically represented using LD Viewer which generates a color coded plot of the pair-wise statistics in the upper portion of the matrix, and a color coded plot of the calculated P-value for each LD measurement in the bottom portion of the matrix. Table 1 Association analysis for SLC2A10 SNPs in Caucasian T2DM-ESRD cases and controls. Genotype Frequency: Cases Genotype Frequency: Ccontrols SNP Alleles 1/1 1/2 2/2 Frequency in Cases (n = 300) 1/1 1/2 2/2 Frequency in Controls (n = 310) P-value rs1004571 A/G 25 125 146 0.30/0.70 28 139 141 0.32/0.68 0.426 rs6012006 A/C 26 104 159 0.27/0.73 29 130 135 0.32/0.68 0.063 rs4810542 G/A 3 40 246 0.08/0.92 0 33 271 0.05/0.95 0.094 rs4810544 C/T 4 33 257 0.07/0.93 0 33 277 0.05/0.95 0.263 rs2425897 A/T 0 17 279 0.03/0.97 0 19 287 0.03/0.97 0.78 rs2425902 G/A 25 119 151 0.29/0.71 37 120 146 0.32/0.68 0.207 rs2235491 A/G 4 16 276 0.04/0.96 1 16 288 0.03/0.97 0.37 rs2076293 A/G 12 229 56 0.43/0.57 2 213 65 0.39/0.61 0.025 rs2425906 G/T 47 138 113 0.39/0.61 57 139 112 0.41/0.59 0.484 rs2425907 T/C 45 140 113 0.39/0.61 56 140 112 0.41/0.59 0.401 rs998422 G/A 46 139 114 0.39/0.61 58 138 108 0.42/0.58 0.289 rs2425911 G/C 26 123 146 0.30/0.70 40 121 144 0.33/0.67 0.213 rs3091904 C/T 44 147 107 0.39/0.61 59 139 111 0.42/0.58 0.541 rs1059217 C/T 46 137 104 0.40/0.60 60 137 111 0.42/0.58 0.567 rs6090547 C/T 71 108 90 0.47/0.53 67 107 102 0.44/0.56 0.381 *Variant not consistent with Hardy-Weinberg proportions P-values in bold indicates P ≤ 0.05 Table 2 Haplotype analysis using HAPLO.SCORE and Dandelion with 7 SNP (rs2425902, rs2425906, rs2425907, rs998422, rs2425911, rs3091904, rs1059217) SLC2A10 haplotypes. Haplotype Case Frequency Control Frequency Hap-Score Empirical Hap-specific P-value Global simulated P-value ATCACTT 60.04% 58.23% 0.51 0.61 GGTGGCC 29.15% 32.74% -1.19 0.21 0.37 AGTGCCC 8.51% 8.22% 0.12 0.88 ==== Refs Bowden DW Sale M Howard TD Qadri A Spray BJ Rothschild CB Akots G Rich SS Freedman BI Linkage of genetic markers on human chromosomes 20 and 12 to NIDDM in Caucasian sib pairs with a history of diabetic nephropathy Diabetes 1997 46 882 886 9133559 Ji L Malecki M Warram JH Yang Y Rich SS Krolewski AS New susceptibility locus for NIDDM is localized to human chromosome 20q Diabetes 1997 46 876 881 9133558 Klupa T Malecki MT Pezzolesi M Ji L Curtis S Langefeld CD Rich SS Warram JH Krolewski AS Further evidence for a susceptibility locus for type 2 diabetes on chromosome 20q13.1-q13.2 Diabetes 2000 49 2212 2216 11118028 Permutt MA Wasson JC Suarez BK Lin J Thomas J Meyer J Lewitzky S Rennich JS Parker A DuPrat L Maruti S Chayen S Glaser B A genome scan for type 2 diabetes susceptibility loci in a genetically isolated population Diabetes 2001 50 681 685 11246891 Zouali H Hani EH Philippi A Vionnet N Beckmann JS Demenais F Froguel P A susceptibility locus for early-onset non-insulin dependent (type 2) diabetes mellitus maps to chromosome 20q, proximal to the phosphoenolpyruvate carboxykinase gene Hum Mol Genet 1997 6 1401 1408 9285775 10.1093/hmg/6.9.1401 Dawson PA Mychaleckyj JC Fossey SC Mihic SJ Craddock AL Bowden DW Sequence and functional analysis of GLUT10: a glucose transporter in the Type 2 diabetes-linked region of chromosome 20q12-13.1 Mol Genet Metab 2001 74 186 199 11592815 10.1006/mgme.2001.3212 Fossey SC Kuroda S Price JA Pendleton JK Freedman BI Bowden DW Identification and characterization of PRKCBP1, a candidate RACK-like protein Mamm Genome 2000 11 919 925 11003709 10.1007/s003350010174 Price JA Fossey SC Sale MM Brewer CS Freedman BI Wuerth JP Bowden DW Analysis of the HNF4 alpha gene in Caucasian type II diabetic nephropathic patients Diabetologia 2000 43 364 372 10768098 10.1007/s001250050055 Fossey SCMJCPJKSJRBJTHSRSSFBIBDW A high resolution 6.0 Mb transcript map of the Type 2 diabetes susceptibility region on human chromosome 20 Genomics 2000 76 45 57 10.1006/geno.2001.6584 Price JA Brewer CS Howard TD Fossey SC Sale MM Ji L Krolewski AS Bowden DW A physical map of the 20q12-q13.1 region associated with type 2 diabetes Genomics 1999 62 208 215 10610714 10.1006/geno.1999.6007 Bento JLMJCRBHPFBIBDW A SNP map of human chromosome 20q12-13.1 American Journal of Human Genetics 2003 73 supplemental 486 Price JABCSHTDFSCRSSFBIWJPBDW Construction of a physical map of chromosome 20q12-13.1 and linkage disequilibrium analysis in diabetic nephropathy patients American Journal of Human Genetics 1997 58 supplemental A241 McVie-Wylie AJ Lamson DR Chen YT Molecular cloning of a novel member of the GLUT family of transporters, SLC2a10 (GLUT10), localized on chromosome 20q13.1: a candidate gene for NIDDM susceptibility Genomics 2001 72 113 117 11247674 10.1006/geno.2000.6457 Wood IS Hunter L Trayhurn P Expression of Class III facilitative glucose transporter genes (GLUT-10 and GLUT-12) in mouse and human adipose tissues Biochem Biophys Res Commun 2003 308 43 49 12890477 10.1016/S0006-291X(03)01322-6 Matthaei S Stumvoll M Kellerer M Haring HU Pathophysiology and pharmacological treatment of insulin resistance Endocr Rev 2000 21 585 618 11133066 10.1210/er.21.6.585 Rothman DL Magnusson I Cline G Gerard D Kahn CR Shulman RG Shulman GI Decreased muscle glucose transport/phosphorylation is an early defect in the pathogenesis of non-insulin-dependent diabetes mellitus Proc Natl Acad Sci U S A 1995 92 983 987 7862678 Cline GW Petersen KF Krssak M Shen J Hundal RS Trajanoski Z Inzucchi S Dresner A Rothman DL Shulman GI Impaired glucose transport as a cause of decreased insulin-stimulated muscle glycogen synthesis in type 2 diabetes N Engl J Med 1999 341 240 246 10413736 10.1056/NEJM199907223410404 Andersen G Rose CS Hamid YH Drivsholm T Borch-Johnsen K Hansen T Pedersen O Genetic variation of the GLUT10 glucose transporter (SLC2A10) and relationships to type 2 diabetes and intermediary traits Diabetes 2003 52 2445 2448 12941788 Bento JL Palmer ND Mychaleckyj JC Lange LA Langefeld CD Rich SS Freedman BI Bowden DW Association of protein tyrosine phosphatase 1B gene polymorphisms with type 2 diabetes Diabetes 2004 53 3007 3012 15504984 Schaid DJ Rowland CM Tines DE Jacobson RM Poland GA Score tests for association between traits and haplotypes when linkage phase is ambiguous Am J Hum Genet 2002 70 425 434 11791212 10.1086/338688 Green LELEMLCD Power comparison of phase-known versus phase-unknown haplotype analyses for case-control designs Am J Hum Genet 2001 69 1948a Mohlke KL Skol AD Scott LJ Valle TT Bergman RN Tuomilehto J Boehnke M Collins FS Evaluation of SLC2A10 (GLUT10) as a candidate gene for type 2 diabetes and related traits in Finns Mol Genet Metab 2005 85 323 327 15936967 10.1016/j.ymgme.2005.04.011 Rose CS Andersen G Hamid YH Glumer C Drivsholm T Borch-Johnsen K Jorgensen T Pedersen O Hansen T Studies of relationships between the GLUT10 Ala206Thr polymorphism and impaired insulin secretion Diabet Med 2005 22 946 949 15975113 10.1111/j.1464-5491.2005.01547.x	16336637	PMC1325051	CC BY	2021-01-04 16:03:33	no	BMC Med Genet. 2005 Dec 7; 6:42	utf-8	BMC Med Genet	2,005	10.1186/1471-2350-6-42	oa_comm
==== Front BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2871632115710.1186/1471-2105-6-287SoftwareArrayQuest: a web resource for the analysis of DNA microarray data Argraves Gary L [email protected] Saurin [email protected] Jeremy L [email protected] W Scott [email protected] Array Genetics, Inc., 59 Great Quarter Road, Newtown, CT 06482, USA2 Department of Cell Biology and Anatomy, Medical University of South Carolina, Charleston, SC 29425 USA2005 1 12 2005 6 287 287 22 7 2005 1 12 2005 Copyright © 2005 Argraves et al; licensee BioMed Central Ltd.2005Argraves 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 Numerous microarray analysis programs have been created through the efforts of Open Source software development projects. Providing browser-based interfaces that allow these programs to be executed over the Internet enhances the applicability and utility of these analytic software tools. Results Here we present ArrayQuest, a web-based DNA microarray analysis process controller. Key features of ArrayQuest are that (1) it is capable of executing numerous analysis programs such as those written in R, BioPerl and C++; (2) new analysis programs can be added to ArrayQuest Methods Library at the request of users or developers; (3) input DNA microarray data can be selected from public databases (i.e., the Medical University of South Carolina (MUSC) DNA Microarray Database or Gene Expression Omnibus (GEO)) or it can be uploaded to the ArrayQuest center-point web server into a password-protected area; and (4) analysis jobs are distributed across computers configured in a backend cluster. To demonstrate the utility of ArrayQuest we have populated the methods library with methods for analysis of Affymetrix DNA microarray data. Conclusion ArrayQuest enables browser-based implementation of DNA microarray data analysis programs that can be executed on a Linux-based platform. Importantly, ArrayQuest is a platform that will facilitate the distribution and implementation of new analysis algorithms and is therefore of use to both developers of analysis applications as well as users. ArrayQuest is freely available for use at . ==== Body Background Numerous Open Source projects contribute source code for routines that analyze DNA microarray data [1]. Projects such as Bioconductor [2] and Bioperl [3] offer web accessible repositories for DNA microarray software packages. However, obstacles such as lack of expertise in configuring the software for use on one's local workstation (i.e., installing operating systems such as R and packages that run under R) and inadequate hardware resources (i.e., computers with high CPU throughput) may prevent implementation of these tools by biologists. Providing browser-based interfaces that allow the programs to be executed over the Internet can enhance the applicability of these advanced analytic software tools by researchers. Here we describe a new web-based DNA microarray analysis process controller that permits application of Open Source programs to analyze DNA microarray data. ArrayQuest is designed as a broadly applicable process controller that will implement DNA microarray analysis programs executable on a Linux system. We believe that this feature can help biologists employ the wealth of Open Source software that has been developed for analyzing DNA microarray datasets. There are a number of existing web-server applications that perform microarray data transformations such as data normalization, two-condition comparisons and unsupervised and supervised clustering [4-9]. In general these tools execute analysis steps sequentially, and the output from one analysis step is then input or specified prior to subsequent analysis. For example, tools such as Expression Profiler: Next Generation developed at EBI [10] will execute a normalization of hybridization data that is then output to a file. This data file must then be uploaded (perhaps after reformatting the data) before differentially expressed genes may be identified. Similarly, the differentially expressed genes must be output to a file and this data must be uploaded before any subsequent analysis can be conducted. We have designed ArrayQuest to execute analysis processes comprised of bundles of different routines, thereby simplifying the execution of multi-step analyses. Implementation ArrayQuest programming ArrayQuest was written in PHP4 and Linux Bash Shell Scripting language and runs on multiple Linux RedHat servers. Project data, DNA microarray data supersets and analysis process control data are saved in a MySQL database. The backend analysis computers use Secure Shell (Shh) as the communication link to the center point web server. The system is designed to permit expanded analysis capacity by distributing analysis requests to servers outside of the local area network. ArrayQuest usage ArrayQuest [11] is accessible through web browsing portals (i.e., Internet Explorer, Safari, Mozilla, etc.) and can be used after registering to obtain a free password-protected account. The ArrayQuest system allows Internet clients to connect to a center point web-server to create a new analysis project or select an existing project folder (Fig. 1). Project folders contain Principal Investigator (PI) and project information entered via a web form. Once a project folder has been created, the user may choose to upload data files from their computer to create their own private database. Next, a single analysis method is selected from methods that are stored in the Methods Library. Depending on the method chosen, input microarray data for an analysis can be selected from either the MUSC DNA Microarray Database [12], a remote database (e.g., GEO) or the user's private database containing data uploaded to the center point web-server. Upon launching an analysis process, the data required to perform the analysis is sent to a computer belonging to a cluster of analysis computers. At the Medical University of South Carolina, the hardware configuration for a computer within the ArrayQuest analysis computer cluster is a dual Opteron CPU (AMD 64 bit) having 2 gigabyte of RAM, running under Fedora Core. Each computer in the cluster is loaded with the R programming language [13] and Bioconductor software packages [2,14]. Besides R and Bioconductor analysis tools, the system is capable of executing other types of analysis programs (e.g., BioPerl and C++). Figure 1 Schematic diagram depicting ArrayQuest system topography and steps in the process of performing an analysis of DNA microarray data. As indicated, DNA microarray data can be obtained from multiple sources including the MUSC DNA Microarray Database, the NIH GEO database or a user's private database. Results reporting ArrayQuest automatically recognizes result files that are produced by an analysis script based on file time tagging. When an analysis script has completed execution, the files are copied from the analysis computer back to the center point web server for viewing (Fig. 1). When requests require lengthy run times, the system will notify the user upon completion by email. The script status can also be queried in real time through the web browser to determine whether an analysis job is running or completed, the duration of the analysis, whether an error has occurred and which analysis computer in the cluster is running the analysis process. User management and privileges The system has two access levels, administrator and user. All users are allowed to work in a password-protected environment, private from all other users. Users have access to all publicly available data in the MUSC DNA Microarray Database as well as any privately held data that they have deposited in the database. Users cannot create or modify analysis methods but can work with administrators to have methods developed and/or implemented. Lists of users will never be shared with a third party. The system will also send (email) forgotten passwords to users on request. Adding new analysis methods to ArrayQuest New analysis methods can be added to the Methods Library at the request of users or developers. A method developer may submit a program description or script to the ArrayQuest system administrator for consideration [11]. Once approved, ArrayQuest personnel will aide in the implementation of the method on the ArrayQuest platform. Non-developers may also request that ArrayQuest personnel create a method from existing Open Source programs. Since the specification of analysis parameters can be conducted entirely via a parameters entry text window, new methods do not require that graphical user interfaces (GUIs) be created. This alleviates some of the burden of implementing new analysis methods that may be encountered with other online analysis tools that use GUIs for setting analysis parameters. Results and discussion There are a number of existing web-server applications that perform common data transformations such as data pre-processing, two-condition comparisons and unsupervised and supervised clustering [4-9]. One feature that distinguishes ArrayQuest from these other applications is that ArrayQuest is designed as a broadly applicable process controller that will implement DNA microarray analysis programs executable on a Linux system. Therefore, ArrayQuest is not limited to execution of a subset of specific analysis functions, but is instead capable of executing a large number of analysis functions that have been generated by the Open Source community and that can be implemented in a Linux-based system. The ArrayQuest Methods Library is currently populated with six methods for analysis of Affymetrix DNA microarray data. These include a number of Bioconductor-based statistical and graphical methods written in R that accept Affymetrix .CEL files and one method that accepts GEO GDS .SOFT files (Table 1). The methods range in complexity from simple data normalization (method 10) to the more comprehensive procedure of normalization, identification of differentially expressed genes, hierarchical clustering, generation of a heatmap and significance analysis of gene ontologies (method 12). In general, individual methods were created by combining or "stacking" Bioconductor packages in order to execute a series of linked analysis routines. Depending on the method selected, users will be required to specify input parameters for parsing of the data such as control and experimental filenames, thresholds for differential expression (e.g., fold change, t-test p-value and false discovery rate) and/or GO IDs. Table 1 Representative analysis methods held in the ArrayQuest Methods Library. Method Title Method Description Required Data Format Data Source Programming Language/ Software1 Output RMA Normalization of Affymetrix Data This method performs Robust Multichip Analysis (RMA) to generate normalized expression intensities for a set of Affymetrix GeneChip CEL files. Affymetrix GeneChip data in .CEL file format MUSC DNA Microarray Database or User's Private Database R/Bioconductor A Microsoft Excel file of normalized intensities transformed into log base 2 for all genes and four JPEG files of box plots and histograms of expression intensities before and after normalization. Identification of differentially expressed genes based on fold-change, p-value and/or FDR parameters This method is used to analyze data from any two-condition microarray experiment. The algorithm normalizes hybridization data, finds differentially expressed genes based on fold-change, t-test and FDR thresholds, collects annotations for these genes, performs hierarchical clustering and renders a heat map of the expression profiles. Affymetrix GeneChip data in .CEL file format MUSC DNA Microarray Database or User's Private Database R/Bioconductor Annotation reports (Excel and HTML); Heatmap of differentially expressed genes (.JPEG); KEGG pathway heat maps of differentially expressed genes (as many as are found) (.JPEG); GO Information (HTML). Identification of differentially expressed genes based on p-value, fold-change and/or FDR parameters: .SOFT files only This method is used to analyze Affymetrix DNA microarray data that can be obtained from NIH GEO as a .SOFT.gz file. The method normalizes hybridization data (RMA), finds differentially expressed genes based on fold-change, t-test and FDR thresholds, collects annotations for these genes, performs hierarchical clustering and renders a heat map of the expression profiles. Affymetrix GeneChip data in GEO .SOFT file format Gene Expression Omnibus (GEO) R/Bioconductor Annotation reports (Excel and HTML); Heatmap of differentially expressed genes (.JPEG); KEGG pathway heat maps of differentially expressed genes (as many as are found) (.JPEG); GO Information (HTML). Assessment of gene expression associated with a specified GO ID(s) This method analyzes Affymetrix GeneChip data to find gene expression values associated with specified GO IDs. The script normalizes GeneChip hybridization data (RMA), extracts hybridization values for genes associated with a user-provided GO ID, performs hierarchical clustering and renders a heat map of the expression profiles. Affymetrix GeneChip data in .CEL file format MUSC DNA Microarray Database or User's Private Database R/Bioconductor Boxplots and histograms of expression intensities before and after normalization (each in JPEG). Heat map based on the number of GO IDs provided by the user (each in JPEG). 1ArrayQuest methods displayed in this table are written using the R statistical computing language [13] and implement packages/algorithms that are the product of the Bioconductor Open Source software development project [2]. Bioconductor packages used in ArrayQuest methods and links to package descriptions and developers are given in the description window for each method in the ArrayQuest Methods Library [11]. The ability to execute bundled analyses in one step is another feature that distinguishes ArrayQuest from other online microarray analysis tools. For example, after specifying a group of raw hybridization data files Method 12 will identify differentially expressed genes, find significantly represented gene ontologies and perform hierarchical clustering. Online tools that perform each of these steps independently require that the user continuously interface with the website to enact each step. This tends to increase the time required to execute the analysis and can increase the complexity and difficulty of the analysis. By combining all stages of the analysis into a single process, as is the ability of ArrayQuest, the overall analysis process is significantly simplified and speeded up. A typical execution of Method 12 involving two sample groups (≤ 5 replicates each) may be completed in approximately 10 minutes. Conclusion ArrayQuest will serve as a useful system for analysis of DNA microarray data on-line and will also enable software developers to make their DNA microarray analysis routines readily available to the research community. Availability and requirements Project name: ArrayQuest Project home page: Operating system(s): Operates online via a browser web portal. Web servers use Redhat Linux and Fedora Linux. Programming language: PHP4 patched to support file uploading status ); bash and standard Linux system utilities. Other requirements: Apache (HTTP server), mySQL (Structured Query Language server), ssh (Secure Shell), awk, scp (secure copy) and microArrayDB (μArrayDB; ). License: GNU General Public License (GPL) and BSD as applicable to subsystems of ArrayQuest. Any restrictions to use by non-academics: Not at this time. Anonymous review of ArrayQuest: ArrayQuest can be accessed in an anonymous fashion at using the guest user account (Username / password: [email protected] / test). This account is populated with a project containing two analyses. One of these (Sample Analysis Process I) is intended as an analysis that the user can modify, execute and then check analysis results. The other (Sample Analysis Process II) should not be modified and is intended only as an example of the analysis procedure and a demonstration of analysis output. Analysis script availability: Analysis scripts employed in ArrayQuest methods are freely available and can be found on the Methods Library List page () by toggling the "Script" button associated with every analysis method. List of abbreviations used GEO, Gene Expression Omnibus; GO ID, Gene Ontology ID; MUSC, Medical University of South Carolina. Authors' contributions Gary L. Argraves is the primary designer and systems programmer of ArrayQuest. Saurin Jani worked to implement R-based Bioconductor routines to run on ArrayQuest and consulted on system features and debugging. Jeremy L. Barth directed the creation of ArrayQuest analysis process routines. W. Scott Argraves is a designer of ArrayQuest and Principal Investigator on the NIH grant that funded the project. Acknowledgements This work was supported by grants from the National Cancer Institute (R24CA095841) and the National Heart Lung and Blood Institute (P20RR016434). ArrayQuest development costs were also provided by ProSoft Systems, Newtown, CT. The authors wish to acknowledge the contribution of Joshua Spruill (Department of Cell Biology and Anatomy, MUSC) for his efforts in maintaining the ArrayQuest servers and back-end computers. ==== Refs Dudoit S Gentleman RC Quackenbush J Open source software for the analysis of microarray data Biotechniques 2003 Suppl 45 51 12664684 Bioconductor Bioconductor open source software for bioinformatics Bioperl Bioperl Project Vaquerizas JM Dopazo J Diaz-Uriarte R DNMAD: web-based diagnosis and normalization for microarray data Bioinformatics 2004 20 3656 3658 15247094 Hokamp K Roche FM Acab M Rousseau ME Kuo B Goode D Aeschliman D Bryan J Babiuk LA Hancock RE Brinkman FS ArrayPipe: a flexible processing pipeline for microarray data Nucleic Acids Res 2004 32 W457 9 15215429 Cheung KH Hager J Pan D Srivastava R Mane S Li Y Miller P Williams KR KARMA: a web server application for comparing and annotating heterogeneous microarray platforms Nucleic Acids Res 2004 32 W441 4 15215426 Herrero J Al-Shahrour F Diaz-Uriarte R Mateos A Vaquerizas JM Santoyo J Dopazo J GEPAS: A web-based resource for microarray gene expression data analysis Nucleic Acids Res 2003 31 3461 3467 12824345 10.1093/nar/gkg591 Psarros M Heber S Sick M Thoppae G Harshman K Sick B RACE: Remote Analysis Computation for gene Expression data Nucleic Acids Res 2005 33 W638 43 15980552 10.1093/nar/gki490 Comander J Weber GM Gimbrone MAJ Garcia-Cardena G Argus--a new database system for Web-based analysis of multiple microarray data sets Genome Res 2001 11 1603 1610 11544205 10.1101/gr.186601 Kapushesky M Kemmeren P Culhane AC Durinck S Ihmels J Korner C Kull M Torrente A Sarkans U Vilo J Brazma A Expression Profiler: next generation--an online platform for analysis of microarray data Nucleic Acids Res 2004 32 W465 70 15215431 ArrayQuest ArrayQuest Argraves GL Barth JL Argraves WS The MUSC DNA Microarray Database Bioinformatics 2003 19 2473 2474 14668234 10.1093/bioinformatics/btg325 R-Project The R Project for Statistical Computing Gentleman RC Carey VJ Bates DM Bolstad B Dettling M Dudoit S Ellis B Gautier L Ge Y Gentry J Hornik K Hothorn T Huber W Iacus S Irizarry R Leisch F Li C Maechler M Rossini AJ Sawitzki G Smith C Smyth G Tierney L Yang JY Zhang J Bioconductor: open software development for computational biology and bioinformatics Genome Biol 2004 5 R80 15461798 10.1186/gb-2004-5-10-r80	16321157	PMC1325052	CC BY	2021-01-04 16:27:47	no	BMC Bioinformatics. 2005 Dec 1; 6:287	utf-8	BMC Bioinformatics	2,005	10.1186/1471-2105-6-287	oa_comm
==== Front PLoS BiolPLoS BiolpmedplosmedPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0040032SynopsisMolecular Biology/Structural BiologyArchaeaRighting the Wrongs: Structural Insights into Replicating Damaged DNA Synopsis1 2006 3 1 2006 3 1 2006 4 1 e32Copyright: © 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. Stepwise Translocation of Dpo4 Polymerase during Error-Free Bypass of an oxoG Lesion ==== Body Every organism's blueprint for life is encoded in the order of the building blocks of its genome. These building blocks consist of four DNA nucleotides, or bases: cytosine (C), guanine (G), adenine (A), and thymine (T). Whenever a cell divides, it must duplicate its genome, and it must do so with high fidelity to maintain genome integrity. Replication is assisted by DNA polymerases. After binding to the DNA being copied (the template strand), a polymerase lines up a complementary base to the template base (C with G and A with T), then adds that base to the growing strand of nucleotides, called the primer strand. The polymerase then moves to the next position along the template DNA. This process is further complicated by the continuous battle all cells wage against a multitude of mutagens. The result of mutagenic damage, such as radiation, can include oxidized DNA modifications, which have been linked to increased risk of cancer. Oxidative stress produces oxidized nucleotides—the most common of these is 7,8-dihydro-8-oxoguanine (oxoG), an oxidized form of guanine—which sometimes persists in DNA as a lesion. The cell is then faced with the challenge of working out how to replicate and maintain genomic integrity in spite of this anomaly. Fortunately, the cell is armed with a toolbox to combat such situations. One of these tools, the Y-family polymerases, can bypass DNA lesions. Beyond knowing that these polymerases had this capability, not much was known about the details of the translocation mechanism, until now. In a new study, Olga Rechkoblit, Dinshaw Patel, and colleagues report a detailed insight into the mechanisms by which Dpo4, a member of the Y-family polymerases, bypasses a DNA lesion. To do this, they solved crystal structures of the polymerase at different stages of association with an oxoG-modified DNA template during its replication. These structures act as snapshots of this polymerase as it progresses through lesion recognition in a pre-nucleotide insertion complex to actual nucleotide insertion, and finally, to the post-insertion complex, indicating that lesion bypass has taken place. These structures helped reveal the translocation mechanics of the bypass polymerase during a complete cycle of nucleotide incorporation that could be compared with what was already known for replication polymerases. In response to DNA injury, the polymerase Dpo4 inserts a nucleotide, and three of its protein domains (light green, blue, magenta) shift upward (their prior position is shown in gray), while the fourth domain anchors Dpo4. This “bypass replication” maneuver allows DNA replication to proceed error-free Rechkoblit et al. initially determined which nucleotide was most readily inserted opposite an oxoG during the Dpo4-mediated replication process. To do this, they measured how commonly each of the four different nucleotides is inserted at this position, and the enzyme kinetics (the efficiency and speed with which the polymerase manages to carry this out) associated with each different nucleotide. What they saw was, reassuringly, that Dpo4 preferentially inserts a cytosine (dCTP) opposite the oxoG. Dpo4 is then able to continue extending beyond the lesion, facilitating error-free bypass. The authors next concentrated on the incorporation of a dCTP opposite oxoG in their crystal structures. Structurally, Dpo4 and other Y-family polymerases have four domains: palm, finger, thumb, and little finger. This hand formation enables the protein to fit around the DNA being replicated. In the three different structures solved, Rechkoblit et al. observed how the domains moved relative to one another and identified any key parts of Dpo4 that enabled this lesion bypass. In particular, they identified two amino acids (arginines 331 and 332) in Dpo4 that play a critical role in forming hydrogen bonds formed through attraction of the positive charge of hydrogen with the nearby negatively charged phosphate group (part of the backbone of DNA) of the oxoG. With respect to the individual domains, they saw that as the dCTP inserts opposite oxoG, the little finger domain that contacts DNA phosphate groups shifts by one nucleotide step. During the next step, when dCTP is chemically bonded into place opposite oxoG, the thumb domain–phosphate contacts move along by one nucleotide. Thus, the little finger and thumb domains do not move at the same time but rather in a stepwise manner, tracking the template- and primer-strand translocation separately. Such accumulated knowledge about DNA repair mechanisms, such as error-free lesion bypass by Dpo4, may eventually lead to cancer therapeutic approaches to reduce DNA lesions. In the meantime, such depth of mechanistic insight gleaned from these structures can lead us to wonder anew at the cell's capacity to produce such innovative solutions to the day-to-day problems it encounters. —Emma Hill	0	PMC1325053	CC BY	2021-01-05 08:21:48	no	PLoS Biol. 2006 Jan 3; 4(1):e32	utf-8	PLoS Biol	2,006	10.1371/journal.pbio.0040032	oa_comm
==== Front CytojournalCytoJournal1742-6413BioMed Central London 1742-6413-2-191628750210.1186/1742-6413-2-19Case ReportFine needle aspiration diagnosis of extracranial glioblastoma multiforme: Case report and review of the literature Schultz Stacey [email protected] Gregory S [email protected] Nancy C [email protected] Marc C [email protected] A Sonali [email protected] Sue E [email protected] Department of Pathology, USC/Keck School of Medicine, Los Angeles, USA2 Department of Medicine, USC/Keck School of Medicine, Los Angeles USA2005 14 11 2005 2 19 19 13 5 2005 14 11 2005 Copyright © 2005 Schultz et al; licensee BioMed Central Ltd.2005Schultz 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 Hitherto uncommon, the incidence of extracranial metastases of primary brain malignancies may increase, with improved treatment methods and longer patient survival. Fine needle aspiration biopsy is a simple, safe and reliable method to diagnose metastatic malignancy. It has definite advantages over tissue biopsy, which is more invasive and is of higher risk to the patient. Ours is a case of glioblastoma multiforme, which metastasized to the scalp and was diagnosed on fine needle aspiration biopsy. Only a few articles document the cytological features of extracranial glioblastoma multiforme, diagnosed by fine needle aspiration biopsy. Case presentation We report the case of an elderly female who presented with focal neurological symptoms. She was diagnosed radiologically with an intracranial lesion in the left temporal region, which was subsequently resected. Histology revealed a glioblastoma multiforme confirmed by immunohistochemistry. The tumor recurred subsequently and the patient was treated with chemotherapy, intraoperatively. At a later stage, she presented with a scalp mass on which fine needle aspiration biopsy was performed. The cytomorphological features aided by immunohistochemistry supported a diagnosis of metastatic glioblastoma multiforme. The mass was later resected and histology confirmed the fine needle aspiration diagnosis of glioblastoma multiforme. Conclusion Reports of extracranial metastases of primary brain tumors are few. When they do occur, the primary cause is implantation during surgery or biopsy. However, spontaneous metastases to other organs do occur rarely. We believe fine needle aspiration biopsy to be very useful in the diagnosis of metastatic glioblastoma multiforme. The ability to use a cellblock for immunohistochemical studies is greatly advantageous and helpful in differentiating this tumor, from other malignancies that can occur in the scalp. A detailed discussion of the material obtained from fine needle aspiration biopsy of metastatic glioblastoma multiforme is presented, as well as a review of previous accounts in the literature. ==== Body Background Extracranial metastasis of primary brain malignancies is an uncommon event [1]. With increased patient survival due to new treatment options, however, the incidence of extracranial metastases of brain malignancies may increase, as has been the case with systemic cancers. An alternative to tissue biopsy to determine whether or not extracranial spread has occurred, is fine needle aspiration (FNA) biopsy. An FNA biopsy has several advantages over tissue biopsy, including immediate evaluation of tissue adequacy, minimal risk to the patient and the possibility of multiple passes to reduce the risk of sampling error. Because of the rarity of metastases of primary brain malignancies, the cytological findings of metastatic primary brain tumors are limited to only a few accounts. A literature review produced a total of five articles pertaining to the FNA biopsy features of extracranial metastatic primary brain tumors [2-6]. Four of the articles involved gliomas (three glioblastoma multiforme [3,4,6], and one low-grade astrocytoma [5]) and one pertained to a sacrococcygeal myxopapillary ependymoma [2]. This case report is submitted to detail the cytological findings in an extracranial glioblastoma multiforme (GBM). Case presentation Clinical history A 74-year-old female with no significant past medical history presented with complaints of headache in conjunction with speech and memory problems for approximately one month. Magnetic resonance imaging (MRI) revealed an enhancing cystic lesion in the left temporal lobe. A partial resection of the left temporal mass, revealed a malignant neoplasm on frozen section and on final pathological evaluation, a diagnosis of GBM was made. Immunohistochemical results showed positive staining for glial fibrillary acidic protein (GFAP) and no evidence of positive staining for cytokeratin. The patient received adjuvant external beam radiation therapy. Approximately nine months later the patient returned with both speech and short-term memory problems. MRI revealed tumor recurrence in the left temporal area. Stereotactic biopsy demonstrated a necrotic tumor consistent with GBM on frozen section, a diagnosis confirmed on permanent sections. The stereotactic biopsy was followed by intraoperative injection of the chemotherapeutic agent DTI-015 [7]. Three months later, the patient returned with a 2.0 cm subcutaneous mass in the left parietal scalp, approximately six centimeters from the closest postoperative scar. Fine needle aspiration biopsy of the left parietal scalp mass revealed a GBM. The lesion was excised a month later with histological confirmation of the diagnosis. Following excision of the scalp mass, a repeat MRI documented progression of the primary tumor. The patient died of pulmonary embolism one month after excision of the scalp mass. FNA biopsy of the palpable left parietal subcutaneous scalp mass was performed by a pathologist, using a 23 gauge needle. Four smears and one cell block were prepared. The air dried smears were stained with a modified Wright stain (Diff-Quik®); the 95% ethanol fixed smears were stained with the Papanicolaou stain. Microscopic description The smears were abundantly cellular with atypical cells arranged in loosely cohesive clusters, ranging from 5 to 40 cells per cluster, with overlapping nuclei (Fig. 1A–1D). Scattered atypical mitoses were also present (Fig. 1C). Morphologically, the cells were polygonal to spindle-shaped with increased nuclear to cytoplasmic ratios, moderate nuclear pleomorphism, coarsely clumped hyperchromatic chromatin, irregular nuclear membranes, and distinct nucleoli. Binucleated and multinucleated cells were also noted. A few scattered intranuclear inclusions were present (Fig. 1D). Fibrillary processes extending from the atypical cells were apparent. A tumor diathesis consisting of neutrophils, degenerated cells, necrotic debris, a few lymphocytes, and bare nuclei made up the background. The cell block showed similar findings as well as characteristic necrotic foci surrounded by palisading spindle-shaped atypical cells (geographic necrosis) (Fig. 2A). Immunohistochemically, the atypical cells displayed strong reactivity for GFAP (Fig. 2B) and vimentin with no reactivity for CD45, cytokeratin, and HMB-45, supporting the diagnosis of GBM. The histological sections from the tumor excised a month later verified the FNA findings of GBM (Fig. 2C–2D). Figure 1 Figure 1A: Smear showing abundant cellularity and necrosis (Diff-Quik ® stain, X100). Figure 1B: Smear showing atypical cells in loosely cohesive clusters (Diff Quik ® stain, X400). Figures 1C and 1D: Smears showing pleomorphic cells with coarsely clumped chromatin, irregular nuclear membranes, prominent nucleoli, atypical mitosis (1C, arrow), intranuclear inclusion (1D, arrow) and cellular processes (Papanicolaou stain X400) Figure 2 Figure 2A: Cell block showing geographic necrosis (H&E stain, X400). Figure 2B: Cell block showing positive staining of malignant cells for glial fibrillary acidic protein (immunostain for GFAP, X400). Figure 2C: Histology of surgical resection showing malignant cells with numerous mitotic figures in a desmoplastic background (H&E stain, X400). Figure 2D: Histology of surgical resection showing positive staining of malignant cells for glial fibrillary acidic protein (immunostain for GFAP, X400). Conclusion The cytologic and immunohistochemical features of this FNA, augmented by the patient's history of GBM, led to a straightforward diagnosis in this case. In the absence of a detailed clinical history, however, a consideration of other malignant neoplasms that occur in the scalp may be necessary in the initial work up. Table 1 gives a brief differential comparing the cytological features of other malignancies that may occur in the scalp of adults. Table 1 Cytologic Features of Malignant Neoplasms of the Scalp Tumor Cellularity Cells Nuclei Nucleoli Cytoplasm Mitosis Back-Ground IHC Metastatic GBM Abundant, Primarily single cells, with occasional small loosely cohesive clusters Round to oval to spindle Pleomorphic, coarsely clumped chromatin, INCIs, occasional binucleation Usually one, may have two Scant, cytoplasmic processes + Atypical forms Tumor diathesis GFAP Squamous cell carcinoma Abundant, tight to loosely cohesive, disorderly groups, with single cells Pleomorphic to uniform small cells, may see squamous pearls Central location, densely hyperchromatic to irregular chromatin clumping Inconspicuous to prominent Moderate, dense with cytoplasmic keratinization + Atypical forms Depends on grade of tumor Keratin Basal cell carcinoma Large tight crowded clusters with peripheral palisading Basaloid Small, round to oval, hyperchromatic Inconspicuous may be prominent Scant + No atypical forms Pink, amorphous material Ber-EP4 Angiosarcoma Scant to moderate, small clusters with single cells Whorl formation, bland to pleomorphic, erytrhophagocytosis Hyperchromatic, shallow longitudinal grooves Prominent in high grade neoplasms Scant to abundant, vacuolated + Atypical forms Bloody, necrotic Factor VIII, CD 31 CD 34 Melanoma Moderate to high, loosely cohesive groups, numerous single cells Epithelioid to spindle to pleomorphic Eccentric location, binucleation common, few INCIs Macronucleoli Moderate, granular, vacuoles, melanin + Atypical forms Clean to bloody, pigment-laden macrophages S100 HMB45 Extracranial spread of GBM, although rare, is a recognized phenomenon with an estimated incidence of less than 0.5%. It usually occurs in patients with a history of a previous craniotomy or a diversionary shunt. Proposed routes of dissemination include implantation of tumor during open surgery and during stereotactic biopsy [8]. Rare cases of spontaneous metastasis without a history of biopsy or surgery have been reported in the literature [9]. The first well-documented extracranial metastasis was reported in 1928 [10]. The sites of metastasis vary with the type of brain primary and include bone, lung, liver, and lymph node. In the pediatric population, medulloblastoma is the leading tumor for extracranial metastasis. In the adult, it is malignant astrocytoma/GBM. True metastases as well as extracranial recurrences have been reported as being diagnosed by FNA biopsy [7]. In previous reports, as well as this one, the specimens procured by FNA from extracranial GBM are quite cellular consisting of malignant-appearing cells arranged in small, loosely cohesive, disorderly clusters with a predominance of single cells. Cellular size ranges from small to large with variably shaped nuclei ranging from round to oval to spindle. All descriptions consistently include marked nuclear pleomorphism with high nuclear to cytoplasmic ratios, coarsely clumped hyperchromatic chromatin, nuclear membrane irregularity, prominent single or multiple nucleoli, multinucleation and necrosis. Occasional intranuclear and cytoplasmic inclusions and rare mitoses may be seen. Scant cytoplasm is noted in all cases; however, the presence of cytoplasmic processes extending from the malignant cells creating a fibrillary background (a characteristic feature for astrocytic neoplasms) is not described in all cases. A GFAP stain may be helpful in highlighting this feature. Thick-walled capillaries with endothelial proliferation are documented in one case study [4], but this feature is absent in our case, even with immunostaining for CD34. Cell block sections do reveal the characteristic geographic necrosis typical of GBM. In addition, the cell block allows the performance of a panel of immunostains. For this reason we recommend submitting at least one FNA pass, solely for the purpose of obtaining a cell block. As in this case, the cell block may be extremely helpful in rendering a definitive FNA diagnosis of this rare extracranial tumour. List of abbreviations Fine needle aspiration (FNA) Glial fibrillary acidic protein (GFAP) Glioblastoma multiforme (GBM) Competing interests The author(s) declare that they have no competing interests. Authors' contributions SS (since deceased) carried out the initial literature review and first draft of manuscript. GP assisted with preparation of the manuscript and photography. NW assisted with the cytologic evaluation and description. MC provided clinical information. SR carried out additional literature review and subsequent drafts of manuscript. SM participated in the coordination of the study and preparation of the final draft of the manuscript. Acknowledgements Ms. Gina Madrid assisted with the typing of the manuscript. ==== Refs Chestnut RM Abitbol JJ Chamberlain M Marshall LF Vertebral collapse with quadriparesis due to metastatic glioblastoma multiforme: Case report and review of the literature J of Neuro-Oncology 1993 16 135 140 10.1007/BF01324700 Wahl RW Fine-needle aspiration of metastatic sacrococcygeal myxopapillary ependymoma Diagn Cytopathol 1986 2 261 3 3533481 Vural G Hagmar B Walaas L Extracranial metastasis of glioblastoma multiforme diagnosed by fine-needle aspiration: A report of two cases and a review of the literature Diagn Cytopathol 1996 15 60 5 8807254 10.1002/(SICI)1097-0339(199607)15:1<60::AID-DC12>3.0.CO;2-A Campora RG Salaverri CO Ramirez FV Villadiego MS Davidson HG Metastatic glioblastoma multiforme in cervical lymph nodes: Report of a case with diagnosis by fine needle aspiration Acta Cytol 1993 37 938 42 8249517 Sunita Kapila K Singhal RM Verma K Extracranial metastasis of anastrocytoma detected by fine-needle aspiration: A case report Diagn Cytopathol 1991 7 290 2 1879267 al-Rikabe AC al-Sohaibani MO Jamjoom A al-Rayess MM Metastatic deposits of a high-grade malignant glioma in cervical lymph nodes diagnosed by fine needle aspiration (FNA) cytology – case report and literature review Cytopathol 1997 8 421 7 Hassenbusch SJ Sawaya R Pietrinigro D Levin VA Intratumoral injection of DT-015 for malignant gliomas: Meeting abstract Asco Annual Meeting 1998 Abstract No 884 Aichholzer M Mazal PR Haberler C Dietrich W Bertalanffy A Roessler K Ungersboeck K Epidural metastasis of a glioblastoma after stereotactic biopsy: case report Min Inv Neurosurg 2001 44 175 7 10.1055/s-2001-18127 Hoffman JH Duffner PK Extraneural metastases of central nervous system tumors Cancer 1985 56 1778 82 4027909 Kumar R Jain R Tandon V Thalamic glioblastoma with cerebrospinal fluid dissemination in the peritoneal cavity Pediatr Neurosurg 1999 31 242 245 10681678 10.1159/000028870	16287502	PMC1325054	CC BY	2021-01-04 16:24:26	no	Cytojournal. 2005 Nov 14; 2:19	utf-8	Cytojournal	2,005	10.1186/1742-6413-2-19	oa_comm
==== Front Nutr JNutrition Journal1475-2891BioMed Central London 1475-2891-4-361635172610.1186/1475-2891-4-36ReviewAdipose energy stores, physical work, and the metabolic syndrome: lessons from hummingbirds Hargrove James L [email protected] Department of Foods and Nutrition, University of Georgia, Athens, GA 30605, USA2005 13 12 2005 4 36 36 14 6 2005 13 12 2005 Copyright © 2005 Hargrove; licensee BioMed Central Ltd.2005Hargrove; 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. Hummingbirds and other nectar-feeding, migratory birds possess unusual adaptive traits that offer important lessons concerning obesity, diabetes and the metabolic syndrome. Hummingbirds consume a high sugar diet and have fasting glucose levels that would be severely hyperglycemic in humans, yet these nectar-fed birds recover most glucose that is filtered into the urine. Hummingbirds accumulate over 40% body fat shortly before migrations in the spring and autumn. Despite hyperglycemia and seasonally elevated body fat, the birds are not known to become diabetic in the sense of developing polyuria (glucosuria), polydipsia and polyphagia. The tiny (3–4 g) Ruby-throated hummingbird has among the highest mass-specific metabolic rates known, and loses most of its stored fat in 20 h by flying up to 600 miles across the Gulf of Mexico. During the breeding season, it becomes lean and maintains an extremely accurate energy balance. In addition, hummingbirds can quickly enter torpor and reduce resting metabolic rates by 10-fold. Thus, hummingbirds are wonderful examples of the adaptive nature of fat tissue, and may offer lessons concerning prevention of metabolic syndrome in humans. ==== Body Recent emphasis on human obesity obscures the fact that fat cells and the triglyceride energy system provide crucial functions in animals as diverse as invertebrate worms (C. elegans)[1], insects including fruit flies (D. melanogaster)[2], bony fish [3], toads (Bufo species), lizards [4], and birds [5]. Plainly, fat tissue did not originate merely as a cause of disease in sedentary people. It is part of an ancient, genetically inherited energy regulatory system in most if not all animal species. In many animals, day length and season strongly affect fat deposition through mechanisms that involve changes in pineal function, activation of the sympathetic nervous system, and changes in sensitivity to peptides such as leptin and neuropeptide Y [6,7]. Studies of seasonal weight gain offer insights into human obesity, and there may be a seasonal component in the development of human obesity in temperate regions[8]. The adaptive value of fat in providing energy for work, reproduction and survival is dramatized in the migratory energetics of the Ruby-throated hummingbird (Archilocus colubris), a bird which is familiar to most people who reside in eastern North America and Central America. The amount of fat (1–2 g) that would allow a human to climb about 50 feet is enough for the Ruby-throat to fly across the Gulf of Mexico, and failure to make the crossing would mean certain death. This paper will review aspects of hummingbird energetics and seasonal weight regulation that may be unfamiliar to students of human obesity. Biology of the Ruby-throated Hummingbird The hummingbird family (Trochilidae) includes some of the smallest and most metabolically active vertebrates, with the Bumblebee hummingbird weighing under 2.0 grams [9]. At 2.5–4.8 g, the adult Ruby-throated hummingbird weighs much less than the common shrew and a little more than a U.S. penny (2.5 g). During mid-summer, females average about 3.3 g compared to 3.0 g for males [9]. Both genders contain an average of about 21% body fat (0.47–0.58 g) when not migrating [10]. The higher body weights are observed just prior to migration when the birds stop nesting and feed actively. The birds gain an additional ~1.7 g of fat and double their percent body fat prior to migration [11]. A. colubris spends the winter in Central America and migrates to North America for the breeding season, going as far north as Ontario, Canada (Fig. 1). The total trip may exceed 2000 miles, and is reversed in the fall. The migrations are timed to coincide with the blossoming of several flowering plants, which provide nectar that fuels much of the journey. Nectars contain as much as 38% (~1 M) sugars (mostly sucrose) [12,13]. After fasting overnight, hummingbirds primarily metabolize free fatty acids and have a respiratory quotient (RQ) of about 0.7. However, the RQ quickly goes to about 1.0 when they begin to feed, indicating oxidation of carbohydrate [14,15]. During the breeding season, males maintain an extremely accurate body mass by ingesting small meals roughly every 15–20 min, and using the energy to court females and chase away other males [16]. Just before nightfall, they consume enough nectar to last the night; when that fails, they may enter torpor to conserve energy [15,17]. Figure 1 Approximate wintering range and breeding range of the Ruby-throated hummingbird. Arrow indicates a probable migratory pathway from Yucatan to the southern U.S. Hummingbirds have one of the highest metabolic rates relative to metabolic body size of any animal on earth. Heart rates up to 1260 beats per minute have been recorded, and breathing rate is about 250 breaths per minute even at rest. Resting body temperatures are about 39°C [9]. At rest, oxygen consumption is about 4 ml O2/g/h [18]. During flight, hummingbird oxygen consumption per gram of muscle tissue is approximately 10 times higher than that seen for elite human athletes [19]. Preparation for migration requires that the birds switch from carbohydrate to fat metabolism during flight, and this entails changes in feeding behavior, energy storage and mitochondrial energy usage. During periods of rapid fattening, hummingbird RQ values are above 1.0, consistent with lipogenesis or at least fat storage. The mechanisms of fattening include increased energy intake, increased food efficiency, altered diet selection, and increased lipogenic enzymes [20]. Preferential metabolism of carbohydrate spares lipids for storage. Hummingbirds do eat insects and they may increase insect consumption prior to migration [21,22]. Hummingbirds [13,23] and many other birds [5,24] maintain very high blood glucose both in the fasted and fed conditions. In hummingbirds, fasted glucose is about 17 mM (300 mg/dl), and it increases to about 42 mM (740 mg/dl) after feeding [23,25]. Although these levels would be classified as diabetic in humans, nectivorous birds do not become diabetic [13,26] in the traditional sense of spilling glucose into the urine with symptoms of polyuria, polydipsia and polyphagia. Also, they do not develop the degree of glycated hemoglobin seen in humans [23]. Migrating birds may become insulin resistant and there may be a parallel to human metabolic syndrome [27]. However, the birds are forced to switch from almost total reliance on carbohydrate to almost total reliance on fatty acid metabolism during migration over oceans or desert terrain that provides no other energy sources. Hummingbird Energetics: Across the Gulf of Mexico on a Gram of Fat During the northward migration, many ruby-throated hummingbirds reach the Gulf of Mexico on the coast of Yucatan. The distance to the U.S. can equal 500–600 miles, and the most direct routes provide no sites at which food or water may be obtained. Whereas some birds may take a coastal route or possibly fly to Cuba, ornithologists believe that most birds fly non-stop across the Gulf of Mexico [9]. Networks of bird-watchers report data on various species during the migrations, and it is noteworthy that arrivals may be reported in Louisiana and Florida prior to arrivals in Texas (Fig. 2). Males arrive in the U.S. prior to the females, and time is critical because the birds compete for habitat. A premium may be placed on early arrival because it provides selective advantage in breeding. Prior to departing, the birds must store enough energy to fly at speeds that range from 25–50 mph. Figure 2 First Spring sightings of Ruby-throated hummingbirds in 2005 occurred during February in Louisiana and Florida (arrows) before the birds had been reported in Texas. Source: Hummingbirds.net . Pearson [28] measured oxygen consumption of 2 species of hummingbird during hovering flight and found values of 68–85 ml O2/g/hr. He calculated a flight range of 385 miles on the assumption that the birds stored 1 g of fat and consumed 80 ml O2/g/hr (caloric equivalent of 4.69 kcal/l) when flying at 50 mph. However, Odum et al [11] showed that Ruby-throated hummingbirds can store more than 40% of body weight as fat, and found a mean content of fat of 2.25 g in birds accidentally killed at television towers. Lasiewski [29] showed that the metabolic rate is probably lower (42 cc.O2/gm/hr) than Pearson estimated. Assuming a flight speed of 25 mph, he estimated that males have a flight range of about 650 miles while females have a range of about 610 miles. The highest estimate of flight range for the Ruby-throated hummingbird is about 2500 km (1500 miles) [11]. If one assumes an average flight weight of 3.5 g and a caloric equivalent of 4.69 kcal/l, then the energy needed for crossing the Gulf of Mexico is about 0.7 kcal per hour. A 20 hour crossing would require about 14 kcal, or 1.5 g of fat (1.8 g of fat tissue assuming that fat tissue contains 7.7 kcal/g). If the birds were to rely on glycogen for this energy, they would need to store about 3.5 g of carbohydrate. Adding 2 g of water of hydration for each g of glycogen, the birds would have to increase their body weight to above 10 g! The flight is possible only because the energy yield per gram of fat is 10 fold higher than for hydrated glycogen [19]. If they relied on glycogen, it is unlikely that the birds could generate sufficient lift to leave Yucatan, much less carry the extra weight across the Gulf of Mexico [22]. Conclusion Migrating birds stay lean until pre-migratory fattening becomes necessary, and then may add fat at a rate of 1–13% of body weight per day [27]. For Ruby-throated hummingbirds, the fattening is crucial for survival. The average "field metabolic rate" is about 8 times resting metabolic rate [30], and during constant flight, hummingbirds expend about 0.7 kcal/hr [18,29]. Glucose at 42 mM distributed in blood and extracellular fluid (20% of body weight) would provide 6 mg of glucose (24 calories or 100 J). This is sufficient only for a few minutes of flight. The birds would quickly shift to metabolism of glycogen and fatty acids to provide adequate energy. Thus, the pre-migratory fattening is purposive, and a typical flight across the Gulf of Mexico will require about 75% of the birds' energy stores (assuming that 1.5 g of fat is used out of ~2.0 g stored). The first lesson that the Ruby-throated hummingbird teaches is that becoming fat can be beneficial if it is necessary as an energy buffer to survive. The second lesson is that fat birds with very high plasma glucose levels do not become diabetic. Part of the preventative mechanism is anatomical and physiological. Nectar feeding birds are unusual in that they consume large amounts of water along with the sugars they typically consume [26]. Birds have a relatively low glomerular filtration rate and are able to reabsorb essentially all of the glucose that is filtered into the urine [26]. It is not clear how they avoid showing symptoms of "glucose toxicity" such as glycated hemoglobin, but the levels of hemoglobin A1c are lower than in humans [23]. One hypothesis could be that the turnover rates of red blood cells and proteins are substantially higher in birds than in mammals. For example, the lifespan of red blood cells in birds can be 21 days or less vs. about 120 days for humans[31], so there may be less opportunity for glycation. Turnover rates for metabolic pools are thought to be proportional to body mass to the 1/4 power [32], which would indicate that metabolic pools exchange about 12 times faster in hummingbirds than in humans (Kleiber, p. 216 and 390). This is congruent with the high ATP turnover rate in active muscle [33]. Whether or not birds avoid obesity and diabetes by dint of their rates of living, the neurobiology and endocrinology of avian fat deposition are complex, and students of migratory birds have suggested that they could offer important clues concerning prevention of obesity and diabetes in humans [26,27]. Diabetes and the metabolic syndrome are rightfully considered to be kinetic disorders that do not develop unless several major controls fail. Typically, sensitivity to insulin and to glucose ("glucose effectiveness") both diminish, creating insulin resistance [34]. Hepatic glucose production may continue (instead of shutting off) even if plasma glucose and insulin are both elevated. The rate of production of insulin by the beta cell must also fail to compensate for the decreased sensitivity [34]. These are conditions that come about in humans because of sedentary habits and obesity [35]. In migrating birds, there may be a decline in insulin sensitivity, but it is unlikely that regulation of beta cell function or hepatic glucose production becomes abnormal. Hummingbirds combat kinetic disorders by dint of their highly aerobic lifestyles and necessity of maintaining close feedback between energy intake and energy expenditure. Unlike humans who have "uncoupled" food intake from functional needs, animals that must flap their wings at 50 beats per second in order to feed have a hard time staying fat. Competing interests The author(s) declare that they have no competing interests. Authors' contributions This review was prepared entirely by Dr. Hargrove. Acknowledgements The author thanks Dr. Cameron B. Kepler (Athens, GA) for discussing the energetics of avian migration. The work was supported by funds from Hatch Project GEO00922, US Agricultural Experiment Station. ==== Refs McKay RM McKay JP Avery L Graff JM C elegans: a model for exploring the genetics of fat storage Dev Cell 2003 4 131 142 12530969 10.1016/S1534-5807(02)00411-2 Hwangbo DS Gershman B Tu MP Palmer M Tatar M Drosophila dFOXO controls lifespan and regulates insulin signalling in brain and fat body Nature 2004 429 562 566 15175753 10.1038/nature02549 Albalat A Liarte C MacKenzie S Tort L Planas JV Navarro I Control of adipose tissue lipid metabolism by tumor necrosis factor-alpha in rainbow trout (Oncorhynchus mykiss) J Endocrinol 2005 184 527 534 15749811 10.1677/joe.1.05940 Migliorini RH Lima-Verde JS Machado CR Cardona GM Garofalo MA Kettelhut IC Control of adipose tissue lipolysis in ectotherm vertebrates Am J Physiol 1992 263 R857 62 1329567 Bairlein F Simons D Nutritional Adaptations in Migrating Birds Israel Journal of Zoology 1995 41 357 367 Bartness TJ Demas GE Song CK Seasonal changes in adiposity: the roles of the photoperiod, melatonin and other hormones, and sympathetic nervous system Exp Biol Med (Maywood) 2002 227 363 376 12037125 Adam CL Mercer JG Appetite regulation and seasonality: implications for obesity Proc Nutr Soc 2004 63 413 419 15373951 10.1079/PNS2004367 Visscher TL Seidell JC Time trends (1993-1997) and seasonal variation in body mass index and waist circumference in the Netherlands Int J Obes Relat Metab Disord 2004 28 1309 1316 15314624 10.1038/sj.ijo.0802761 Johnsgard PA The Hummingbirds of North America 1997 2nd Washington, D.C., Smithosonian Institution Press 278 pp Connell CE Odum EP Kale H Fat-free weights of birds Auk 1960 77 1 9 Odum EP Connell CE Stoddard Flight Energy And Estimated Flight Ranges Of Some Migratory Birds Auk 1961 78 515 527 Galetto L Bernardello G Floral nectaries, nectar production dynamics and chemical composition in six ipomoea species (convolvulaceae) in relation to pollinators Ann Bot (Lond) 2004 94 269 280 15229123 10.1093/aob/mch137 McWhorter TJ Martinez del Rio C Food ingestion and water turnover in hummingbirds: how much dietary water is absorbed? J Exp Biol 1999 202 (Pt 20) 2851 2858 10504321 Suarez RK Lighton JR Moyes CD Brown GS Gass CL Hochachka PW Fuel selection in rufous hummingbirds: ecological implications of metabolic biochemistry Proc Natl Acad Sci U S A 1990 87 9207 9210 2251266 Powers DR Diurnal-Variation in Mass, Metabolic-Rate, and Respiratory Quotient in Anna and Costa Hummingbirds Physiological Zoology 1991 64 850 870 Calder WA Calder LL Frazier TD The hummingbird's restraint: a natural model for weight control. Experientia 1990 46 999 1002 10.1007/BF01940653 Hiebert S Seasonal-Changes in Body-Mass and Use of Torpor in a Migratory Hummingbird Auk 1993 110 787 797 Lasiewski RC Oxygen consumption of torpid, resting, active and flying hummingbirds. Physiological Zoology 1963 36 122 140 Suarez RK Hummingbird Flight - Sustaining the Highest Mass-Specific Metabolic Rates among Vertebrates Experientia 1992 48 565 570 1612136 10.1007/BF01920240 Bairlein F Gwinner E Nutritional mechanisms and temporal control of migratory energy accumulation in birds Annual Review of Nutrition 1994 14 187 215 7946517 10.1146/annurev.nu.14.070194.001155 Yanega GM Rubega MA Feeding mechanisms: Hummingbird jaw bends to aid insect capture Nature 2004 428 615 15071586 10.1038/428615a Altshuler DL Dudley R The ecological and evolutionary interface of hummingbird flight physiology J Exp Biol J Exp Biol 2002 205 2325 2336 Beuchat CA Chong CR Hyperglycemia in hummingbirds and its consequences for hemoglobin glycation Comp Biochem Phys A Comp Biochem Phys A 1998 120 409 416 Bairlein F Seasonal-Variations of Serum Glucose-Levels in a Migratory Songbird, Sylvia-Borin Comp Biochem Phys A Comp Biochem Phys A 1983 76 397 399 10.1016/0300-9629(83)90345-6 Beuchat CA Chong CR Hyperglycemia in hummingbirds: Implications for hummingbird ecology and human health Faseb Journal 1997 11 529 529 McWhorter TJ Martinez del Rio C Pinshow B Roxburgh L Renal function in Palestine sunbirds: elimination of excess water does not constrain energy intake J Exp Biol 2004 207 3391 3398 15326215 10.1242/jeb.01169 Bairlein F How to get fat: nutritional mechanisms of seasonal fat accumulation in migratory songbirds Naturwissenschaften 2002 89 1 10 12008967 10.1007/s00114-001-0279-6 Pearson O The metabolism of hummingbirds. Condor 1950 52 145 1152 Lasiewski RC The energetics of migrating hummingbirds. The Condor 1964 64 324 Powers DR Conley TM Field metabolic-rate and food-consumption of 2 sympatric hummingbird species in southeastern Arizona Condor 1994 96 141 150 Kalomenopoulou M Koliakos G Total body haematocrit iron kinetics and erythrocyte life span in pigeons (Columba livia) Comp Biochem Physiol A 1989 92 215 218 2566412 10.1016/0300-9629(89)90156-4 Kleiber M The Fire of Life, An Introduction to Animal Energetics 1975 New York: Huntington, Robert Kreiger 453 Suarez RK Upper limits to mass-specific metabolic rates Annual Review of Physiology 1996 58 583 605 8815810 10.1146/annurev.ph.58.030196.003055 Bergman RN Watanabe R Rebrin K Ader M Steil G Toward an integrated phenotype in pre-NIDDM Diabet Med 1996 13 S67 77 8894486 Esparza J Fox C Harper IT Bennett PH Schulz LO Valencia ME Ravussin E Daily energy expenditure in Mexican and USA Pima indians: low physical activity as a possible cause of obesity Int J Obes Relat Metab Disord 2000 24 55 59 10702751 10.1038/sj.ijo.0801085	16351726	PMC1325055	CC BY	2021-01-04 16:23:37	no	Nutr J. 2005 Dec 13; 4:36	utf-8	Nutr J	2,005	10.1186/1475-2891-4-36	oa_comm
==== Front Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1361628588610.1186/1465-9921-6-136ResearchMembrane TNF confers protection to acute mycobacterial infection Fremond Cecile [email protected] Nasiema [email protected] Ivy [email protected] Sergei I [email protected] Vladimir [email protected] Valerie FJ [email protected] Muazzam [email protected] Bernhard [email protected] Molecular Immunology and Embryology, Centre National de la Recherche Scientifique, Orléans, France2 Department of Immunology, Institute of Infectious Disease and Molecular Medicine, University of Cape Town, South Africa3 Laboratory of Molecular Immunoregulation, Center for Cancer Research, National Cancer Institute-Frederick, Fort-Detrick, Frederick, MD 21702, USA4 Max Planck Institute for Infection Biology, Department of Immunology, Schumannstrabe 21/22, 10117 Berlin, Germany2005 14 11 2005 6 1 136 136 16 6 2005 14 11 2005 Copyright © 2005 Fremond et al; licensee BioMed Central Ltd.2005Fremond 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 Tumour necrosis factor (TNF) is crucial for the control of mycobacterial infection as TNF deficient (KO) die rapidly of uncontrolled infection with necrotic pneumonia. Here we investigated the role of membrane TNF for host resistance in knock-in mice with a non-cleavable and regulated allele (mem-TNF). Methods C57BL/6, TNF KO and mem-TNF mice were infected with M. tuberculosis H37Rv (Mtb at 100 CFU by intranasal administration) and the survival, bacterial load, lung pathology and immunological parameters were investigated. Bone marrow and lymphocytes transfers were used to test the role of membrane TNF to confer resistance to TNF KO mice. Results While TNF-KO mice succumbed to infection within 4–5 weeks, mem-TNF mice recruited normally T cells and macrophages, developed mature granuloma in the lung and controlled acute Mtb infection. However, during the chronic phase of infection mem-TNF mice succumbed to disseminated infection with necrotic pneumonia at about 150 days. Reconstitution of irradiated TNF-KO mice with mem-TNF derived bone marrow cells, but not with lymphocytes, conferred host resistance to Mtb infection in TNF-KO mice. Conclusion Membrane expressed TNF is sufficient to allow cell-cell signalling and control of acute Mtb infection. Bone marrow cells, but not lymphocytes from mem-TNF mice confer resistance to infection in TNF-KO mice. Long-term infection control with chronic inflammation likely disrupting TNF mediated cell-cell signalling, additionally requires soluble TNF. Mycobacterium tuberculosis H37Rvmembrane TNFTNF-deficiencyT cell recruitmentgranuloma ==== Body Background Protective immunity to M. tuberculosis (Mtb) infection is regulated by T cells, macrophages and cytokines including IFNγ, IL-12 and TNF [1,2]. IFNγ derived from T and NK cells has been shown to be essential, as mice with a disruption of the IFNγ signalling are unable to restrict the growth of M. tuberculosis and succumb to the infection [3-6]. A critical role for TNF in mycobacterial defence was inferred from neutralisation and gene deletion experiments in mice [7-10]. TNF neutralising therapies for rheumatoid arthritis and Crohn's disease turned out to increase the risk of developing tuberculosis (TB) and other opportunistic infections [11-14]. TNF produced by macrophages, and a variety of other cells, is a major regulator of inflammation and leukocyte trafficking [15,16]. Although soluble TNF in controlling intracellular bacterial infections is uncontested, the function of membrane TNF, which is subsequently cleaved by the metalloproteinase-disintegrin TACE (TNFα converting enzyme) [17] into the secreted trimeric TNF, is not established. Several biological functions of membrane TNF have been described, such as strong cytotoxity, polyclonal activation of B cells, induction of IL-10 by monocytes, ICAM-1 expression on endothelial cells and regulation of chemokine expression [18,19]. The transgenic overexpression of membrane TNF (mem-TNF) demonstrated an in vivo role in the control of Listeria and mycobacterial infection [20,21]. However these studies have been performed on TNF/lymphotoxin (LT) deficient background, and since lymphotoxin is implicated into TB resistance, do not address the role of membrane TNF only [20,21]. The recent generation of a mouse with functional, normally regulated and expressed membrane-bound TNF, obtained by knocking-in an uncleavable Δ1–9,K11E TNF allele, represents a major advance and allowed interesting insights in the role of membrane TNF in lymphoid structure maintenance and inflammation [19]. Here we asked whether membrane TNF is sufficient for containing Mtb infection. We compared the host resistance to acute mycobacterial infection in mem-TNF mice, eg. Δ1–9,K11E TNF knock-in mice [19], and TNF-deficient (TNF-KO) mice [22]. We show that membrane TNF substitutes soluble TNF to recruit and activate macrophages and T cells, to generate granuloma and control acute infection, but is insufficient to control the chronic phase of infection. Transfer of bone marrow cells, but not of lymphocytes from mem-TNF and WT mice was able to confer resistance to infection in TNF KO mice. Methods Mice Mem-TNF [19] and TNF-KO mice [22] on a C57BL/6 background and C57BL/6 mice were bred in house. For experiments, adult (8–15 week old) animals were kept in sterile isolators in a biohazard animal unit. All animal experiments complied with the French Government's ethical and animal experiment regulations. Bacteria and infection Pulmonary infection with M. tuberculosis H37Rv (Pasteur) was performed by delivering 100 bacteria into both nasal cavities (20 μl each) under xylazine-ketamine anaesthesia as described [23]. The bacterial load in the lung was determined at day 1 post infection. Three independent experiments were conducted, one short term for 90 days, a long-term study over 240 days (n = 6 per group) and a second long-term study over 200 days (n = 10 mice per group). Bacterial load in tissues and histological investigation Bacterial loads in organs of infected mice were evaluated at different time points after infection with M. tuberculosis H37Rv as described [23]. For histological analysis lungs were fixed in 4% phosphate buffered formalin, paraffin-embedded as described, and stained with haematoxylin and eosin and a modified Ziehl-Neelsen [10]. FACS analysis of infiltrating cells from infected lung FACS analysis of inflammatory cells from infected lung was performed as described [23,24]. Rat anti-mouse CD4-PerCP (clone RM4-5), CD8-FITC (clone 53-6.7), Ly6G-PE (clone RD6-8C5), CD11b-PE (clone M1/70), I-A/I-E- FITC (clone 2G9) were from BD Pharmingen (San Diego, CA) and stained cells were analyzed by flow cytometry on a LSR analyser (Becton Dickinson). Cytokine determination IL-12p40 and IFNγ were quantified using commercial ELISA (Duoset, R&D Systems, Abingdon, UK). Bioactive TNF was assessed using the WEHI 164 cells based bioassay [25]. Reconstitution of irradiated mice with lymphocytes or bone marrow cells of TNF KO mice Haemopoietic reconstitution of bone marrow was performed as described [26]. Briefly, designated recipient mice received an optimised lethal total body irradiation dose of 8 Grey using a γ-irradiation source (CHRO, Orleans). Irradiated mice were reconstituted with 2 × 106 fresh unseparated bone marrow cells by intravenous injection in the lateral tail vein (n = 6 per group). Mice were left to fully reconstitute for at least three month prior to infection, reconstitution was verified by the analysis of the haematogram. In addition, freshly isolated splenic T lymphocytes (107 i.v.) from WT and mem-TNF mice were transferred into TNF KO mice (exposed to 4 Grey γ-irradiation) immediately before Mtb infection and survival was followed (n = 6 per group). The experiments were repeated once. Primary macrophage and dendritic cells cultures Murine bone marrow cells were isolated from femurs and differentiated into macrophages using 20% horse serum and 30% L929 cell-conditioned medium as a source of M-CSF [26]. Alternatively, murine bone marrow cells were differentiated into myeloid dendritic cells using 4% J558L cell-conditioned medium as a source of GM-CSF [27]. Stimulation of macrophages and dendritic cells Bone marrow derived macrophages (BMDM) and dendritic cells (BMDC) were plated (at 105 cells/well) and stimulated with LPS (Escherichia coli, serotype O111:B4, Sigma, St Louis, MO, at 100 ng/ml) or M. tuberculosis H37Rv (heat-killed 40 min at 80°C; 2 bacteria per cell), or infected with M. bovis BCG (Pasteur Institute, at a MOI of 2 bacteria per cell) as described before [28]. Cell supernatants were harvested after 24 h of stimulation in the presence of IFNγ (100 U/ml) for TNF and IL-12 p40 quantification, and nitrite measurements by Griess reagents [29]. Membrane expressed CD40 and CD86 staining was performed as described [28] using CD40-PE (clone 3/23) and CD86-FITC (clone GL1, BD PharMingen San Diego, CA). The mean fluorescence of non-stimulated and activated macrophages was compared. Antigen-specific IFNγ production T cell priming was assessed by the production of IFNγ upon antigen restimulation ex vivo as described [23]. Single cell suspension of splenocytes and mediastinal lymph nodes were prepared from mice 4 weeks after infection. Cells were stimulated with either 2.5 μg/ml Con A (Sigma), a lyophilised soluble fraction from BCG culture supernatant (SupBCG, 10 μg/ml), or heat killed Listeria monocytogenes (100 bacteria per cell) for 3 days at 37°C as described [28]. IFNγ production in the supernatant was quantified by ELISA. Statistical Analysis ANOVA was used for the analysis of the early data points for the comparison of three experimental groups. For the data points beyond one month we used the Student's t test. Survival data (Kaplan Meier plots) were compared using a log rank test for the comparison of the groups. A p value <0.05 was considered statistically significant. Results Absence of secreted TNF in mem-TNF mice To confirm that mem-TNF mice do not secrete TNF, serum was obtained 90 min after LPS (100 μg) injection. In contrast to WT mice TNF was undetectable in the sera of both TNF-KO and mem-TNF mice (not shown). While bone marrow derived macrophages (BMDM) secrete high TNF levels in response to M. bovis BCG and Mtb infection or to LPS, TNF is undetectable in culture supernatants of BMDM from TNF-KO and strongly defective in mem-TNF mice (Fig 1A), the latter however express TNF on the membrane of activated macrophages and T-cells as shown before [19]. Interestingly, the production of IL-12p40 was augmented in TNF-KO as compared to WT macrophage, but normalised in mem-TNF macrophages (Fig 1B). Therefore, membrane-bound TNF is sufficient to normalise IL12p40 production by macrophages, while complete absence of TNF results in deregulated IL12p40 expression as described before [30]. The production of IL-6 and nitrites was essentially unaffected in stimulated mem-TNF macrophages, as compared to WT or TNF-KO macrophages (data not shown). Second, mycobacteria-induced expression of costimulatory molecules CD40 and CD86 was investigated in BMDM and BMDC. CD40 and CD86 upregulation was induced to a comparable extent in macrophages (Fig 1C,D) and dendritic cells (data not shown) in all groups as compared to unstimulated controls indicating that the expression of costimulatory molecules is TNF independent. Taken together, these observations confirm the defective secretion of functional TNF in vivo and in vitro in mem-TNF mice, while the expression IL-12 and costimulatory molecules were unaffected. Figure 1 Defective soluble TNF and augmented IL-12 production from mycobacterial stimulated macrophages, and normal upregulation of CD40 and CD86 costimulatory molecules. Mycobacteria induced TNF (A) and IL-12p40 (B) production at 24 h TNF in the supernatant of macrophages from WT (open bars), TNF-deficient (black bars) and mem-TNF mice (grey bars) infected with BCG or H37Rv (at a MOI of 2) or stimulated with LPS (100 ng/ml). Data are expressed as the mean ± SD (n = 2 mice from one out of 4 independent experiments). TNF independent expression of costimulatory molecules CD40 (C) and CD86 (D) analysed by flow cytometry in macrophages stimulated as above. Results are expressed as mean percentage. Membrane TNF is sufficient to control acute M. tuberculosis infection To ascertain the role of soluble vs membrane-bound TNF in resistance to tuberculosis, we compared mem-TNF, TNF-KO mice and WT mice infected with 100 CFU Mtb given by intranasal administration. Within 4–5 weeks of infection TNF-KO mice displayed rapid weight loss (not shown), impeded locomotor activity and succumbed to infection rapidly (Fig 2A). Mem-TNF mice had normal weight development and survived the 90 days observation period. Therefore, we conclude that membrane bound TNF confers substantial protection to acute Mtb infection. Figure 2 Necrotic pneumonia and uncontrolled infection in TNF-KO mice, but not in mem-TNF mice at 30 days after Mtb infection (100 CFU, intranasal administration). Survival (A), relative lung weights (B), macroscopic lung changes (C), and bacterial load at 30 days in the lungs (D). The results are expressed as the mean ± SD (n = 6 per group) and are representative of two independent experiments. The bacterial load at day 1 upon infection was 84 ± 28 CFU per lung (n = 6). Microscopic investigations of the lungs show confluent neutrophil and mononuclear cell inflammation with extensive, confluent necrosis and abundant bacilli in TNF-KO mice, while mem-TNF and WT mice show focal, largely perivascular mononuclear cell infiltration (E). Representative haematoxylin and eosin stained lung sections at low (50×) and high power (200×) are shown (n = 5 per group). The relative lung weights -a surrogate marker of inflammation- of surviving TNF-KO mice at 30 days were increased (p < 0.01), but not in WT and only slightly in mem-TNF mice (Fig 2B). Lungs of TNF-KO mice displayed large nodular and confluent lesions on the pleura at 30 days post infection, which were much smaller and more discrete in mem-TNF mice, comparable to those in WT mice (Fig. 2C). We then determined the mycobacterial load in the lungs. At 30 days WT mice displayed a 4 log increase of viable mycobacteria reaching 106 CFU counts in the lungs (Fig 2D), which stabilises thereafter, as shown before [24]. By contrast, bacillary burden in the lung was significantly increased in TNF-KO mice (p < 0.01) reaching a value of 109 CFU, while mem-TNF had comparable CFU values as WT mice (Fig 2D). Mtb dissemination measured by splenic and hepatic bacterial loads was also significantly increased in TNF-KO mice, but not in mem-TNF mice (data not shown). Therefore, acute Mtb infection is controlled in the presence of membrane TNF only. The establishment of granulomas is the manifestation of a vigorous cell mediated immune response, which is crucial for inhibiting mycobacterial growth and depends on TNF [31]. We asked whether membrane bound TNF is sufficient for granuloma formation upon Mtb infection. At 30 days post infection the lungs of mem-TNF mice displayed well-defined granulomatous lesions that were characterised by foamy epitheloid like macrophages with surrounding and interspersed and perivascular lymphocytic infiltration, similar to the granuloma structures formed in WT mice (Fig 2E). TNF-KO mice had abundant inflammatory cells with extensive necrosis, and no structured granuloma. Therefore, membrane TNF is sufficient for granuloma formation and infection control. Membrane TNF allows lymphocyte recruitment and cell activation upon mycobacterial infection In view of the granulomatous response elicited in mem-TNF mice after Mtb infection we investigated whether membrane TNF could affect the cellular recruitment in the lungs. Single cell suspension of lung infiltrating cells was obtained at 30 days post infection. CD4 and CD8 lymphocytes were significantly higher in TNF-KO than in WT mice (p < 0.01), in line with the severe pathology in these mice, while the sole presence of mem-TNF prevented the augmented lymphocyte recruitment as shown in mem-TNF mice (Fig. 3A,B). Enhanced cell mediated immune responses have been shown in the complete absence of TNF [32] and is probably linked to uncontrolled bacterial growth in TNF-KO mice, leading to excessive inflammation. We show that the presence of membrane TNF corrects the TNF-deficient phenotype, i.e. cellular recruitment to the infected lungs was normal. Concomitant with increased lymphocyte recruitment IFNγ levels in lung homogenates from TNF-KO mice were significantly increased (p < 0.01) as compared to WT mice (Fig. 3C). By contrast, lungs from mem-TNF mice showed comparable amounts of IFNγ as WT mice. Figure 3 Augmented CD4 and CD8 T cell recruitment, pulmonary IFNγ production and antigen-specific T cell response in infected TNF-KO mice are corrected in mem-TNF mice. Recruitment of CD4 (A) and CD8 T cells (B) in the lung of WT, mem-TNF and TNF-KO mice infected with Mtb (100 CFU). Lymphocytes were obtained form lungs from Mtb infected mice at 30 days post infection as described in Materials and Methods. IFNγ levels in lung homogenates at 30 days of infection (C): Increased IFNγ levels in TNF KO mice, which were lower and comparable in mem-TNF and WT mice. Antigen-specific IFNγ production by SupBCG restimulated splenocytes from mem-TNF and TNF-KO mice, but augmented response in TNF-KO splenocytes (D). Data are representative of two independent experiment (n = 4 mice, mean ± SD). We then asked whether a mycobacteria-specific T cell response is acquired in mem-TNF and TNF-KO mice. Ex vivo restimulation 30 days after infection of splenic lymphocytes (Fig. 3D) with mycobacterial antigens (SupBCG), but not with an irrelevant antigen, induced IFNγ secretion in cells from both TNF-KO and mem-TNF mice. Although IFNγ levels were significantly higher in TNF-KO as compared with WT (p < 0.01), the IFNγ levels of mem-TNF mice did not differ from those of infected WT mice. These data suggest that membrane TNF allows substantial, but controlled recruitment and activation of T cells and macrophages resulting in mycobactericidal effector mechanisms. The acquisition of an antigen specific immune response as assessed by the production of IFNγ is TNF independent. Membrane TNF is unable to confer long-term protection against Mtb infection While acute infection was controlled by the expression of membrane TNF (Fig. 2), absence of soluble TNF might affect the control of chronic infection, as we have shown in other conditions [24]. Conversely, in TLR2 [33] or TLR4 deficient mice [34] acute, but not chronic infection is controlled. We therefore conducted a long-term infection study in mem-TNF mice and compared it to WT mice infected with 100 CFU (intranasal route), a condition where TNF-KO mice die within 4–5 weeks (Fig. 2A) [10,23]. While mem-TNF mice appeared healthy over four months, they started to loose body weight and succumbed to infection between 130 – 170 days (Fig. 4A). At 112 days, the relative lung and spleen weights as indicator of inflammation and pathology were significantly increased in mem-TNF mice (Fig. 4B) and viable mycobacteria in lungs were almost two logs higher (Fig. 4C), reaching 107 CFU, a bacterial load still much lower than that observed in moribund TNF-KO mice after 30 days of infection (Fig. 2D). Bacterial dissemination was also increased (Fig. 4D). Mycobacterial cultures obtained from lungs from moribund mem-TNF mice sacrificed at later time points indicated a further increase of CFU in the lungs (not shown). The microscopic investigation revealed a more abundant macrophages and lymphocyte infiltration in mem-TNF mice with confluent foci and less defined granulomatous lesions at 112 days (Fig. 4E). Therefore, lack of soluble TNF may allow slow growth of mycobacteria during the chronic infection, with likely progressive hypoxemia due to chronic pneumonia, leading to death in the chronic phase of infection. Figure 4 Mem-TNF mice succumb of uncontrolled chronic Mtb infection after 4–6 months. Survival (A) of infected mice (Mtb 100 CFU given by intranasal route), data two independent experiments (n = 16 per group). Organs weights (B) and bacterial loads (CFU) in lungs (C) and spleen (D) were determined at 112 days after infection. The results are expressed as the mean ± SD (n = 6 per group). Microscopic changes of the lung from mem-TNF and WT mice at 112 days (E). Microscopic analysis of the lung reveal a more abundant macrophages and lymphocyte infiltration in mem-TNF mice with confluent foci and less defined granulomatous lesions. Representative haematoxylin and eosin stained lung sections at low (50×) and high power (200×) are shown (n = 6 per group). Reconstitution of TNF deficiency by bone marrow transplantation First we attempted to confer resistance in irradiated TNF KO mice to acute Mtb infection by the transfer of lymphocytes from mem-TNF mice. Neither lymphocytes from mem-TNF nor WT mice increased survival in TNF KO mice (data not shown), which is at variance with the recent data from Saunders et al. [37]. Since TNF derived from hemopoietic cells likely contributes most of bioactive TNF [35] and we have shown before that bone marrow transplantation confers resistance to TNF-KO mice to BCG infection [36], we asked whether membrane expressed TNF on haemopoietic cells might be sufficient to correct the susceptibility to Mtb infection of TNF-KO mice. Lethally irradiated TNF-KO mice reconstituted with bone marrow cells from mem-TNF mice controlled Mtb infection as demonstrated by survival over three months, while TNF-KO mice succumbed to acute necrotic pneumonia within 40 days (Fig. 5A). Similarly, irradiated TNF-KO reconstituted with WT bone marrow survived the whole experiment. Although TNF-KO mice could not contain the infection, reconstituted TNF-KO mice developed granuloma (data not shown) and were able to control acute infection and bacterial growth in the lungs (Fig. 5B). Therefore, bone marrow derived cells expressing membrane TNF, but not lymphocytes, are sufficient to control infection in reconstituted TNF-KO mice. Figure 5 Bone marrow cells from mem-TNF mice correct the heightened susceptibility of TNF deficient mice. Long-term survival of TNF-KO mice reconstituted with bone marrow from mem-TNF or WT mice (A), and control of infection in the lung (B) of bone marrow reconstituted TNF-KO mice (p < 0.01). Lethally irradiated (8 Grey) and reconstituted TNF-KO were exposed to infection (Mtb 100 CFU, intranasal administration) 4 months after irradiation and bone marrow reconstitution. The results are expressed as mean values ± SD (n = 6). Discussion We report that membrane TNF plays a crucial role in the control of mycobacterial infection using a knock-in mouse model where the endogenous TNF allele was replaced by a non-cleavable membrane TNF (mem-TNF) mutated in the TACE cleavage site [19]. The expression of mutated membrane TNF confers protection against acute mycobacterial infection with initial control of mycobacterial growth and normal granuloma development in the lung. However, long-term control of infection appears to be dependent additionally on soluble TNF, as mem-TNF mice eventually succumb to chronic infection as shown by a recent contribution [37]. However, we show here significant differences such as uncontrolled infection in the late phase with disseminating infection and the capacity of membrane expressing cells from the bone marrow to confer long-term -and not transient- resistance to Mtb infection in TNF deficient recipient mice. A critical role of TNF for the effective control and resolution of mycobacterial infection has been demonstrated previously [7,10,24], which is mediated by TNFR1 [9], rather than TNFR2 signalling [38]. TNF provided by recombinant BCG expressing TNF may reconstitute granuloma formation and host response in TNF-KO, but not in TNFR1-KO mice, demonstrating the critical role of TNF and TNFR1 signalling [39]. Overexpressed membrane TNF conferred partial resistance to Listeria and mycobacterial infection in transgenic mice on a TNF deficient [20] or on a TNF-LTα double deficient background [21,40]. However, transgenic expression of mem-TNF results in artificially high, non-regulated and non-selective expression of membrane TNF which may make conclusions on the physiological function difficult. While Saunders et al [37] using the same genetic model demonstrate death during the chronic phase, we show here that the bacterial load in lung and spleen is augmented significantly within 3 month of infection with disseminated infection (Fig 4C,D) and increases further in the mice dying with necrotic pneumonia. The mechanism how membrane TNF confers protection may be due to cell-to-cell contacts of T cells, macrophages and other cells and needs further investigations. Several biological functions of membrane TNF have been reported previously and a preferential TNFR2 signalling has been suggested in vitro [18], while in vivo both TNFR1 and TNFR2 were reported to contribute to the signalling by membrane bound TNF [21,41-43]. Since TNFR1 is crucial for the resistance against TB infection, our study suggests memTNF -> TNFR1 signalling as important for host defence. Furthermore, membrane TNF has been shown to be involved in reverse (outside-to-inside) signalling. Upon ligation of its receptor mem-TNF expressing cells are activated to express E-selectin [44]. Thus, membrane TNF at least in T cells might function as a bipolar positive regulator of inflammation either transmitting signals as a ligand to target cells or receiving signals through membrane TNF itself into T cells. Membrane TNF is sufficient for the development of granulomas. In the absence of TNF there was no formation of well-defined granulomas [7,36]. Neutralisation of TNF by antibodies, which likely affects also membrane TNF, leads to the disruption of established granulomas and uncontrolled infection [31]. Activated macrophages expressing iNOS in granuloma are critical for bacterial killing [45-47]. iNOS expression is reduced in lungs of infected TNF-KO mice as compared to mem-TNF and WT mice (data not shown), which has been confirmed [21,40]. Further, the transfer of bone marrow cells expressing mem-TNF is sufficient to correct the increased susceptibility of TNF-KO mice, which clearly points to hemopoietic cells being critically involved in host resistance, and corroborates our previous data using WT bone marrow cells to correct the TNF-KO phenotype [36]. By contrast, we were unable to confer host resistance to TNF KO mice by passive transfer of mature lymphocytes as reported recently [37]. Therefore, sustained membrane expressed TNF on myeloid and lymphoid cells may be necessary to control infection. Interestingly, lymphocyte recruitment was lower and comparable in mem-TNF mice and WT mice (Fig. 3A,B) unlike in the absence of TNF. Augmented lymphocyte recruitment with expansion of activated CD4 and CD8 T cells in the complete absence of TNF has been reported before [32]. Since the expression of costimulatory molecules (Fig. 1C,D) was normal in the absence of TNF, we tested the adaptive immune response. Lymphocytes form infected mem-TNF mice had normal antigen-induced INF-γ response, comparable to that of WT mice (Fig. 3D), suggesting a normal adaptive immune response, while the response was augmented in TNF deficient T cells, in line with previous findings [32]. CD4+ T-cells are critical for cell mediated immunity [1], but not sufficient as shown before for MyD88 deficiency [23]. Finally, the question arises why mem-TNF mice succumb in the chronic phase of infection. It may be hypothesized that membrane expressed TNF activates macrophages in the absence of soluble TNF to a certain level, but the killing may not be as effective as in WT macrophages. Secreted TNF and hence distal signalling is likely required in the chronic phase where direct cell contact between T cells and macrophages may be more difficult to achieve in the chronic inflammatory and fibrotic tissue and therefore soluble TNF may be more efficient to maintain the activation of the innate host defence. Conclusion Membrane expressed TNF allows cell-cell signalling and control of acute Mtb infection. Bone marrow cell reconstitution, but not lymphocyte transfer from mem-TNF mice confer resistance to infection in TNF-KO mice. Long-term infection control, however, with chronic inflammation disrupting TNF mediated cell-cell signalling, requires additionally soluble TNF. The data are of clinical significance, as neutralising therapies used to treat patients with rheumatoid arthritis may reactivate latent TB infection [11]. TNF blockers have different capability to inactivate TNF, anti-TNF antibodies (infliximab, remicade) binding both soluble and membrane bound TNF, whereas soluble TNFR2 (etanercept) binding preferentially soluble TNF [14]. Therefore novel strategies to disrupt more selectively the TNF – TNF-R interaction sparing mem-TNF may be associated with less infectious complications. Abbreviations Mtb M. tuberculosis, CFU colony forming unit, mem-TNF membrane bound TNF Competing interests The author(s) declare that they have no competing interests. Authors' contributions CF and NA were driving the project by designing the protocol and conducting the infectious protocol together with ID and MJ. SG was advising on the genetic mouse model, PCR control and cell transfer experiments. VQ with VY conducted the in vitro cell culture experiments with FACS analysis. BR with MJ was responsible for the overall design and control of the studies. Acknowledgements Dr. Jonathan Sedgwick provided the mem-TNF mice for this study. This work was supported by Le Studium, Orleans, the Wellcome Trust, MRC, NRF and Harry Crossley Fund 435873, Cape Town, and CNRS, Fondation de la Recherche Medicale, and the Ligue contre le Cancer, France. ==== Refs Flynn JL Chan J Immunology of tuberculosis Annu Rev Immunol 2001 19 93 129 11244032 10.1146/annurev.immunol.19.1.93 North RJ Jung YJ Immunity to tuberculosis Annu Rev Immunol 2004 22 599 623 15032590 10.1146/annurev.immunol.22.012703.104635 Cooper AM Roberts AD Rhoades ER Callahan JE Getzy DM Orme IM The role of interleukin-12 in acquired immunity to Mycobacterium tuberculosis infection Immunology 1995 84 423 432 7751026 Cooper AM Magram J Ferrante J Orme IM Interleukin 12 (IL-12) is crucial to the development of protective immunity in mice intravenously infected with mycobacterium tuberculosis J Exp Med 1997 186 39 45 9206995 10.1084/jem.186.1.39 Cooper AM Kipnis A Turner J Magram J Ferrante J Orme IM Mice lacking bioactive IL-12 can generate protective, antigen-specific cellular responses to mycobacterial infection only if the IL-12 p40 subunit is present J Immunol 2002 168 1322 1327 11801672 Flynn JL Chan J Triebold KJ Dalton DK Stewart TA Bloom BR An essential role for interferon gamma in resistance to Mycobacterium tuberculosis infection J Exp Med 1993 178 2249 2254 7504064 10.1084/jem.178.6.2249 Bean AG Roach DR Briscoe H France MP Korner H Sedgwick JD Britton WJ Structural deficiencies in granuloma formation in TNF gene-targeted mice underlie the heightened susceptibility to aerosol Mycobacterium tuberculosis infection, which is not compensated for by lymphotoxin J Immunol 1999 162 3504 3511 10092807 Ehlers S Kutsch S Ehlers EM Benini J Pfeffer K Lethal granuloma disintegration in mycobacteria-infected TNFRp55-/- mice is dependent on T cells and IL-12 J Immunol 2000 165 483 492 10861087 Flynn JL Goldstein MM Chan J Triebold KJ Pfeffer K Lowenstein CJ Schreiber R Mak TW Bloom BR Tumor necrosis factor-alpha is required in the protective immune response against Mycobacterium tuberculosis in mice Immunity 1995 2 561 572 7540941 10.1016/1074-7613(95)90001-2 Jacobs M Brown N Allie N Ryffel B Fatal Mycobacterium bovis BCG infection in TNF-LT-alpha-deficient mice Clin Immunol 2000 94 192 199 10692238 10.1006/clim.2000.4835 Keane J Gershon S Wise RP Mirabile-Levens E Kasznica J Schwieterman WD Siegel JN Braun MM Tuberculosis associated with infliximab, a tumor necrosis factor alpha-neutralizing agent N Engl J Med 2001 345 1098 1104 11596589 10.1056/NEJMoa011110 Mohan AK Cote TR Siegel JN Braun MM Infectious complications of biologic treatments of rheumatoid arthritis Curr Opin Rheumatol 2003 15 179 184 12707568 10.1097/00002281-200305000-00002 Mohan AK Cote TR Block JA Manadan AM Siegel JN Braun MM Tuberculosis following the use of etanercept, a tumor necrosis factor inhibitor Clin Infect Dis 2004 39 295 299 15306993 10.1086/421494 Keane J TNF-blocking agents and tuberculosis: new drugs illuminate an old topic Rheumatology (Oxford) 2005 15741198 Vassalli P The pathophysiology of tumor necrosis factors Annu Rev Immunol 1992 10 411 452 1590993 10.1146/annurev.iy.10.040192.002211 Sedgwick JD Riminton DS Cyster JG Korner H Tumor necrosis factor: a master-regulator of leukocyte movement Immunol Today 2000 21 110 113 10689296 10.1016/S0167-5699(99)01573-X Black RA Rauch CT Kozlosky CJ Peschon JJ Slack JL Wolfson MF Castner BJ Stocking KL Reddy P Srinivasan S Nelson N Boiani N Schooley KA Gerhart M Davis R Fitzner JN Johnson RS Paxton RJ March CJ Cerretti DP A metalloproteinase disintegrin that releases tumour-necrosis factor-alpha from cells Nature 1997 385 729 733 9034190 10.1038/385729a0 Grell M Douni E Wajant H Lohden M Clauss M Maxeiner B Georgopoulos S Lesslauer W Kollias G Pfizenmaier K The transmembrane form of tumor necrosis factor is the prime activating ligand of the 80 kDa tumor necrosis factor receptor Cell 1995 83 793 802 8521496 10.1016/0092-8674(95)90192-2 Ruuls SR Hoek RM Ngo VN McNeil T Lucian LA Janatpour MJ Korner H Scheerens H Hessel EM Cyster JG McEvoy LM Sedgwick JD Membrane-bound tnf supports secondary lymphoid organ structure but is subservient to secreted tnf in driving autoimmune inflammation Immunity 2001 15 533 543 11672536 10.1016/S1074-7613(01)00215-1 Akassoglou K Probert L Kontogeorgos G Kollias G Astrocyte-specific but not neuron-specific transmembrane TNF triggers inflammation and degeneration in the central nervous system of transgenic mice J Immunol 1997 158 438 445 8977220 Olleros ML Guler R Corazza N Vesin D Eugster HP Marchal G Chavarot P Mueller C Garcia I Transmembrane TNF induces an efficient cell-mediated immunity and resistance to Mycobacterium bovis bacillus Calmette-Guerin infection in the absence of secreted TNF and lymphotoxin-alpha J Immunol 2002 168 3394 3401 11907097 Marino MW Dunn A Grail D Inglese M Noguchi Y Richards E Jungbluth A Wada H Moore M Williamson B Basu S Old LJ Characterization of tumor necrosis factor-deficient mice Proc Natl Acad Sci U S A 1997 94 8093 8098 9223320 10.1073/pnas.94.15.8093 Fremond CM Yeremeev V Nicolle DM Jacobs M Quesniaux VF Ryffel B Fatal Mycobacterium tuberculosis infection despite adaptive immune response in the absence of MyD88 J Clin Invest 2004 114 1790 1799 15599404 10.1172/JCI200421027 Botha T Ryffel B Reactivation of Latent Tuberculosis Infection in TNF-Deficient Mice J Immunol 2003 171 3110 3118 12960337 Espevik T Nissen-Meyer J A highly sensitive cell line, WEHI 164 clone 13, for measuring cytotoxic factor/tumor necrosis factor from human monocytes J Immunol Methods 1986 95 99 105 3782828 10.1016/0022-1759(86)90322-4 Muller M Eugster HP Le Hir M Shakhov A Di Padova F Maurer C Quesniaux VF Ryffel B Correction or transfer of immunodeficiency due to TNF-LT alpha deletion by bone marrow transplantation Mol Med 1996 2 247 255 8726467 Lutz MB Kukutsch N Ogilvie AL Rossner S Koch F Romani N Schuler G An advanced culture method for generating large quantities of highly pure dendritic cells from mouse bone marrow J Immunol Methods 1999 223 77 92 10037236 10.1016/S0022-1759(98)00204-X Nicolle DM Pichon X Bouchot A Maillet I Erard F Akira S Ryffel B Quesniaux VF Chronic pneumonia despite adaptive immune response to Mycobacterium bovis BCG in MyD88-deficient mice Lab Invest 2004 84 1305 1321 15258598 10.1038/labinvest.3700149 Green LC Wagner DA Glogowski J Skipper PL Wishnok JS Tannenbaum SR Analysis of nitrate, nitrite, and [15N]nitrate in biological fluids Anal Biochem 1982 126 131 138 7181105 10.1016/0003-2697(82)90118-X Hodge-Dufour J Marino MW Horton MR Jungbluth A Burdick MD Strieter RM Noble PW Hunter CA Pure E Inhibition of interferon gamma induced interleukin 12 production: a potential mechanism for the anti-inflammatory activities of tumor necrosis factor Proc Natl Acad Sci U S A 1998 95 13806 13811 9811882 10.1073/pnas.95.23.13806 Kindler V Sappino AP Grau GE Piguet PF Vassalli P The inducing role of tumor necrosis factor in the development of bactericidal granulomas during BCG infection Cell 1989 56 731 740 2647299 10.1016/0092-8674(89)90676-4 Zganiacz A Santosuosso M Wang J Yang T Chen L Anzulovic M Alexander S Gicquel B Wan Y Bramson J Inman M Xing Z TNF-alpha is a critical negative regulator of type 1 immune activation during intracellular bacterial infection J Clin Invest 2004 113 401 413 14755337 10.1172/JCI200418991 Drennan MB Nicolle D Quesniaux VJ Jacobs M Allie N Mpagi J Fremond C Wagner H Kirschning C Ryffel B Toll-like receptor 2-deficient mice succumb to Mycobacterium tuberculosis infection Am J Pathol 2004 164 49 57 14695318 Abel B Thieblemont N Quesniaux VJ Brown N Mpagi J Miyake K Bihl F Ryffel B Toll-like receptor 4 expression is required to control chronic Mycobacterium tuberculosis infection in mice J Immunol 2002 169 3155 3162 12218133 Saunders BM Tran S Ruuls S Sedgwick JD Briscoe H Britton WJ Transmembrane TNF is sufficient to initiate cell migration and granuloma formation and provide acute, but not long-term, control of Mycobacterium tuberculosis infection J Immunol 2005 174 4852 4859 15814712 Grivennikov SI Tumanov AV Liepinsh DJ Kruglov AA Marakusha BI Shakhov AN Murakami T Drutskaya LN Forster I Clausen BE Tessarollo L Ryffel B Kuprash DV Nedospasov SA Distinct and nonredundant in vivo functions of TNF produced by t cells and macrophages/neutrophils: protective and deleterious effects Immunity 2005 22 93 104 15664162 10.1016/j.immuni.2004.11.016 Jacobs M Marino MW Brown N Abel B Bekker LG Quesniaux VJ Fick L Ryffel B Correction of defective host response to Mycobacterium bovis BCG infection in TNF-deficient mice by bone marrow transplantation Lab Invest 2000 80 901 914 10879741 Jacobs M Brown N Allie N Chetty K Ryffel B Tumor necrosis factor receptor 2 plays a minor role for mycobacterial immunity Pathobiology 2000 68 68 75 10878503 10.1159/000028116 Bekker LG Moreira AL Bergtold A Freeman S Ryffel B Kaplan G Immunopathologic effects of tumor necrosis factor alpha in murine mycobacterial infection are dose dependent Infect Immun 2000 68 6954 6961 11083819 10.1128/IAI.68.12.6954-6961.2000 Olleros ML Guler R Vesin D Parapanov R Marchal G Martinez-Soria E Corazza N Pache JC Mueller C Garcia I Contribution of Transmembrane Tumor Necrosis Factor to Host Defense against Mycobacterium bovis Bacillus Calmette-Guerin and Mycobacterium tuberculosis Infections Am J Pathol 2005 166 1109 1120 15793291 Lucas R Juillard P Decoster E Redard M Burger D Donati Y Giroud C Monso-Hinard C De Kesel T Buurman WA Moore MW Dayer JM Fiers W Bluethmann H Grau GE Crucial role of tumor necrosis factor (TNF) receptor 2 and membrane-bound TNF in experimental cerebral malaria Eur J Immunol 1997 27 1719 1725 9247583 Kusters S Tiegs G Alexopoulou L Pasparakis M Douni E Kunstle G Bluethmann H Wendel A Pfizenmaier K Kollias G Grell M In vivo evidence for a functional role of both tumor necrosis factor (TNF) receptors and transmembrane TNF in experimental hepatitis Eur J Immunol 1997 27 2870 2875 9394812 Maeda S Chang L Li ZW Luo JL Leffert H Karin M IKKbeta is required for prevention of apoptosis mediated by cell-bound but not by circulating TNFalpha Immunity 2003 19 725 737 14614859 10.1016/S1074-7613(03)00301-7 Harashima S Horiuchi T Hatta N Morita C Higuchi M Sawabe T Tsukamoto H Tahira T Hayashi K Fujita S Niho Y Outside-to-inside signal through the membrane TNF-alpha induces E-selectin (CD62E) expression on activated human CD4+ T cells J Immunol 2001 166 130 136 11123285 MacMicking JD Nathan C Hom G Chartrain N Fletcher DS Trumbauer M Stevens K Xie QW Sokol K Hutchinson N Altered responses to bacterial infection and endotoxic shock in mice lacking inducible nitric oxide synthase Cell 1995 81 641 650 7538909 10.1016/0092-8674(95)90085-3 Chan J Xing Y Magliozzo RS Bloom BR Killing of virulent Mycobacterium tuberculosis by reactive nitrogen intermediates produced by activated murine macrophages J Exp Med 1992 175 1111 1122 1552282 10.1084/jem.175.4.1111 Garcia I Guler R Vesin D Olleros ML Vassalli P Chvatchko Y Jacobs M Ryffel B Lethal Mycobacterium bovis Bacillus Calmette Guerin infection in nitric oxide synthase 2-deficient mice: cell-mediated immunity requires nitric oxide synthase 2 Lab Invest 2000 80 1385 1397 11005207	16285886	PMC1325056	CC BY	2021-01-04 16:36:25	no	Respir Res. 2005 Nov 14; 6(1):136	utf-8	Respir Res	2,005	10.1186/1465-9921-6-136	oa_comm
==== Front BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1651630305710.1186/1471-2164-6-165Research ArticleExpression and genomic organization of zonadhesin-like genes in three species of fish give insight into the evolutionary history of a mosaic protein Hunt Peter ND [email protected] Michael D [email protected] Schalburg Kristian R [email protected] William S [email protected] Ben F [email protected] Centre for Biomedical Research, University of Victoria, Victoria, British Columbia V8W 3N5, Canada2 Department of Molecular Biology and Biochemistry, Simon Fraser University, Burnaby, British Columbia V5A 1S6, Canada2005 22 11 2005 6 165 165 21 7 2005 22 11 2005 Copyright © 2005 Hunt et al; licensee BioMed Central Ltd.2005Hunt 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 mosaic sperm protein zonadhesin (ZAN) has been characterized in mammals and is implicated in species-specific egg-sperm binding interactions. The genomic structure and testes-specific expression of zonadhesin is known for many mammalian species. All zonadhesin genes characterized to date consist of meprin A5 antigen receptor tyrosine phosphatase mu (MAM) domains, mucin tandem repeats, and von Willebrand (VWD) adhesion domains. Here we investigate the genomic structure and expression of zonadhesin-like genes in three species of fish. Results The cDNA and corresponding genomic locus of a zonadhesin-like gene (zlg) in Atlantic salmon (Salmo salar) were sequenced. Zlg is similar in adhesion domain content to mammalian zonadhesin; however, the domain order is altered. Analysis of puffer fish (Takifugu rubripes) and zebrafish (Danio rerio) sequence data identified zonadhesin (zan) genes that share the same domain order, content, and a conserved syntenic relationship with mammalian zonadhesin. A zonadhesin-like gene in D. rerio was also identified. Unlike mammalian zonadhesin, D. rerio zan and S. salar zlg were expressed in the gut and not in the testes. Conclusion We characterized likely orthologs of zonadhesin in both T. rubripes and D. rerio and uncovered zonadhesin-like genes in S. salar and D. rerio. Each of these genes contains MAM, mucin, and VWD domains. While these domains are associated with several proteins that show prominent gut expression, their combination is unique to zonadhesin and zonadhesin-like genes in vertebrates. The expression patterns of fish zonadhesin and zonadhesin-like genes suggest that the reproductive role of zonadhesin evolved later in the mammalian lineage. ==== Body Background Molecules that are directly involved in reproduction are often subject to rapid evolutionary change [1]. Zonadhesin (ZAN) is one such molecule that has undergone domain expansion [2,3] and positive selection [4,5] in mammals. ZAN is a multi-domain sperm protein that is implicated in the species-specific binding of egg and sperm. Porcine (Sus scrofa) ZAN was first described by Hardy et al. [6] as a protein expressed by developing sperm that would bind to the zona pellucida of the egg. Since its initial discovery, zonadhesin has been identified in several other mammals, including mouse, human [2] and rabbit [7]. Recent data suggest the processed zonadhesin localizes to the acrosomal matrix and binds the zona pellucida during the acrosome reaction [8]. The discrete domains of mosaic proteins are known to be important in the evolution of new genes. Domain subunits can be rearranged, duplicated or deleted to produce a variety of proteins with different functions [9]. Zonadhesin structure is unique in its combination of protein domains. All mammalian zonadhesin genes are predicted to encode: a signal peptide, a multiple meprin A5 antigen receptor tyrosine phosphatase mu (MAM) domain, multiple trypsin-like inhibitor (TIL) domains, multiple von Willebrand D (VWD) cell adhesion domains, multiple hepta-peptide repeats that form the mucin domain, multiple epidermal growth factor (EGF) domains, a single transmembrane domain and short intracellular domain at the carboxyl terminus. The domain order is the same for all mammals studied, with the main difference being the number of MAM and VWD domains. These individual domains each have a particular function and are found in many other mosaic proteins. The extracellular VWD domain occurs in a family of immediate-early genes that are growth regulators and is thought to have an adhesive function. This modular domain is found in a variety of mosaic proteins including von Willebrand factor [10], apolipoprotein B, vitellogenins, microsomal triglyceride transfer protein (MTP) [11], and mucins [12]. Biochemical studies of pig zonadhesin have shown that the ZAN precursor is processed and the MAM domains are removed leaving the VWD domain to interact with the zona pellucida [13]. While the role of the VWD in sperm-egg binding has been addressed, the role of the mucin-repeats and the MAM domains is still unknown. The mucin or MUC domain is the primary functional domain of mucin proteins. Mucins are a diverse group of heavily glycosylated proteins that are the major component of mucus. Mucins function to lubricate surfaces and are the first line of defence against pathogens [14]. Most of the secreted mucins contain a domain with sequence similarity to VWD and a domain composed of a variable number of tandem repeats that code for serine-, threonine- and proline-rich repeat peptides that are potential glycosylation sites [15]. Two mucin-like genes that are similar to zonadhesin include alpha-tectorin, which is involved in non-syndromical autosomal dominant hearing impairment [16,17], and Fc fragment of IgG binding protein (FCGBP). By virtue of sequence identity, the closest relative to mammalian zonadhesin is FCGBP and this similarity is mostly seen in the TIL and VWD domains. FCGBP is expressed in the mucosa of the small and large intestines, epithelial colon cells and the placenta and is thought to play a role in the mucosal immune system through the promotion of multivalent IgG and the trapping of antigen IgG complexes in the mucosa [18-20]. The 170 amino acid MAM domain distinguishes zonadhesin from other VWD and mucin repeat containing genes such as FCGBP. MAM domains have adhesive function and are found in several proteins including protein-tyrosine phosphatases, neuropilins and meprins. Meprins are metalloendopeptidases that have been found in the intestinal brush border and renal membranes of mammals [21]. While their role in the zonadhesin protein is unknown, MAM domains are important for multimerization [22] and have conserved cysteine residues that are responsible for covalent interactions [23]. When this study began, zonadhesin expression had only been described in the testes. Only zonadhesin was thought to encode MAM domains, mucin repeats, and VWD domains and no non-mammalian zonadhesin orthologs had been reported. For these reasons we were interested in Atlantic salmon (Salmo salar) zonadhesin-like ESTs from a gut-derived library that encoded the MAM, mucin and VWD domains. Here we describe the cDNA, genomic structure, and expression patterns of this Atlantic salmon zonadhesin-like gene. We also use comparative genomic and expression analyses to uncover additional zonadhesin-like genes, as well as orthologs of zonadhesin, in zebrafish (Danio rerio) and puffer fish (Takifugu rubripes). Results and discussion Characterization of a zonadhesin-like gene in S. salar A gene similar to mammalian zonadhesin was identified during an expressed sequence tag (EST) analysis of Atlantic salmon. We assembled this zonadhesin-like gene (zlg) from five overlapping ESTs [24]; the largest of which [GenBank: CK990464] was sequenced by primer walking from both directions and assembled to give a 4388 bp sequence. PCR primers were designed to the 3'-end of the EST and used to amplify a probe for hybridization to Atlantic salmon genomic DNA and bacterial artificial chromosome (BAC) library filters (Probe 1, Figure 1A). Figure 1 Organization and Southern blot analysis of the Atlantic salmon zonadhesin-like gene (zlg). A) Domain structure, probe locations, and EST coverage of the Atlantic salmon Zlg. The three VWD domains, the two MAM domains and the mucin domain are shown. The signal peptide (SP) and poly (A) tail are also indicated. Probe 1 included 177 nucleotides of the 3'-UTR and 206 nucleotides of VWD3 in the coding region. Probe 2 included 233 nucleotides of the 3'-end of VWD1. The 4388 bp EST covers the full length of the cDNA sequence except for the 5'-UTR which is estimated to be approximately 200 bp by comparison to Northern blot data. B) Southern blot analysis of Atlantic salmon and rainbow trout genomic DNA. Gene copy number was assessed by genomic hybridization. Twenty micrograms of Atlantic salmon (A.S.) and rainbow trout (R.T.) genomic DNA were digested with four enzymes, EcoR I, Hind III, BamH I and Bgl II and hybridized with radiolabeled probe 1 representing the 3'-end of the zlg mRNA. Probing of 91,776 BAC clones on five Atlantic salmon genome BAC library filters [25] resulted in one positive BAC. This BAC (722P12) was subcloned, sequenced, and assembled into a 138,345 base pair contiguous sequence [GenBank: AY785950]. The assembly had more than 3000 high quality sequence reads and 10 fold sequence coverage in most regions. One gap of 500 bp was filled by PCR followed by sequencing from both directions. An in silico restriction digest matched experimental digests. The zonadhesin-like gene was the only gene found on this BAC. All other open reading frames were associated with LINE, SINE and transposon related repetitive elements. Comparison of 722P12 against itself using Dotter and PipMaker [26] did not reveal any recent domain expansions or duplications. The Simple Modular Architecture Research Tool (SMART) [27] was used to identify conserved domains of the predicted protein (Figure 1A). The SMART algorithm was used to detect three VWD domains (amino acid positions 31–198, 415–578 and 1277–1498), a VWC domain at (position 365–425), and two MAM domains (positions 708–870 and 895–1056). This tool also located three low complexity regions that corresponded to the mucin domains predicted by the computer program NetOGlyc 3.1 [28]. These mucin domains occurred at positions 1091–1099 and 1115–1158 with a smaller low complexity region located between nucleotides 693 and 704. These results were corroborated using the InterPro domain prediction computer program [29]. The SMART and InterPro domain prediction tools, in agreement with Kyte-Doolittle hydropathy data, did not find any transmembrane domains in the salmon Zlg. In contrast, the SMART tool detected a transmembrane domain at the expected location for zonadhesins from other species. Using two distinct probes (Figure 1A), only one copy of zlg was found in Atlantic salmon by Southern blot analysis (Figure 1B). However, two bands occurred when the same probes were used with rainbow trout genomic DNA (Figure 1B). Probe 1, which contained sequence homologous to the 3'-UTR of zlg (Figure 1), would be expected to be specific for this gene; however, probe 2 spanned VWD1 and could be expected to hybridize with related zonadhesin-like genes. Both probes gave similar results and only probe 1 data is shown in Figure 1B. Southern blot analysis did not reveal any other genes in Atlantic salmon other than zlg. This suggests that if another zonadhesin-like gene does exist in Atlantic salmon it is likely quite divergent from zlg. The 5'-end of the zlg cDNA coding sequence was obtained by PCR using primers designed to the predicted translation start site on the BAC sequence. The final assembly of the cDNA was 4,791 bp [GenBank: AY785949]. The total length of the mRNA found by Northern blot was just over 5 Kb (Figure 2A). The additional length of the mRNA is likely comprised of the 5'-UTR of zlg. The cDNA aligned completely to the BAC and the 23 consecutive exons of the zonadhesin-like gene utilized canonical splice sites. The cDNA has a predicted ORF of 4,518 bases that encodes a 1,506 amino acid protein. The predicted protein starts with a methionine and has a putative signal peptide of 18 amino acids. A poly (A) signal of AATAAA was identified by Genscan [30,31] at 4770 bp from the start codon of the cDNA, 241 bp downstream of the stop codon. Figure 2 Expression of the Atlantic salmon zlg mRNA. A) Expression of the zonadhesin-like gene was analyzed in a variety of tissues. Ten micrograms of total RNA from liver, brain, spleen, kidney, midgut, hindgut, foregut and gonads from male and female Atlantic salmon was blotted on a positively charged nylon membrane and hybridized with radiolabeled probe 1 representing the 3'-end of the Atlantic salmon zlg mRNA. B) Analysis of the Atlantic salmon zlg tissue expression pattern by semi-quantitative reverse transcription PCR. Zlg primers ('probe 1' primer set) were used to test for the presence of the zlg cDNA. The EST to which the primers were designed [GenBank: CK990464] and BAC 722P12 were used as positive controls. The genomic region encompassed by the primers contains two introns of 395 and 441 bp yielding a 1219 bp band upon amplification. C) Ubiquitin primers were used as a positive control for mRNA presence. zlg (ZAN neg) and ubiquitin (Ubi neg) template-free negative controls were included. Salmo salar zlg expression Semi-quantitative reverse transcription PCR was used to identify tissues expressing the zlg mRNA (Figure 2B). Liver, brain, kidney, spleen, foregut, midgut, hindgut and gonads were taken from one male and one female Atlantic salmon. Male and female salmon showed expression in the liver, midgut and hindgut. However, expression in the spleen only occurred in the male and expression in the foregut only occurred in the female. A weak band was present in the ovarian sample and expression was not observed in the testes. This gene does not appear to be expressed in either male or female brain or kidney. Northern blot results were similar to the RT-PCR results and showed a single band in male midgut and hindgut, and female midgut and foregut (Figure 2A). A weak band was seen in the Northern blot analysis of liver from both male and female fish (Figure 2A). Unlike the RT-PCR results, Northern blot analysis did not detect a transcript in the male spleen and female hindgut. This discrepancy could be due to the higher sensitivity of the PCR experiment. The zlg expression pattern differed from the testes-specific expression known for mammalian zonadhesin. To clarify the relationship between zan and zan-like genes, we looked for related genes in the genomes of other fish. Genomic analysis of the zebrafish zan locus An initial inspection of the ZV4 (September 6th, 2004 release) whole genome shotgun (WGS) assembly of the D. rerio genome suggested there are two copies of zonadhesin that exist at two distinct loci. One copy is found at linkage group 7 (scaffold 588) and consists entirely of whole genome shotgun reads. The second copy is found at linkage group 24 (scaffold 1965) and resides within a completely sequenced BAC from the CHORI-211 library [GenBank: BX649275]. The ZV3 assembly contained only one copy of zonadhesin that also assembles into linkage group 7. It is important to note that the ZV3 assembly consisted entirely of whole genome shotgun reads; the ZV4 assembly incorporated finished BAC sequence into the ZV3 assembly; and that both assemblies are considered 'pre-assemblies' that need to be analyzed with caution. The location of zonadhesin at linkage group 7 is supported by the fact that other genes with significant similarity to human chromosome 7q22 map to the same region of the D.rerio genome. These genes include acetylcholine esterase, serpine, AP1S1, unnamed product FLJ39237, unnamed protein product FLJ10925 and mucin (Figure 3). Figure 3 Alignment of zebrafish BACs to the linkage group 7 of the ZV4 whole genome shotgun assembly. Approximately 600 Kb of the zonadhesin locus from the ZV4 assembly is shown on the x-axis. Alignments of this region to zebrafish BACs, the puffer fish (fugu) zan locus (Scaffold 860) and human 7q22 were performed using MultiPipMaker. Regions of high conservation consisting of gap-free alignments of at least 100 bp and 70 percent nucleotide sequence identity are shown in red and other locally aligned regions are shown in green. RepeatMasker [45] was used to mask zebrafish repeats before the alignments were made. The regions surrounding the zan locus at both linkage groups were inspected for possible segmental duplications. BAC clones (both finished and unfinished) that aligned to a 600 kb region surrounding the zan locus at linkage group 7, or a 1 MB region at linkage group 24, were obtained and aligned to both loci using the BLASTZ algorithm through the MultiPipMaker web server. Our analysis showed that the zonadhesin-containing BAC BX649275 integrated completely into linkage group 7 and was ≅ 97% identical overall, not including indels. Seven additional BACs from the DKey library aligned to the scaffold and produced an acceptable tile through the entire zan locus (Figure 3). In contrast, only a portion of BAC BX649275 and a portion of BX640466.9 aligned to the zan locus at linkage group 24. This is suggestive of an assembly artifact that resulted in the assignment of a second zonadhesin gene to linkage group 24. The differences between individual BACs, and between BACs and linkage group 7, can be as high as 3%. This amount of polymorphism is higher than the 0.5% polymorphism rate expected from the whole genome shotgun sequence, which came from approximately 1000, 5 day old embryos [32]. Despite this high rate of polymorphism, the existence of one zonadhesin locus at linkage group 7 is supported by the large tile of overlapping clones at linkage group 7, and the rapidly evolving nature of zonadhesin genes [1]. Prediction of the zebrafish zan transcript and domain structure We further analysed the zonadhesin gene found on the completely sequenced BAC (BX649275). The 4,616 amino acid translation product of the putative zan gene lacked a signal peptide, but contained the domain structure of: two MAM domains; a mucin domain; nine VWD domains; and a transmembrane region. This domain organization is typical of zonadhesin (Figure 4). The entire zonadhesin gene was also contained within 13 unordered pieces from the clone DKey-3K24 [GenBank: CR848737.2]. From this clone we obtained a third putative zan transcript that was most similar to the complete zan sequence from BAC BX649275. The predicted cDNA from CR848737.2 differs from the cDNA of BX649275 by 0.8%. The putative ZAN protein encoded by BAC CR848737.2 has a domain structure the same as the one encoded by the putative zan transcript of BAC BX649275. Figure 4 Domain structures of representative zonadhesins and related proteins. Mouse [GenBank: NP035871.1], pig [GenBank: Q28983], human [GenBank: Q9Y493], puffer fish and zebrafish zonadhesins are shown with the salmon and zebrafish Zlg. The human MUC2 [GenBank: L21998.1], human prepro von Willebrand factor [GenBank: P04275] and rat (Rattus norvegicus) MUC2 [GenBank: Q62635] and human and chicken FCGBP proteins [GenBank: NP003881.1 and XP422715.1] are also included. Signal peptide sequences (SP) are drawn as a white box and are found at the N-terminus of all proteins except for puffer fish and zebrafish Zan. All the zonadhesins have an epidermal growth factor (EGF) domain (in red), a transmembrane (TM) domain (in yellow) and a cytoplasmic domain (CD) (in green) which are labeled on the mouse Zan protein. The MAM and mucin domains are indicated and the VWD domains are represented by 'D'. The partial D3 domains of the mouse are represented by D3p1-20. Other partial VWD domains are denoted D0 or D' as found in the current literature. It is important to note that although this domain representation is consistent with previous representations of zonadhesin and other VWD containing proteins, other domain types such as von Willebrand C (VWC) and trypsin inhibitor-like (TILa) domains have not been included here. Zebrafish zonadhesin has expanded its VWD domains through recent tandem duplication. The VWD domains 5, 6 and a fragment of 7 are encoded in 17 exons (45–61). These exons have identical sizes to the exons (11–27 and 28–44) encoding the first four VWD domains. Furthermore, each group of 17 exons are symmetrical and flanked by phase '1' introns, which is evidence for recent domain expansion. Pair-wise alignment of BAC BX649275 against itself revealed these exons are found in three ≈ 5 kb blocks that are 85–87% identical. The zan locus at linkage group 7 only contains two of these ≈ 5 kb repeats. It is possible the third repeat was collapsed in the whole genome shotgun assembly process or the presence of two repetitive blocks is a true population variant. Danio rerio zonadhesin expression To compare the expression pattern and verify the transcript size of the predicted zebrafish zan we extracted gut and testes RNA and performed RT-PCR and Northern blot analysis. Three different PCR primer sets were designed against three regions of the predicted zan cDNA. The first primer set was designed to amplify the exons encoding the MAM domains through to the first VWD domain. This primer set produced a doublet in both male and female (the larger band in the female was very faint) (Figure 5A). The female bands were approximately 50 bp larger than the male bands. The differences between individuals may be due to a variation in mucin domain length since mucin domains have been shown to be variable in other genes [12]. The second primer set flanked the ninth VWD domain and the third stretched from the epidermal growth factor domain to the cytoplasmic domain. Each primer set identified a zonadhesin transcript in the gut, but not in the testes (Figure 5A). These results were corroborated by Northern blot analysis that found a single transcript of ≅ 15 Kb in the gut but not in the testes (Figure 5B and 5C). This ≅ 15 kb transcript correlates to the mRNA length predicted from the genomic sequence and supports the existence of a single zonadhesin gene in zebrafish. Figure 5 Zebrafish zonadhesin mRNA expression. A) Semi-quantitative reverse transcription PCR analysis of zebrafish tissues. Three primer sets were designed against zonadhesin specific sites. The amplicons of the first primer set crossed the mucin domain stretching from the MAM domains to the first VWD domain (MAM-VWD). The second primer set amplified within the VWD9 domain (VWD9). Products of the third primer set stretched from the last VWD domain to the cytoplasmic domain (VWD9-CD). Ubiquitin primers were used as a positive control and a template-free reaction was included as a negative control. B-C) Expression of zebrafish zonadhesin analyzed by Northern blot. B) Expression of zonadhesin was investigated in the zebrafish testis and gut. Five micrograms of total RNA from each tissue was blotted on a positively charged nylon membrane and hybridized with a DNA probe encompassing the zebrafish EGF to CD region of zan. C) The same Northern blot membrane was reprobed with zebrafish alpha-tubulin as a positive control. Genomic prediction T. rubripes zan cDNA transcript and domain structure Analysis of the puffer fish genome assembly release 2 (SCAFFOLDS 17 05 02; [33]) revealed a putative zonadhesin gene in the same contig (scaffold 870) as the ache gene. This syntenic relationship was also found at the mammalian and zebrafish zan loci (Figure 3). However, Scaffold 870 has been split in the current Fugu genome project release (MAYFFOLDS) leaving zonadhesin in a gap-free region of scaffold 2,670 without ache. Puffer fish zonadhesin was predicted to contain 47 exons that coded for a protein of 2,525 amino acids. The predicted zonadhesin protein contained two MAM domains at the N-terminus, a mucin domain, five VWD domains, an EGF domain, a transmembrane domain and a short cytoplasmic domain (Figure 4). This protein has the same domains in the same order as human zonadhesin. No signal peptide was identified in the puffer fish zonadhesin. However, this sequence may be incomplete since it was found at the end of the scaffold sequence. Expression of T. rubripes zonadhesin Evidence from the GenBank database suggests that puffer fish zonadhesin is also expressed in the gut. Sequences can be found that have been isolated from gut-specific libraries [GenBank: CA591505, CA588342 and CA588225], but there have been no zonadhesin sequences found with testes-specific expression. Structural similarity of D. rerio and T. rubripes to mammalian zan genes Although fish and mammals have not shared a common ancestor for an estimated 450 million years [34], the domain structure of zonadhesin has been highly conserved (Figure 4). However, domain numbers between species are variable and this variability appears to have been influenced by tandem duplication. Tandem duplications have occurred in both mammalian and fish species and are most prevalent in the VWD domain region (Figure 4). It is this region which, at least in mammals, has been shown to be important for zona pellucida binding [35]. While multiple VWD domains are found in all characterized zonadhesins, recent expansion of this domain is seen in the mouse and in the zebrafish. Repeated tandem duplication in the mouse zonadhesin gene resulted in 20 copies of a two-exon segment encoding a partial VWD3 domain that increased the length of the protein by over 2000 amino acids [2,3]. In zebrafish, the double duplication of two domains homologous to the puffer fish VWD1 and VWD2, as well as a portion of the VWD3 domain (containing 17 VWD-coding exons), resulted in an additional 34 exons that encoded 4 additional full VWD domains and two partial VWD domains in the zebrafish Zan (Figure 4). The ancestral zonadhesin likely looked similar to the puffer fish Zan as it is very similar in length and domain structure to most mammalian zonadhesins (Figure 4). The puffer fish gene also has slightly higher identity with the human zonadhesin gene at 52% (indels removed) compared with the 50% identity between zebrafish and human whereas the zebrafish and puffer fish genes are 62% identical. The five VWD domains of the puffer fish also cluster with the four in human (Figure 6A). The last four VWD domains of puffer fish Zan seem to be homologous to the last four human domains; although, the first VWD domain of puffer fish also has high similarity to the second domain in human. This may be the result of an ancient duplication of the first two VWD domains in puffer fish and subsequent loss of the new first domain. This inheritance is similar to that of the MAM domains. Figure 6 Phylogeny of the VWD and MAM domains of zonadhesin and zonadhesin-like genes. A) Human and puffer fish zonadhesin VWDs were aligned with the VWD domains of salmon and zebrafish Zlgs and chicken FCGBP-like protein. A neighbor-joining tree utilized 140 informative sites. B) Phylogeny of the MAM domains from puffer fish, zebrafish, pig, mouse and human ZANs, salmon and zebrafish Zlg and chicken FCGBP-like proteins are shown with a neighbor-joining tree. This phylogeny utilized 88 informative sites. Both phylogenies used gap-free alignments and a Poisson substitution correction model. Consensus trees based on 1000 pseudo-replicates are reported with the bootstrap support values indicated above the respective nodes. The scale is given in amino acid substitutions per site. Puffer fish zonadhesin has two MAM domains, a structure matching the rabbit and pig proteins [7]. Neighbor-joining tree analysis of individual domains reveals that the MAM1 domain of puffer fish is most similar to the MAM1 and MAM3 domains of human, while the MAM2 domain of puffer fish groups with MAM2 of human (Figure 6B). This pattern of similarity could be explained by a duplication of both MAM domains in the human lineage and subsequent loss of the fourth domain. This phylogeny, in combination with the conserved domain order and synteny between fish and mammal zonadhesin loci, supports the orthologous relationship of these genes. Genomic and phylogenetic analysis of genes with a domain content similar to zonadhesin Until this study, zonadhesin was generally thought to be unique in its domain content as no other genes were reported to contain MAM, mucin, TIL and VWD domains. However, the characterization of the salmon zlg revealed all of these domains, but in a different order (Figure 4). The expression of zlg is similar to zebrafish zonadhesin which, in addition to domain content, established a possible evolutionary link between these genes. Examination of GenBank sequences also revealed that zonadhesin-like genes have been found in gut tissues of other species; however, the automated annotation is based on the similarity of the TIL and VWD domains of FCGBP. For example, there are three human colon ESTs [GenBank: AI984139, AI983786 and AI983612] that are annotated as similar to zonadhesin; however, these genes align perfectly to the FCGBP gene at chromosome 19q13.2. Similarly, the only mouse colon EST that is annotated as similar to zonadhesin aligns to the Fcgbp gene in the orthologous region at mouse chromosome 7. We looked for zonadhesin-like genes in puffer fish, zebrafish and chicken (Gallus gallus) genome projects. This search uncovered several regions with VWD-containing proteins without any detectable MAM domains, some of these possibly related to FCGBP proteins. One interesting exception was a zonadhesin-like gene found on zebrafish chromosome 2. This putative zebrafish zlg encodes a 1,308 amino acid protein containing three MAM domains, the first of which is flanked by short (15 amino acid) mucin-like low complexity regions. The MAM domains are followed by two VWD domains. This gene structure is reminiscent of S. salar Zlg and shows that a zonadhesin ortholog and a zan-like gene exist together in the zebrafish genome (Figure 4). The search for zonadhesin-like genes in chicken (Gallus gallus) did not reveal an obvious zonadhesin ortholog but rather a prediction of a FCGBP-like protein residing on chromosome 9 [GenBank: XP422715.1]. We analysed the corresponding region of the G. gallus genome and utilized additional EST evidence and in silico predictions to obtain a putative transcript that encodes a 4,770 amino acid gene product. This gene product contains three MAM domains in addition to VWD and mucin-type O-glycosylation sites after each of the MAM domains. Overall, this MAM-mucin-TIL-VWD series of domains is reminiscent of all zonadhesins and is evidence for a common evolutionary origin of zonadhesin, the zonadhesin-like genes and the Fc fragment of IgG binding protein(FCGBP). We extracted the MAM and VWD domains of representative zonadhesin and zonadhesin-like genes and performed a phylogenetic analysis (see Figures 6A and 6B respectively). In addition to the clustering of fish and mammalian zonadhesin, both phylogenetic trees suggest an evolutionary relationship among the zonadhesin-like genes. In particular, the zebrafish Zlg MAM1 and MAM2 domains grouped with the salmon Zlg MAMs as well as two of the chicken FCGBP-like MAM domains (Figure 6B). The grouping of the MAM domains of these proteins indicates that the zonadhesin-like genes represent a novel gene family that is distinct from zonadhesin. Although the phylogeny is more complex with many nodes not well supported, the evolutionary relationship between the fish Zlgs and the chicken FCGBP-like gene was also observed for the VWD domain (Figure 6A). For example, the VWD1 domains of the fish Zlgs clustered with the VWD2 and VWD5 domains of the chicken FCGBP-like gene. A second clade consisting of the VWD2 domains of salmon and zebrafish Zlg, and the VWD3 and VWD6 domain of the chicken FCGBP-like gene also formed. The clade containing puffer fish ZAN VWD5, human ZAN VWD4 and salmon Zlg VWD3 also suggests that zonadhesin is closely related to the zonadhesin-like genes. The human FCGBP VWD domains all formed a single clade except for VWD1 which grouped with the chicken FCGBP-like VWD1 and 4 (data not shown). Overall, the phylogenies of the VWD and MAM domains combined with the expression patterns of: fish zonadhesins, fish and chicken zonadhesin-like genes, FCGBP, mucin, and several MAM containing proteins, suggest that these mosaic genes share a common ancestor. Zonadhesin in the context of other sperm-egg interacting proteins Many mammalian zona pellucida adhesion molecule candidates appear to have evolved from different physiological processes. Well known examples of sperm proteins with enzymatic function that have been 'hijacked' into playing non-enzymatic roles in sperm-egg interactions include: B4GALT1/GalTase (beta 1,4 galactosyltransferase), SPAM1/PH-20 (hyaluronidase), HK1/ZRK/p95 (hexokinase), and ARSA/SLIP1 (aryl-sulfatase-A; reviewed by [36]). Evidence for immune-system hijacking events comes from the discovery of several complement system proteins in human spermatozoa, seminal plasma and follicular fluid (reviewed by [37]). The partial activation of the complement system (without engaging the membrane attack complex) in acrosome-reacted spermatozoa suggests how components of a conserved immune system pathway could play a new role in sperm-egg recognition [38]. Although the function of the fish zonadhesin and zonadhesin-like genes is not known, their expression in the gut and absence of expression in the testes, combined with their homology to gut-expressed genes of the mucosal immune system (i.e. FCGBP and mucin), suggest that zonadhesin was also 'hijacked' by the mammalian reproductive system. Conclusion We identified zonadhesin genes in zebrafish and puffer fish that are similar in domain order and content to all known mammalian orthologs. Unlike all mammalian zonadhesin genes studied to date, zebrafish zan was expressed in the gut but not in the testes. In addition to these orthologs, we characterized zonadhesin-like genes (zlg) in Atlantic salmon, zebrafish and chicken. While the Atlantic salmon zlg contained the same domains found in zonadhesin, the order of these domains was altered and the expression was found predominantly in the gut and not in the testes. Overall, this suggests that zonadhesin's reproductive role evolved later in the mammalian lineage. Methods An Atlantic salmon CHORI-214 bacterial artificial chromosome (BAC) library was obtained from BACPAC Resources, Children's Hospital Oakland Research Institute (CHORI) [25]. Five BAC library filters (13A-17A) were hybridized with a probe designed from a zonadhesin-like EST [GenBank: CK990464]. These five filters contained 91,776 BAC clones in a pTARBAC2.1 vector with an average size of 190 Kb. Each filter was estimated to represent the salmon genome once. Filter hybridizations were conducted as described by CHORI [39]. The PCR product that was used as a probe was generated by PCR (Invitrogen) using the manufacturer's protocol and the following primer set: 5'-GTGCCCATTGTAGGAAGGAA-3' and 5'-GGGGTTGAGGATTCTGGAG-3'. The probe was gel purified and end-labeled with γ32P-ATP (Amersham). Probed BAC library filters were visualized using a Molecular Dynamics Storm PhosphorImaging system. BAC DNA was isolated by an alkaline lysis procedure using Nucleobond columns (Clontech) using the manufacturer's protocol. The isolated 722P12 BAC DNA was nebulized and the DNA was blunt-ended. The blunt-ended repaired DNA was size fractioned by electrophoresis and the gel region corresponding to 1200–3000 bp was excised and gel purified (Qiagen). The fragments were blunt-end ligated into pUC19 plasmid cut with Hinc II (NEB) and transformed into electrocompetent DH5α E. coli cells using a Bio-Rad Gene Pulser system. Extracted recombinant plasmid templates were sequenced on an ABI 3700 DNA sequencer. Bases were called using PHRED [40,41]. The resulting 3000+ high quality sequence reads were assembled using PHRAP [42] and then viewed and edited using Consed [43]. One gap of about 500 bp in the assembly was filled by designing primers to the contig ends followed by amplification of this BAC region by PCR and subsequent cloning and sequencing this fragment. Restriction digests of the isolated BAC were compared to in silico digests for assembly confirmation. BAC 722P12 was deposited in GenBank under the accession number AY785949. Dotter [44] and PipMaker [26] were used to compare the BAC sequence to itself and to identify duplicated and repeated regions. Identification of other repeat elements was done with RepeatMasker [45] using repeat library 4.01 from Repbase [46]. Low complexity regions that corresponded to the mucin domains were predicted by the computer program NetOGlyc 3.1 [28]. Genscan was used to predict novel genes and gene structures [30,31]. Translated and untranslated BLAST searches were performed using 722P12 BAC as the query. The Atlantic salmon zlg cDNA was partially sequenced by first completing a series of primer walks from the 5'- and 3'-ends to complete a 4,388 bp EST clone [GenBank: CK990464]. Primers were designed to the predicted translation start site on the genomic DNA in order to amplify fragments spanning the 5' end of the coding region from gut total cDNA. Sim4 [47] and Dotter [44] were used to align the cDNA sequence with the genomic DNA to identify exonic and intronic regions. The zlg cDNA was deposited under the GenBank accession number AY785950. Southern blot analysis Liver genomic DNA from male Atlantic salmon and rainbow trout were isolated from 100 mg of tissue using the Easy-DNA Kit (Invitrogen). Southern blot analysis was performed as described by Hames and Higgins [48]. DNA was digested by restriction enzymes EcoR I, Hind III, BamH I and Bgl II (NEB). The digested DNA was electrophoresed for 18 h and then transferred to Hybond, positively charged nylon membrane (Amersham). Two probes were prepared to the 5' and 3' ends of the zlg cDNA sequence. Probe 1 included 206 nucleotides of the 3' end of the zlg ORF and 177 nucleotides of the 3'-UTR and probe 2 included 233 nucleotides of the VWD1 domain (Figure 1A). Both probes were gel purified and labeled with a Rediprime II random labeling kit (Amersham) with 50 μCi of α32P-labeled-dCTP. Blots were prehybridized at 68°C for 4 h in hybridization buffer (5× SSC, 5× Denhardt's solution and 1% SDS) with 100 μg/mL denatured human placental DNA (Sigma). This was followed by replacement with fresh, preheated (68°C) hybridization buffer and the addition of the radiolabeled probe. Hybridization was allowed to proceed overnight. Following hybridization, the membrane was washed twice with 20 mL of 2× SSC, 0.1% SDS at room temperature for 15 min followed by two 15 min washes of 200 mL 0.2× SSC, 0.1% SDS at 65°C in a shaking bath. Prehybridization, hybridization and wash conditions were the same for both probes. Northern blot analysis RNA was extracted from Atlantic salmon tissues (liver, brain, spleen, kidney, midgut, hindgut, foregut and gonads) and from zebrafish testes and gut using Trizol (Invitrogen). Total RNA samples were quantified and checked for quality by spectrophotometric analysis and agarose gel electrophoresis. Northern blots were prepared using the NorthernMax-Gly kit (Ambion) following the manufacturer's instructions. Ten μg of Atlantic salmon or 5 μg of zebrafish total RNA from each tissue was blotted on a Hybond positively charged nylon membrane (Amersham). Northern blot analysis of Atlantic salmon tissues utilized the same probe 1 described for the Southern blot analysis. The zebrafish zonadhesin probe was amplified from gut tissue using a primer set spanning from the 3' end of the epidermal growth factor domain through to the cytoplasmic domain using primers 5'-GGTTTGAGGGCACAAACTGT-3'and 5'-TAGGGATGCGCTGTCTTTTT-3'. Prehybridization for both Northern blots proceeded for 2 h at 42°C in 15 mL of ULTRAhyb buffer (Ambion). Hybridization with the α32P-dCTP-labeled probe at a final concentration of 106cpm/mL of hybridization buffer was performed at 42°C overnight. The zebrafish Northern blot was stripped and reprobed with a probe designed from zebrafish alpha-tubulin that was expected to produce a doublet of 1485 bp [GenBank: AY398374] and 1544 bp [GenBank: AF029250]. Semiquantitative reverse transcription PCR Total RNA from Atlantic salmon and zebrafish tissue extracted as above was reverse transcribed using Superscript II enzyme (Invitrogen) and an 18 nucleotide oligo d(T) primer as described in the manufacturer's protocol with exception of the production of the cDNA template for the zebrafish MAM primer set which required a gene-specific internal primer for reverse transcription to reach this region (5'-AGACACTTTCACCCCCAGTG-3'). For Atlantic salmon, one μL of cDNA was amplified in a 25 μL reaction volume with either zlg probe 1 primers or ubiquitin primers (5'-ATGTCAAGGCCAAGATCCAG-3' and 5'-TAATGCCTCCACGAAGACG-3'). The zlg EST [GenBank: CK990464] and the 722P12 BAC were included as positive controls and both primer sets were run with template-free negative controls. For zebrafish, three primer sets were designed against the genomic zan sequence. The first primer set flanked the second MAM domain through to the first VWD domain (5'-TTGCAATTGATAGCGTCTGC-3' and 5'-TTCAGTCACAGGGTCACAGG-3'); the second primer set flanked the ninth VWD domain (5'-GGAGACCGTTACTGCAAACC-3' and 5'-CGAACAGTGATGCCGTACAC-3'); and the third primer set stretched from the epidermal growth factor domain to the cytoplasmic domain (see Northern blot probe description). The integrity of each cDNA was confirmed by control PCR reactions that used an ubiquitin primer set (5'-CCTCGAGGTAGAGCCAAGTG-3' and 5'-GCAGCACACAAGGTGCAAAGTA-3') and a template-free negative control. Puffer fish and zebrafish zonadhesin prediction and analysis The puffer fish zonadhesin was found by BLASTN search of the three puffer fish genome assembles available at the MRC RFCGR Fugu genome database [33] with human and mouse zan nucleotide sequences. Scaffold 870 from assembly 2 was found to have similarity to zonadhesin and was subsequently analyzed by Genscan for coding sequences and peptide predictions. Puffer fish ESTs were aligned to scaffold 870 using Sim4. The zebrafish zonadhesin was found by BLASTP search of the Ensembl zebrafish peptide database (Ensembl assembly 25.4.1) using a fragment of Atlantic salmon predicted protein as the query sequence. The Atlantic salmon query fragment consisted of all the amino acids except those representing the MAM domains. The two genomic regions identified were analyzed by Genscan to find the putative coding and protein sequences. We looked for zonadhesin-like genes in puffer fish, zebrafish and chicken (Gallus gallus) genome projects using the ENSEMBL BLAST search tools using both cDNA and protein sequences from several zonadhesin-like genes as in silico probes. These included salmon zlg, human, puffer fish and zebrafish zans and the related human FCGBP gene. Genomic regions from all significant matches were extracted and gene prediction analysis was performed using Genscan. Domain predictions and phylogenetic analysis Protein domains were predicted using SMART and InterPro prediction tools and the domains were extracted from the parent nucleotide and protein sequences. Multiple sequence alignments of the extracted domains were done using ClustalX [50] followed by manual inspection. See additional file 1 and additional file 2 for VWD and MAM multiple alignments respectively. Neighbor-joining trees were created using MEGA3. Consensus trees based on 1000 pseudoreplicates are reported with the bootstrap support values indicated above the respective nodes. Gaps were removed and we reported phylogenetic data using the Poisson correction model with uniform rates across all sites. Neighbor-joining trees were also performed using the Poisson correction model with unequal rates across sites using gamma distance parameters 0.65 and 2.25. While some of the less supported nodes changed, the clades discussed here did not vary substantially using these different parameters for either the MAM or the VWD trees. We also used the equal input model using either uniform rates across all sites or unequal rates across all sites using gamma distance parameters 0.65 and 2.25. Again these parameters did not change the topology of the clades discussed in the text. Authors' contributions PNDH carried out the molecular genetic studies, sequencing and sequence alignment and drafted the manuscript. MDW performed the comparative genomic analysis and contributed to the experimental design and manuscript draft. KRVS conceived the study and participated in experimental genetic studies. WSD contributed to the production of the GRASP EST database. BFK participated in study design and coordination. All authors read and approved the final manuscript. Supplementary Material Additional File 1 Multiple sequence alignment (Fasta format) of VWD domains used to construct the phylogeny in Figure 6. Click here for file Additional File 2 Multiple sequence alignment (Fasta format) of MAM domains used to construct the phylogeny in Figure 6. Click here for file Acknowledgements This research was supported by the Natural Sciences and Engineering Research Council of Canada (BFK), the Canadian Institutes of Health Research (BFK), and the University of Victoria/ Michael Smith Foundation for Health Research (PNDH, MDW). We thank Simon Jones and Kim Taylor (Pacific Biological Station, Nanaimo, B.C., CA) for Atlantic salmon tissues and Jack and Kevin Nickolichuk (Mountain Trout Sales, Sooke, B.C., CA) for rainbow trout tissues. ==== Refs Swanson WJ Vacquier VD The rapid evolution of reproductive proteins Nat Rev Genet 2002 3 137 144 11836507 10.1038/nrg733 Gao Z Garbers DL Species diversity in the structure of zonadhesin, a sperm-specific membrane protein containing multiple cell adhesion molecule-like domains J Biol Chem 1998 273 3415 3421 9452463 10.1074/jbc.273.6.3415 Wilson MD Riemer C Martindale DW Schnupf P Boright AP Cheung TL Hardy DM Schwartz S Scherer SW Tsui LC Miller W Koop BF Comparative analysis of the gene-dense ACHE/TFR2 region on human chromosome 7q22 with the orthologous region on mouse chromosome 5 Nucleic Acids Res 2001 29 1352 1365 11239002 10.1093/nar/29.6.1352 Herlyn H Zischler H Identification of a positively evolving putative binding region with increased variability in posttranslational motifs in zonadhesin MAM domain 2 Mol Phylogenet Evol 2005 15927490 Swanson WJ Nielsen R Yang Q Pervasive adaptive evolution in mammalian fertilization proteins Mol Biol Evol 2003 20 18 20 12519901 Hardy DM Garbers DL A sperm membrane protein that binds in a species-specific manner to the egg extracellular matrix is homologous to von Willebrand factor J Biol Chem 1995 270 26025 26028 7592795 10.1074/jbc.270.44.26025 Lea IA Sivashanmugam P O'Rand MG Zonadhesin: characterization, localization, and zona pellucida binding Biol Reprod 2001 65 1691 1700 11717130 10.1095/biolreprod65.6.1691 Olson GE Winfrey VP Bi M Hardy DM NagDas SK Zonadhesin assembly into the hamster sperm acrosomal matrix occurs by distinct targeting strategies during spermiogenesis and maturation in the epididymis Biol Reprod 2004 71 1128 1134 15175237 10.1095/biolreprod.104.029975 Patthy L Modular assembly of genes and the evolution of new functions Genetica 2003 118 217 231 12868611 10.1023/A:1024182432483 Sadler JE Biochemistry and genetics of von Willebrand factor Annu Rev Biochem 1998 67 395 424 9759493 10.1146/annurev.biochem.67.1.395 Babin PJ Bogerd J Kooiman FP Van Marrewijk WJ Van der Horst DJ Apolipophorin II/I, apolipoprotein B, vitellogenin, and microsomal triglyceride transfer protein genes are derived from a common ancestor J Mol Evol 1999 49 150 160 10368443 Eckhardt AE Timpte CS DeLuca AW Hill RL The complete cDNA sequence and structural polymorphism of the polypeptide chain of porcine submaxillary mucin J Biol Chem 1997 272 33204 33210 9407109 10.1074/jbc.272.52.33204 Bi M Hickox JR Winfrey VP Olson GE Hardy DM Processing, localization and binding activity of zonadhesin suggest a function in sperm adhesion to the zona pellucida during exocytosis of the acrosome Biochem J 2003 375 477 488 12882646 10.1042/BJ20030753 Moncada DM Kammanadiminti SJ Chadee K Mucin and Toll-like receptors in host defense against intestinal parasites Trends Parasitol 2003 19 305 311 12855381 10.1016/S1471-4922(03)00122-3 Moniaux N Escande F Porchet N Aubert JP Batra SK Structural organization and classification of the human mucin genes Front Biosci 2001 6 D1192 206 11578969 Verhoeven K Van Laer L Kirschhofer K Legan PK Hughes DC Schatteman I Verstreken M Van Hauwe P Coucke P Chen A Smith RJ Somers T Offeciers FE Van de Heyning P Richardson GP Wachtler F Kimberling WJ Willems PJ Govaerts PJ Van Camp G Mutations in the human alpha-tectorin gene cause autosomal dominant non-syndromic hearing impairment Nat Genet 1998 19 60 62 9590290 Legan PK Rau A Keen JN Richardson GP The mouse tectorins. Modular matrix proteins of the inner ear homologous to components of the sperm-egg adhesion system J Biol Chem 1997 272 8791 8801 9079715 10.1074/jbc.272.13.8791 Harada N Iijima S Kobayashi K Yoshida T Brown WR Hibi T Oshima A Morikawa M Human IgGFc binding protein (FcgammaBP) in colonic epithelial cells exhibits mucin-like structure J Biol Chem 1997 272 15232 15241 9182547 10.1074/jbc.272.24.15232 Kobayashi K Blaser MJ Brown WR Identification of a unique IgG Fc binding site in human intestinal epithelium J Immunol 1989 143 2567 2574 2529312 Kobayashi K Hamada Y Blaser MJ Brown WR The molecular configuration and ultrastructural locations of an IgG Fc binding site in human colonic epithelium J Immunol 1991 146 68 74 1984453 Bond JS Beynon RJ Mammalian metalloendopeptidases Int J Biochem 1985 17 565 574 3896890 10.1016/0020-711X(85)90287-3 Cismasiu VB Denes SA Reilander H Michel H Szedlacsek SE The MAM (meprin/A5-protein/PTPmu) domain is a homophilic binding site promoting the lateral dimerization of receptor-like protein tyrosine phosphatase mu J Biol Chem 2004 15084579 Marchand P Volkmann M Bond JS Cysteine mutations in the MAM domain result in monomeric meprin and alter stability and activity of the proteinase J Biol Chem 1996 271 24236 24241 8798668 10.1074/jbc.271.39.24236 Rise ML von Schalburg KR Brown GD Mawer MA Devlin RH Kuipers N Busby M Beetz-Sargent M Alberto R Gibbs AR Hunt P Shukin R Zeznik JA Nelson C Jones SR Smailus DE Jones SJ Schein JE Marra MA Butterfield YS Stott JM Ng SH Davidson WS Koop BF Development and application of a salmonid EST database and cDNA microarray: data mining and interspecific hybridization characteristics Genome Res 2004 14 478 490 14962987 10.1101/gr.1687304 Thorsen J Zhu B Frengen E Osoegawa K de Jong PJ Koop BF Davidson WS Hoyheim B A highly redundant BAC library of Atlantic salmon (Salmo salar): an important tool for salmon projects BMC Genomics 2005 6 50 15807896 10.1186/1471-2164-6-50 Schwartz S Zhang Z Frazer KA Smit A Riemer C Bouck J Gibbs R Hardison R Miller W PipMaker--a web server for aligning two genomic DNA sequences Genome Res 2000 10 577 586 10779500 10.1101/gr.10.4.577 Schultz J Milpetz F Bork P Ponting CP SMART, a simple modular architecture research tool: identification of signaling domains Proc Natl Acad Sci U S A 1998 95 5857 5864 9600884 10.1073/pnas.95.11.5857 Julenius K Molgaard A Gupta R Brunak S Prediction, conservation analysis, and structural characterization of mammalian mucin-type O-glycosylation sites Glycobiology 2005 15 153 164 15385431 10.1093/glycob/cwh151 Apweiler R Attwood TK Bairoch A Bateman A Birney E Biswas M Bucher P Cerutti L Corpet F Croning MD Durbin R Falquet L Fleischmann W Gouzy J Hermjakob H Hulo N Jonassen I Kahn D Kanapin A Karavidopoulou Y Lopez R Marx B Mulder NJ Oinn TM Pagni M Servant F Sigrist CJ Zdobnov EM InterPro--an integrated documentation resource for protein families, domains and functional sites Bioinformatics 2000 16 1145 1150 11159333 10.1093/bioinformatics/16.12.1145 Burge CB Karlin S Finding the genes in genomic DNA Curr Opin Struct Biol 1998 8 346 354 9666331 10.1016/S0959-440X(98)80069-9 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 Sanger Center Zebrafish assembly information Fugu Genomics Project Gilligan P Brenner S Venkatesh B Fugu and human sequence comparison identifies novel human genes and conserved non-coding sequences Gene 2002 294 35 44 12234665 10.1016/S0378-1119(02)00793-X Hickox JR Bi M Hardy DM Heterogeneous processing and zona pellucida binding activity of pig zonadhesin J Biol Chem 2001 276 41502 41509 11526117 10.1074/jbc.M106795200 Bi M Wassler MJ Hardy DM Hardy DM Sperm adhesion to the extracellular matrix of the egg Fertilization 2002 1 San Diego, Academic Press 153 180 Harris CL Mizuno M Morgan BP Complement and complement regulators in the male reproductive system Mol Immunol 2006 43 57 67 16023728 10.1016/j.molimm.2005.06.026 Riley-Vargas RC Lanzendorf S Atkinson JP Targeted and restricted complement activation on acrosome-reacted spermatozoa J Clin Invest 2005 115 1241 1249 15849610 10.1172/JCI200523213 Children's Hospital Oakland Research Institute Ewing B Green P Base-calling of automated sequencer traces using phred. II. Error probabilities Genome Res 1998 8 186 194 9521922 Ewing B Hillier L Wendl MC Green P Base-calling of automated sequencer traces using phred. I. Accuracy assessment Genome Res 1998 8 175 185 9521921 University of Washington Genome Center Gordon D Abajian C Green P Consed: a graphical tool for sequence finishing Genome Res 1998 8 195 202 9521923 Sonnhammer EL Durbin R A dot-matrix program with dynamic threshold control suited for genomic DNA and protein sequence analysis Gene 1995 167 GC1 10 8566757 10.1016/0378-1119(95)00714-8 RepeatMasker Jurka J Repbase update: a database and an electronic journal of repetitive elements Trends Genet 2000 16 418 420 10973072 10.1016/S0168-9525(00)02093-X Florea L Hartzell G Zhang Z Rubin GM Miller W A computer program for aligning a cDNA sequence with a genomic DNA sequence Genome Res 1998 8 967 974 9750195 Hames BD Higgins SJ Rickwood DHBD Gene Probes 2 The Practical Approach Series 1995 New York, Oxford University Press Inc. Chenna R Sugawara H Koike T Lopez R Gibson TJ Higgins DG Thompson JD Multiple sequence alignment with the Clustal series of programs Nucleic Acids Res 2003 31 3497 3500 12824352 10.1093/nar/gkg500 Kumar S Tamura K Nei M MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment Brief Bioinform 2004 5 150 163 15260895 10.1186/1471-2105-5-150	16303057	PMC1325057	CC BY	2021-01-04 16:32:47	no	BMC Genomics. 2005 Nov 22; 6:165	utf-8	BMC Genomics	2,005	10.1186/1471-2164-6-165	oa_comm
==== Front BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2921633666510.1186/1471-2105-6-292SoftwareSNP-VISTA: An interactive SNP visualization tool Shah Nameeta [email protected] Michael V [email protected] Simon [email protected] Len A [email protected] Philip [email protected] Bernd [email protected] Inna L [email protected] Institute for Data Analysis and Visualization, (IDAV), Department of Computer Science, University of California, Davis, One Shields Ave., Davis, CA 95616, USA2 Genomics Division, Lawrence Berkeley National Laboratory, One Cyclotron Road, Berkeley, CA, 94720, USA3 DOE Joint Genome Institute, 2800 Mitchell Drive, Walnut Creek, CA 94598, USA2005 8 12 2005 6 292 292 4 7 2005 8 12 2005 Copyright © 2005 Shah et al; licensee BioMed Central Ltd.2005Shah 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 advances in sequencing technologies promise to provide a better understanding of the genetics of human disease as well as the evolution of microbial populations. Single Nucleotide Polymorphisms (SNPs) are established genetic markers that aid in the identification of loci affecting quantitative traits and/or disease in a wide variety of eukaryotic species. With today's technological capabilities, it has become possible to re-sequence a large set of appropriate candidate genes in individuals with a given disease in an attempt to identify causative mutations. In addition, SNPs have been used extensively in efforts to study the evolution of microbial populations, and the recent application of random shotgun sequencing to environmental samples enables more extensive SNP analysis of co-occurring and co-evolving microbial populations. The program is available at [1]. Results We have developed and present two modifications of an interactive visualization tool, SNP-VISTA, to aid in the analyses of the following types of data: A. Large-scale re-sequence data of disease-related genes for discovery of associated and/or causative alleles (GeneSNP-VISTA). B. Massive amounts of ecogenomics data for studying homologous recombination in microbial populations (EcoSNP-VISTA). The main features and capabilities of SNP-VISTA are: 1) mapping of SNPs to gene structure; 2) classification of SNPs, based on their location in the gene, frequency of occurrence in samples and allele composition; 3) clustering, based on user-defined subsets of SNPs, highlighting haplotypes as well as recombinant sequences; 4) integration of protein evolutionary conservation visualization; and 5) display of automatically calculated recombination points that are user-editable. Conclusion The main strength of SNP-VISTA is its graphical interface and use of visual representations, which support interactive exploration and hence better understanding of large-scale SNP data by the user. ==== Body Background Polymorphisms are nucleotide differences in genomic DNA sequences that naturally occur within a population. A single nucleotide substitution is called a single nucleotide polymorphism (SNP) and these variants occur at a frequency of approximately one every 1,000 bases in humans [2]. SNPs are established genetic markers that aid in the identification of loci affecting quantitative traits and/or disease in a wide variety of eukaryote species. The recent completion of a single version of the human genome [3,4] has now provided the substrates for direct comparison of individuals in both health and disease. Ideally, to better understand the genetic contributions to severe diseases, one would obtain the entire human genome sequence for all disease-carrying individuals for comparison to unaffected control groups. While these complete datasets are not readily obtainable today, a strategy that is currently approachable is the re-sequencing of a large set of appropriate candidate genes in individuals with a given disease to screen for potential causative/susceptibility alleles. However, one ongoing challenge remains in the visualization of such datasets by end users to thereby derive biological insights. In addition, SNPs have been used extensively in efforts to study the evolution of microbial populations. Such efforts have largely been confined to multi-locus sequence typing of clinical isolates of species such as Neisseria meningitidis and S.taphylococcus aureus [5]. However, the recent application of random shotgun sequencing to environmental samples [6-8] enable more extensive SNP analysis of co-occurring and co-evolving microbial populations. An intriguing finding in one study [6] was the mosaic nature of the genomes of an archaeal population inferred to be the result of extensive homologous recombination of three ancestral strains. This observation was based on the manual analysis of a small subset of the data (ca. 40,000 basepairs) and remains to be verified across the whole genome. Tools to analyze these type of data are also in their infancy. Manipulation, cross-referencing, and haplotype viewing of SNP data are essential for quality assessment and identification of variants associated with genetic disease [9]. The display and interpretation of large genotype data sets can be simplified by using a graphical display. Several software tools have been developed to assist researchers to carry out this task. A visual genotype (VG2) display [10,11] proved to be useful in showing raw datasets of individuals' genotype data. This format presents all data in an array of samples (rows) × polymorphic sites (columns) and encodes each diallelic polymorphism according to a general color scheme. This array format allows one to visually inspect the data across both individual's diplotypes and polymorphic sites to make comparisons and identify correlated sub-datasets. Another program, ViewGene [12], was developed as a flexible tool that utilizes input data and constructs an assembly reference scaffold that can be viewed through a simple graphical interface. Polymorphisms generated from many sources can be added to this scaffold with a variety of options to control what is displayed. Large amounts of polymorphism data can be organized so that patterns and haplotypes can be readily discerned. Two other tools, Genewindow [13] and SNPper [14], were recently developed as Web-based applications to mark variations on the human DNA. A final software system for automated and visual analysis of functionally annotated haplotypes, HapScope [15], displays genomic structure with haplotype information in an integrated environment, providing alternative views for assessing genetic and functional correlation. Although these tools provide a number of valuable options for the scientist, some user requirements are not satisfied. VG2 uses simple but effective representations to show genotype data with SNP classification and organizes the data using hierarchical clustering. The major drawbacks of this tool are its static display, lack of provision for details on demand and lack of capabilities to map SNPs to genomic structure. ViewGene provides a simple interface for analyzing sequence data to locate regions favorable to re-sequencing but is limited in its capabilities for post-processing of SNP data. HapScope consists of valuable haplotype analysis methods along with interactive visualization, but its major focus is the presentation of results from haplotype analysis. Our goal was to develop exploration tools for discovery of disease-related mutations from re-sequencing data. It is important to note that most experiments in SNP research are exploratory in nature, and it has become essential to provide the scientific community with an advanced SNP exploration tools. With SNP data growing as a result of large-scale gene re-sequencing and ecogenomics projects, there exists a need to overcome limitations of current SNP analysis tools. We present an interactive visualization tool, which aids scientists in generating hypotheses from large-scale SNP data. Implementation SNP-VISTA is implemented as a stand-alone Java application using JBuilder [16] as a development environment. SNP-VISTA uses clustering software, Levenshtein [17], which is bundled with the package. Automatic recombination points are calculated using a C++ program that can be invoked from the Java application. Results SNP-VISTA is available in two versions, as GeneSNP-VISTA or EcoSNP-VISTA, each tailored for a specific application. We describe each version in the next two sections. GeneSNP-VISTA: Visualization of disease-related mutations in genes We use the ABO blood group gene (transferase A, alpha 1-3-N-acetylgalactosaminyltransferase; transferase B, alpha 1.3.galactosyltransferase) from the Finished Genes Page of SeattleSNPs [18] to demonstrate our tool's capabilities. The program requires the following files as input: Reference sequence This file contains the DNA sequence of the gene in FASTA format [19]. Annotation file A tab-delimited file with annotation for exons and coding sequence (cds) in the following format: <exon/cds><tab><start><tab><end> If the coding sequence is not specified explicitly then exons are merged to obtain the coding sequence. SNP data A tab-delimited file with four fields on each line, in the format: <Site Position><tab><Sample ID><tab><Allele 1><tab><Allele 2> Protein alignment This file contains the protein alignment in multi-FASTA format. The first protein in the file must be the protein corresponding to the gene given in the reference sequence. Sample input files are available on the website [20]. GeneSNP-VISTA supports the following applications: Mapping of SNPs to gene structure A SNP can be located in a UTR, exon, intron or splice site. Such information about the location of SNPs is valuable to biologists. To illustrate this feature, we mapped SNPs with the ABO dataset to its gene structure as shown in Figure 1A. A coordinate bar represents the ABO blood group gene, which is 23,758 base pairs long and has seven exons that are shown by blue rectangles. The red rectangle is the user-selected sub-region of the gene. Green lines show the exact location of each SNP in the gene. On mouse over the connecting line is highlighted in red. Classification of SNPs A SNP can be found in several genotypic forms in a given individual (homozygous or heterozygous for either of the two nucleotide states). In addition, SNPs in coding sequence can come in two forms; synonymous (those that do not change an amino acid) and non-synonymous (those that do change an amino acid)]. Our program bins SNPs into these various categories and uses different colors for each class of SNPs The graphical representation is similar to VG2, where selected data is represented as an array of samples (rows) × polymorphic sites (columns), and each cell is colored depending on the classification of a given SNP based on their location in the gene, frequency of occurrence in samples and allele composition (Figure 1B). Upon mouse over, detailed information (sample id, position, frequency, etc.) about the selected SNP is displayed in a semi-transparent callout. Figure 1 GeneSNP-VISTA screenshot for the ABO blood group gene (transferase A, alpha 1-3-N-acetylgalactosaminyltransferase; transferase B, alpha 1.3.galactosyltransferase). A. Coordinate bar showing gene structure. The ABO gene consists of 23,758 basepairs. Seven exons are displayed as blue rectangles. The red rectangle is a user-selected region. B. SNPs are represented as an array of samples (rows) × polymorphic sites (columns), where each cell is colored based on the SNP classification. Blue is used for individuals homozygous for the common SNP allele, yellow color is used for individuals homozygous for the rare SNP allele, red is for individuals heterozygous for the SNP, and a black dot is used for non-synonymous SNP. C. Clustering results are shown as a hierarchical tree, where each node can be collapsed or expanded. D. Window displaying protein alignment. The display is linked with the non-synonymous SNP selected by the user. Clustering Clustering of samples based on the patterns of SNPs allows a user to navigate through the data. We use Levenshtein software package to perform hierarchical clustering. Clustering can be performed using all the SNPs in the data or a user-selected subset. GeneSNP-VISTA displays the hierarchical tree (Figure 1C) where each node can be collapsed or expanded. Integration of multiple alignments of homologous proteins in different species One of the approaches to assess how functional a SNP might be when it changes an amino acid is to investigate the conservation of that amino acid across multiple additional species. A SNP causing change in a conserved amino acid would be predicted to be more likely to be a functional change since over evolutionary time this amino acid has resisted genetic drift. Integration of multiple alignments of homologous proteins allows a scientist to determine whether a SNP has caused a conserved amino acid to change. GeneSNP-VISTA displays the protein alignment along with an entropy or sum-of-pairs similarity score in the protein alignment window (Figure 1D). When a user selects a non-synonymous SNP, the corresponding amino acid is highlighted in green. In Figure 1, the user has selected a heterozygous non-synonymous SNP in the last exon, which changes the amino acid Phenylalanine (F) to Isoleucine (I). The protein alignment window shows the conservation of this amino acid, which is 100% conserved. Consistent with our finding, the SIFT program [21] predicts this position as intolerant, and the Polyphen program [22] deems it as probably damaging change (see results at [23]. EcoSNP-VISTA: Discovery of recombination points in microbial populations As our test bed, we have used the acid mine drainage [6] dataset that is publicly available at [24]. The following files are needed as input: Alignment data This file contains the blast output obtained by blasting the consensus sequence against all reads in the database. Annotation file This file is similar to the GeneSNP-VISTA annotation file, and it has the following format: <exon/cds><tab><start><tab><end> Recombination points (optional) A tab-delimited file with four fields on each line, in the format: <Read name><tab><Position> Sample input files are available at [25]. The following modifications are made to GeneSNP-VISTA to handle ecogenomics data: Nucleotide-based color scheme Each cell in the array is colored based on the nucleotide at the SNP position. Once the reads are clustered this representation allows a user to discern various SNP patterns probably corresponding to different strains (Figure 2A). Figure 2 EcoSNP-VISTA screenshot of scaffold 1 of the microbial genome of ferroplasma II [6]. A. SNPs are represented as an array of reads (rows) × polymorphic sites (columns), where each cell is colored based on the nucleotide. Red is used for nucleotide T (Thyamine), blue is used for nucleotide A (Adenine), yellow is used for nucleotide C (Cytosine), and green is used for nucleotide G (Guanine). B. Coordinate bar providing global view of recombination points shown with blue lines and frequency of SNPs, where black indicates higher frequency. C. Array representation showing exact position of the recombination point with two black triangles. D. Window displaying the BLAST alignment for the selected region. Recombination point calculation and visualization A user can provide recombination points, obtained from another program or calculate them by EcoSNP-VISTA. The recombination point calculation is based on the Bellerophon program [26]. Our tool displays recombination points on the coordinate bar using blue lines showing the global view and the frequency of the SNPs (Figure 2B). The array representation also shows the exact position of the recombination point with two black triangles (Figure 2C). The reads can be examined closely in a window as shown in Figure 2D. A user can visually verify the recombination points and accept or reject them. It is also possible to add a recombination point. Automatic recombination point calculation results typically in a large number of false positives, whereas manual detection of recombination points is a very time-consuming job. EcoSNP-VISTA combines both approaches to provide a feasible method for detecting recombination points. Discussion A major challenge in examining human SNP data is assessing which variants are more likely to be involved in having damaging effects on the structure and function of a gene/protein. GeneSNP-VISTA is an interactive visual tool for highly efficient analysis of large amounts of SNP data to determine a subset that may be of relevance to explaining human disease. As shown in Figure 1, all the information about a SNP (type, location on genomic structure, frequency of occurrence, type of amino acid change and the positions conservation) allows a scientist to determine whether a SNP is a possible causative mutation. By providing a visually integrated representation of SNPs data with genomic structure and protein conservation, GeneSNP-VISTA facilitates the screening of causative mutations from re-sequencing of a large set of appropriate candidate genes in individuals with a given disease. Adaptation of existing computational methods and development of new ones for effective SNP analysis of co-occurring and co-evolving microbial populations from ecogenomics data poses new challenges. Manual analysis [6] led to interesting results, but such an analysis is time-intensive and becomes prohibitive for whole genome-scale exploration. Automatic methods are not available yet for such an analysis. As an alternative, EcoSNP-VISTA provides a visual interface for semi-automatic analysis of SNPs data from ecogenomics data. As shown in Figure 2, a compact color-coded representation of SNPs data allows a scientist to manually detect recombination points and visually verify automatically calculated recombination points. EcoSNP-VISTA provides insight into homologous recombination in microbial populations and has the potential to guide in the development of computational methods for such analysis. Conclusion We have developed SNP-VISTA, a publicly available interactive visualization tool that assists scientists in the analysis of re-sequence data of disease-related genes for discovery of associated and/or causative alleles and ecogenomics data for studying homologous recombination in microbial populations. Availability and requirements • Project name: SNP-VISTA • Project home page: • Operating system(s): MacOSX, Windows, Linux • Programming language: Java • Other requirements: e.g. Java 1.4 or higher • License: Open Source License • Any restrictions to use by non-academics: None • Support group: [email protected] Authors' contributions NS developed the algorithms and designed the prototype of the program; MT and SM coded and integrated the complete package; LP and PH provided biological insight and actively participated in discussion of the project and writing the paper; BH guided NS at different stages of the project; ID designed and led the project. Acknowledgements This work was performed under the auspices of the US Department of Energy's Office of Science, Biological and Environmental Research Program, and by the University of California, Lawrence Berkeley National Laboratory under contract No. DE-AC02-05CH11231. The project was partly supported by the Programs for Genomic Applications grant from the NHLBI/NIH. Additional support for this project was provided by the Laboratory Directed Research and Development (LDRD) Program, through the project entitled "Interactive Visualization Methods for Exploration and Comparison of Multi-billion Base Pair Sequence Data." We thank the members of the Visualization and Computer Graphics Research Group at the Institute for Data Analysis and Visualization (IDAV) at the University of California, Davis. ==== Refs SNP-VISTA Program Sachidanandam R Weissman D Schmidt SC Kakol JM Stein LD Marth G Sherry S Mullikin JC Mortimore BJ Willey DL Hunt SE Cole CG Coggill PC Rice CM Ning Z Rogers J Bentley DR Kwok PY Mardis ER Yeh RT Schultz B Cook L Davenport R Dante M Fulton L Hillier L Waterston RH McPherson JD Gilman B Schaffner S Van Etten WJ Reich D Higgins J Daly MJ Blumenstiel B Baldwin J Stange-Thomann N Zody MC Linton L Lander ES Altshuler D A map of human genome sequence variation containing 1.42 million single nucleotide polymorphisms Nature 2001 409 928 33 11237013 10.1038/35057149 Venter JC Adams MD Myers EW Li PW Mural RJ Sutton GG Smith HO Yandell M Evans CA Holt RA Gocayne JD Amanatides P Ballew RM Huson DH Wortman JR Zhang Q Kodira CD Zheng XH Chen L Skupski M Subramanian G Thomas PD Zhang J Gabor Miklos GL Nelson C Broder S Clark AG Nadeau J McKusick VA Zinder N Levine AJ Roberts RJ Simon M Slayman C Hunkapiller M Bolanos R Delcher A Dew I Fasulo D Flanigan M Florea L Halpern A Hannenhalli S Kravitz S Levy S Mobarry C Reinert K Remington K Abu-Threideh J Beasley E Biddick K Bonazzi V Brandon R Cargill M Chandramouliswaran I Charlab R Chaturvedi K Deng Z Di Francesco V Dunn P Eilbeck K Evangelista C Gabrielian AE Gan W Ge W Gong F Gu Z Guan P Heiman TJ Higgins ME Ji RR Ke Z Ketchum KA Lai Z Lei Y Li Z Li J Liang Y Lin X Lu F Merkulov GV Milshina N Moore HM Naik AK Narayan VA Neelam B Nusskern D Rusch DB Salzberg S Shao W Shue B Sun J Wang Z Wang A Wang X Wang J Wei M Wides R Xiao C Yan C Yao A Ye J Zhan M Zhang W Zhang H Zhao Q Zheng L Zhong F Zhong W Zhu S Zhao S Gilbert D Baumhueter S Spier G Carter C Cravchik A Woodage T Ali F An H Awe A Baldwin D Baden H Barnstead M Barrow I Beeson K Busam D Carver A Center A Cheng ML Curry L Danaher S Davenport L Desilets R Dietz S Dodson K Doup L Ferriera S Garg N Gluecksmann A Hart B Haynes J Haynes C Heiner C Hladun S Hostin D Houck J Howland T Ibegwam C Johnson J Kalush F Kline L Koduru S Love A Mann F May D McCawley S McIntosh T McMullen I Moy M Moy L Murphy B Nelson K Pfannkoch C Pratts E Puri V Qureshi H Reardon M Rodriguez R Rogers YH Romblad D Ruhfel B Scott R Sitter C Smallwood M Stewart E Strong R Suh E Thomas R Tint NN Tse S Vech C Wang G Wetter J Williams S Williams M Windsor S Winn-Deen E Wolfe K Zaveri J Zaveri K Abril JF Guigo R Campbell MJ Sjolander KV Karlak B Kejariwal A Mi H Lazareva B Hatton T Narechania A Diemer K Muruganujan A Guo N Sato S Bafna V Istrail S Lippert R Schwartz R Walenz B Yooseph S Allen D Basu A Baxendale J Blick L Caminha M Carnes-Stine J Caulk P Chiang YH Coyne M Dahlke C Mays A Dombroski M Donnelly M Ely D Esparham S Fosler C Gire H Glanowski S Glasser K Glodek A Gorokhov M Graham K Gropman B Harris M Heil J Henderson S Hoover J Jennings D Jordan C Jordan J Kasha J Kagan L Kraft C Levitsky A Lewis M Liu X Lopez J Ma D Majoros W McDaniel J Murphy S Newman M Nguyen T Nguyen N Nodell M Pan S Peck J Peterson M Rowe W Sanders R Scott J Simpson M Smith T Sprague A Stockwell T Turner R Venter E Wang M Wen M Wu D Wu M Xia A Zandieh A Zhu X The sequence of the human genome Science 2001 291 1304 51 11181995 10.1126/science.1058040 The Human Genome Consortium Initial sequencing and analysis of the human genome Nature 2001 409 860 921 11237011 10.1038/35057062 Spratt BG Zhang Q Jones DM Hutchison A Brannigan JA Dowson CG Recruitment of a Penicillin-Binding Protein Gene from Neisseria flavescens during the Emergence of Penicillin Resistance in Neisseria meningitidis PNAS 1989 86 8988 8992 2510173 Tyson GW Chapman J Hugenholtz P Allen EE Ram RJ Richardson PM Solovyev VV Rubin EM Rokhsar DS Banfield JF Community structure and metabolism through reconstruction of microbial genomes from the environment Nature 2004 428 37 43 14961025 10.1038/nature02340 Venter JC Remington K Heidelberg JF Halpern AL Rusch D Eisen JA Wu D Paulsen I Nelson KE Nelson W Fouts DE Levy S Knap AH Lomas MW Nealson K White O Peterson J Hoffman J Parsons R Baden-Tillson H Pfannkoch C Rogers YH Smith HO Environmental genome shotgun sequencing of the Sargasso Sea Science 2004 304 66 74 15001713 10.1126/science.1093857 Tringe SG von Mering C Kobayashi A Salamov AA Chen K Chang HW Podar M Short JM Mathur EJ Detter JC Bork P Hugenholtz P Rubin EM Comparative metagenomics of microbial communities Science 2005 308 554 7 15845853 10.1126/science.1107851 The International HapMap Consortium The International HapMap Project Nature 2003 426 789 96 14685227 10.1038/nature02168 Nickerson DA Taylor SL Weiss KM Clark AG Hutchinson RG Stengard J Salomaa V Vartiainen E Boerwinkle E Sing CF DNA sequence diversity in a 9.7-kb region of the human lipoprotein lipase gene Nature Genetics 1998 19 233 240 9662394 10.1038/907 Reider MJ Taylor SL Clark AG Nickerson DA Sequence variation in the human angiotensin converting enzyme Nature Genetics 1999 22 59 62 10319862 10.1038/8760 Kashuk C SenGupta S Eichler E Chakravarti A ViewGene: A graphical tool for polymorphism visualization and characterization Genome Research 2002 12 333 8 11827953 10.1101/gr.211202 Riva A Kohane IS SNPper: retrieval and analysis of human SNPs Bioinformatics 2002 18 1681 5 12490454 10.1093/bioinformatics/18.12.1681 Staats B Qi L Beerman M Sicotte H Burdett LA Packer B Chanock SJ Yeager M Genewindow: an interactive tool for visualization of genomic variation Nature Genetics 2005 37 109 110 15678133 10.1038/ng0205-109 Zhang J Rowe WL Struewing JP Buetow KH HapScope: A software system for automated and visual analysis of functionally annotated haplotypes Nucleic Acids Research 2002 30 5213 21 12466546 10.1093/nar/gkf654 Borland : JBuilder Web site RuG/L04 – dialectometrics & cartography Web site SeattleSNPs Web site EBI Help Web site GeneSNP-VISTA program Ng PC Henikoff S Accounting for human polymorphisms predicted to affect protein function Genome Research 2002 12 436 446 11875032 10.1101/gr.212802 Ramensky V Bork P Sunyaev S Human non-synonymous SNPs: server and survey Nucleic Acids Research 2002 30 3894 3900 12202775 10.1093/nar/gkf493 PolyPhen and SIFT results Metagenomics Prototype Web Tools EcoSNP-VISTA program Huber T Faulkner G Hugenholtz P Bellerophon: A program to detect chimeric sequences in multiple sequence alignments Bioinformatics 2004 20 2317 2319 15073015 10.1093/bioinformatics/bth226	16336665	PMC1325058	CC BY	2021-01-04 16:27:46	no	BMC Bioinformatics. 2005 Dec 8; 6:292	utf-8	BMC Bioinformatics	2,005	10.1186/1471-2105-6-292	oa_comm
==== Front BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2911633665010.1186/1471-2105-6-291Methodology ArticleAn SVM-based system for predicting protein subnuclear localizations Lei Zhengdeng [email protected] Yang [email protected] Department of Bioengineering (MC063), University of Illinois at Chicago, 851 South Morgan Street, Chicago IL 60607, USA2005 7 12 2005 6 291 291 9 5 2005 7 12 2005 Copyright © 2005 Lei and Dai; licensee BioMed Central Ltd.2005Lei and Dai; 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 large gap between the number of protein sequences in databases and the number of functionally characterized proteins calls for the development of a fast computational tool for the prediction of subnuclear and subcellular localizations generally applicable to protein sequences. The information on localization may reveal the molecular function of novel proteins, in addition to providing insight on the biological pathways in which they function. The bulk of past work has been focused on protein subcellular localizations. Furthermore, no specific tool has been dedicated to prediction at the subnuclear level, despite its high importance. In order to design a suitable predictive system, the extraction of subtle sequence signals that can discriminate among proteins with different subnuclear localizations is the key. Results New kernel functions used in a support vector machine (SVM) learning model are introduced for the measurement of sequence similarity. The k-peptide vectors are first mapped by a matrix of high-scored pairs of k-peptides which are measured by BLOSUM62 scores. The kernels, measuring the similarity for sequences, are then defined on the mapped vectors. By combining these new encoding methods, a multi-class classification system for the prediction of protein subnuclear localizations is established for the first time. The performance of the system is evaluated with a set of proteins collected in the Nuclear Protein Database (NPD). The overall accuracy of prediction for 6 localizations is about 50% (vs. random prediction 16.7%) for single localization proteins in the leave-one-out cross-validation; and 65% for an independent set of multi-localization proteins. This integrated system can be accessed at . Conclusion The integrated system benefits from the combination of predictions from several SVMs based on selected encoding methods. Finally, the predictive power of the system is expected to improve as more proteins with known subnuclear localizations become available. ==== Body Background The cell nucleus is a highly complex organelle that organizes the comprehensive assembly of our genes and their corresponding regulatory factors. Accordingly, the cell nucleus reflects the intricate regulation of various biological activities. Although protein complexes disperse throughout the entire organelle, it is known that many nuclear proteins participating in related pathways tend to concentrate into specific areas [1,2]. For example, the rDNA processing and ribosome biogenesis often occur within the nucleolus and the proteins responsible for pre-splicing appear to concentrate into multiple nuclear speckles, even while they are migrating in the nucleus. The confinement of biomolecules within specific compartments is crucial for the formation and function of the cell nucleus; in contrast, the mis-localization of proteins can lead to both human genetic disease and cancer [3]. Accordingly, information on protein subnuclear localization is essential for a full understanding of genomic regulation and function. Advances in experimental technology have enabled the large-scale identification of nuclear proteins. However, at the same time, the sequencing of both the human and mouse genomes has generated an enormous inventory of primary sequences with unknown functions. A faster and cheaper bioinformatics tool is required for the annotation of these exponentially accumulating sequences. A computational prediction of protein subnuclear compartments from primary protein sequences can provide important clues to the function of novel proteins. A host of systems for the prediction of protein subcellular localizations has emerged over the last two decades [4-23]. This list includes several web-based predictors that have a broad coverage of subcellular localizations at the genomic level, such as PSORT [4], SubLoc [7], Proteome Analyst [15], CELLO [16], PSORTb v.2.0 [17], and LOCtree [21]. The development led to the ability to predict the particular subcellular compartment, in which a given protein resides within a cell, with a steadily increasing accuracy. The predictions for eukaryotic organisms, however, have certain limitations. They can provide information on whether a protein localizes in the nuclear compartment, but they can not discriminate among the sub-compartments in which it functions. The prediction of protein localization at the subnuclear level is challenging compared with that at the subcellular level. Three facts contribute to the difficulty: (1) proteins within the cell nucleus face no apparent physical barrier like a membrane [24]; (2) the nucleus is far more compact and complicated in comparison with other compartments in a cell [25]; and (3) protein complexes within the cell nucleus are not static [1,24,25]. Recent developments in live-cell imaging have revealed that nuclear processes may rely on a constant flow of molecules between dynamic compartments created by relatively immobile binding or assembly sites. As proteins diffuse through the nuclear space, they appear to alter their compartments during different phases of the cell cycle or accompanying differentiation [3]. For instance, some nucleolar proteins are continually exchanging between the nucleoplasm and the nucleolus. Proteomic studies have also highlighted the dynamic nature of the nucleolar proteome [3]. Employing the database Nuclear Protein Database (NPD) developed by Dellaire, Farrall and Bickmore [26], Bickmore and Sutherland [27] recently addressed the characteristics of the primary sequences of nuclear proteins, such as the molecular weight, isoelectric point, and amino acid composition for proteins in different subnuclear compartments. They also found that motifs and domains are often shared by proteins co-localized within the same subnuclear compartment. Furthermore, certain generally abundant motifs/domains are lacking from the proteins concentrated in some specific areas of the nucleus. Based on these findings, it should be possible to combine totality of this information in a manner that will enhance the prediction of compartmental-specific nuclear localizations of the protein constituents listed in genome databases. Encouraged by our previous success in the design of a metric for the biological similarity of protein sequences [22,23], a prediction system is developed based on support vector machines (SVMs), one of the most advanced machine learning methods [28,29]. The principal feature of our mode of analysis is the introduction of new kernel functions which are effective in capturing the subtle difference between sequences originated from two distinct nuclear compartments. Results and Discussion Normally, conventional k-peptide encoding vectors (k = 1, 2, 3) are used for the description of a protein sequence. Successful applications include (1) the protein fold recognition [30,31], and (2) the prediction of subcellular localization [5,7,16]. The basic concept of the new kernels proposed in our previous work [22,23] is the measurement of biological similarity for k-peptides, having either none or a few shared residues, with the incorporation of evolutionary information. Our finding indicates that the mapping of conventional k-peptide encoding vectors by a matrix formed with high-scored pairs of k-peptides can facilitate the construction of a suitable metric. The score of a pair of k-peptides is calculated by the BLOSUM scores of residues and, therefore, the evolutionary information of the residues is embedded into the sequence description. A related concept that links two k-peptides with a small number of mutated residues has been presented by Leslie et al. [32] for protein homology detection. This study presents the performance of conventional k-peptide encoding methods and the new proposed kernels for the prediction of protein subnuclear compartments. Furthermore, with the use of the jury voting scheme developed in [31], an integrated system was built by combining binary prediction outcomes obtained from different sequence encoding schemes. The results demonstrate that the integrated system enhances the overall performance of the system. The dataset used in this study was extracted from the Nuclear Protein Database (NPD) [26] using a Perl script. The NPD is a curated database that stores information on more than 1000 vertebrate proteins, chiefly from human and mouse, which are reported in the literature to be localized in the cell nucleus. Since certain proteins associate with more than one compartment, a test dataset consisting of proteins with multiple localizations was first extracted out. These proteins have the same SwissProt ID or Entrez Protein ID though localized in different compartments. This preparative procedure resulted in 92 proteins that are localized within the six compartments described below. The majority is localized in 2 compartments and the remaining portion is localized in 3 or 4 compartments. After excluding the multi-localization proteins, a non-redundant dataset was further constructed by PROSET [33] to ensure low sequence identity (<50%). In order to have sufficient number of proteins for training and testing, only six localizations were selected for evaluation. These are PML BODY (38), Nuclear Lamina (55), Nuclear Splicing Speckles (56), Chromatin (61), Nucleoplasm (75), and Nucleolus (219). Each of these proteins has a single localization and the total number is 504. It should be noted that the multi-localization proteins are not included in the set of 504 single-localization proteins for the leave-one-out cross-validation (LOOCV). Therefore, the multi-localization dataset is essentially an independent testing set. The summary of the datasets is presented in Table 1. Table 1 The summary of the nuclear proteins Label Compartment Number of sequences 1 PML BODY 38 2 Nuclear Lamina 55 3 Nuclear Splicing Speckles 56 4 Chromatin 61 5 Nucleoplasm 75 6 Nucleolus 219 - Multiple Localizations 92 AA – amino acid composition encoding method; DI – di-peptide encoding method; TRI – tri-peptide encoding method; D1X1 – amino acid composition encoding vector transformed with D1; D2X2 – di-peptide encoding vector transformed with D2; D3X3 – tri-peptide encoding vector transformed with D3. The evaluations of the predictive power of the methods were performed on the datasets. Since there are 6 localizations in the dataset, the one-versus-one multi-class classification system led to 6(6-1)/2 = 15 SVM models for one single encoding method (see Methods for details). Three encoding techniques corresponding to the conventional k-peptide composition and three encoding methods based on the new kernels were used for k = 1,2,3. SVMLight [34] was used as the SVM solver. The overall accuracy for the multi-class classification proposed by Rost and Sander [35] was used for the evaluation of our system. Suppose there are m = m1 + m2 + ... + mN test proteins, where mi is the number of proteins belonging to class i(i = 1,...,N). Suppose further that out of the proteins considered, pi proteins are correctly predicted to belong to class i. Then p = p1 + p2 + ... + pN is the total number of correctly predicted proteins. The accuracy for class i is acci = pimi, MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciGacaGaaeqabaqabeGadaaakeaacqWGHbqycqWGJbWycqWGJbWydaWgaaWcbaGaemyAaKgabeaakiabbccaGiabg2da9iabbccaGmaalaaabaGaemiCaa3aaSbaaSqaaiabdMgaPbqabaaakeaacqWGTbqBdaWgaaWcbaGaemyAaKgabeaaaaGccqqGSaalaaa@3B99@ and the overall accuracy, denoted by Qacc, is defined as Qacc=∑i=1Nacci×mim=∑i=1Npim=pm. MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciGacaGaaeqabaqabeGadaaakeaacqWGrbqudaWgaaWcbaGaemyyaeMaem4yamMaem4yamgabeaakiabg2da9maaqahabaGaemyyaeMaem4yamMaem4yam2aaSbaaSqaaiabdMgaPbqabaGccqGHxdaTdaWcaaqaaiabd2gaTnaaBaaaleaacqWGPbqAaeqaaaGcbaGaemyBa0gaaaWcbaGaemyAaKMaeyypa0JaeGymaedabaGaemOta4eaniabggHiLdGccqGH9aqpdaaeWbqaamaalaaabaGaemiCaa3aaSbaaSqaaiabdMgaPbqabaaakeaacqWGTbqBaaGaeyypa0ZaaSaaaeaacqWGWbaCaeaacqWGTbqBaaaaleaacqWGPbqAcqGH9aqpcqaIXaqmaeaacqWGobGta0GaeyyeIuoakiabc6caUaaa@56B3@ Note that acci and Qacc are respectively corresponding to the definitions of Qi%obs MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciGacaGaaeqabaqabeGadaaakeaacqWGrbqudaqhaaWcbaGaemyAaKgabaGaeiyjauIaem4Ba8MaemOyaiMaem4Camhaaaaa@3456@ and Qtotal in Rost and Sander [35]. Since the numbers of proteins for various localizations are unbalanced, the Matthew's correlation coefficient (MCC) was also employed for the optimization of parameters and evaluation of performance [36]: MCCi=pisi−uioi(pi+ui)(pi+oi)(si+ui)(si+oi), MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=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@62BA@ where pi is the number of correctly predicted proteins of the location i, si is the number of correctly predicted proteins not in the location i, ui is the number of under-predicted proteins, and oi the number of over-predicted proteins. In order to evaluate the performance of the system for multi-localization proteins, the criterion proposed in Gardy et al. was used [17]. More specifically, for a protein with multi-localization, if the system validly predicts one of the locations, then the entire prediction is considered correct. It should be noted that this criterion overestimates the performance. Since our method can only predict one localization for a given protein, other evaluation methods for multi-localization proteins such as the one proposed by Chou and Cai [14,18] can not be applied. The performances for each encoding method and the combined encoding methods are shown in Table 2 and Table 3, respectively. The results for the single-localization proteins were obtained from the LOOCV procedure; and the results for the multi-localization proteins were obtained from the final prediction system. Overall, the single encoding methods gave an accuracy of prediction Qacc that ranged from 47.8% to 51.4% for single-localization proteins and from 57.6% to 64.1% for multi-localization proteins. The corresponding average MCCs ranged from 0.203 to 0.276 for single-localization proteins and from 0.182 to 0.401 for multi-localization proteins. The combination of the new encoding methods D1X1, D2X2, and D3X3 with the use of jury voting yielded an improved performance for MCC. For example, the average MCC was elevated from 0.266–0.276 to 0.284 for single-localization proteins and from 0.362–0.401 to 0.420 for multi-localization proteins. The change in Qacc was not uniform: it decreased from the highest value 51.4% to 50.0% for single-localization protein and increased from 64.1% to 65.2% for multi-localization proteins. The combination of the conventional k-peptide compositions AA, DI, and TRI did not demonstrate significant improvement. Further optimization of the parameter for the determination of sparsity of matrix D3 is likely to enhance the performance of the prediction system. Table 2 Results for each individual encoding method Method AA DI TRI D1X1 D2X2 D3X3 Compartment Accuracy % [MCC] PML BODY 26.3 [0.144] 13.2 [0.091] 0.0 [-0.045] 31.6 [0.183] 29.0 [0.139] 10.5 [0.066] Nuclear Lamina 40.0 [0.363] 27.3 [0.256] 40.0 [0.228] 45.5 [0.340] 41.8 [0.279] 36.4 [0.331] Nuclear Splicing Speckles 30.4 [0.326] 32.1 [0.358] 30.4 [0.365] 33.9 [0.321] 33.9 [0.316] 33.9 [0.391] Chromatin 14.8 [0.174] 11.5 [0.106] 13.1 [0.191] 19.8 [0.215] 21.3 [0.248] 21.3 [0.271] Nucleoplasm 25.3 [0.189] 26.7 [0.207] 12.0 [0.123] 20.0 [0.182] 22.7 [0.246] 28.0 [0.229] Nucleolus 78.1 [0.374] 83.1 [0.357] 85.8 [0.357] 73.5 [0.357] 72.2 [0.364] 83.1 [0.367] Single-localization Overall Accuracy and MCC 49.2 [0.262] 49.0 [0.229] 48.4 [0.203] 48.4 [0.266] 47.8 [0.265] 51.4 [0.276] Multi-localization Overall Accuracy and MCC 64.1 [0.365] 57.6 [0.343] 58.7 [0.182] 60.9 [0.401] 57.6 [0.362] 64.1 [0.362] AA – amino acid composition encoding method; DI – di-peptide encoding method; TRI – tri-peptide encoding method; D1X1 – amino acid composition encoding vector transformed with D1; D2X2 – di-peptide encoding vector transformed with D2; D3X3 – tri-peptide encoding vector transformed with D3. Table 3 Results using combined methods Methods Combination of AA, DI, TRI Combination of D1X1, D2X2, and D3X3 Compartment Accuracy % [MCC] PML BODY 13.2 [0.073] 29.0 [0.172] Nuclear Lamina 30.9 [0.275] 43.6 [0.338] Nuclear Splicing Speckles 32.1 [0.410] 35.7 [0.363] Chromatin 9.8 [0.170] 19.7 [0.260] Nucleoplasm 20.0 [0.182] 22.7 [0.206] Nucleolus 88.1 [0.374] 76.7 [0.367] Single-localization Overall Accuracy and MCC 50.4 [0.247] 50.0 [0.284] Multi-localization Overall Accuracy and MCC 62.0 [0.362] 65.2 [0.420] AA – amino acid composition encoding method; DI – di-peptide encoding method; TRI – tri-peptide encoding method; D1X1 – amino acid composition encoding vector transformed with D1; D2X2 – di-peptide encoding vector transformed with D2; D3X3 – tri-peptide encoding vector transformed with D3. The final models for the prediction system are the combination of the new encoding methods D1X1, D2X2, and D3X3, since adding any conventional k-peptide encoding method does not improve the performance of the system. The predictions for all the 92 multi-localization testing proteins are detailed in Table S1 in the supplementary file [see Additional file 1]. Conclusion An SVM-based multi-class classification system has been developed for the prediction of protein subnuclear localizations. This is the first system designed specifically for this task. This system, which integrates predictions from three new encoding methods, achieves encouraging levels of accuracy for six specific subnuclear localizations. However, compared to the prediction of protein localizations at the subcellular level, the corresponding prediction at the subnuclear level is far more challenging. This difficulty arises mainly from the biological fact that each compartment within the cell nucleus contains no apparent physical barrier like a membrane. Furthermore, the nucleus is a considerably more compact and complex organelle in comparison to other organelles in the cell. Finally, the dynamic nature of the nucleolar proteome adds an additional level of complexity to the task of prediction. Methods Kernels based on high-scored pairs of k-peptides Recently, Lei and Dai proposed new kernels based on high-scored pairs of k-peptides for protein sequence encoding [22,23] for the SVMs. Superior performance of the SVMs with these new kernels was demonstrated through application to the prediction of protein subcellular localization. The kernels proposed in [22,23] can be described as follows. A matrix Dk of high scored k-peptide pairs is defined with a prescribed threshold. Each entry is associated with the BLOSUM score of some pair of k-peptides. The matrix is of dimension 21k × 21k, where 21 is the number of amino acid symbols (normal 20 amino acids plus the special symbol ''X''). The thresholds are set to zeroes for k = 1, 2. Therefore, matrix D1 is the same as the BLOSUM matrix, except that the entries with negative values are replaced by zeroes; the entries of matrix D2 are the BLOSUM pair scores of two di-peptides with all negative values being replaced by zeroes. Since the size of D3 is very large and the majority of all possible pairs is associated with lower scores, the elimination of those pairs can reduce noise that may confuse the prediction. Therefore, a careful thresholding is necessary to ensure the sparsity of the matrix D3. In this work, the threshold is set to 8 for k = 3. For example, the score is 12 for an AAA-AAA pair, 11 for an AAY-ACY pair, and 0 for a TVW-TVR pair since TVW-TVR BLOSUM62 pair-score is 6, which is smaller than the threshold value 8. Given the dimensional scaling, when k > 3, such a coding scheme is less attractive from a computational point of view. For a pair of k-peptide composition vectors xki, xkj, the new kernels are defined as K (xki, xkj) = exp(-γ \|\| Dkxki - Dkxkj \|\|2), k = 1, 2, 3, .... It can be considered as a Gaussian kernel for a pair of vectors Dkxki and Dkxkj. These kernels define the sequence similarity for the mapped vectors Dkxki and Dkxkj, not directly for the k-peptide composition vectors xki and xkj. In this study, the kernel type used for the conventional k-peptide composition encoding methods is the radial basis kernel: exp(-γ \|\| xki - xkj \|\|2) In the following, the concept described above is illustrated and the comparison with the conventional k-peptide encoding method is provided. Consider two short amino acid sequences AAACY and AACCY. Using the input format of the SVMLight [34], the conventional tri-peptide encoding method generates two coding vectors: x31: 1:0.33 2:0.33 42:0.33 x32: 2:0.33 23:0.33 483:0.33 where the numbers appearing in the vectors are in the format of "index: score". It is obvious that the two sequences share the tri-peptide "AAC", and the corresponding vector index is 2. On the other hand, using BLOSUM62, the transformed vectors D3x31 for x31 and D3x32 for x32 are calculated as follows: Example of encoding AAACY to D3x31: ACY0000......11......0AAC8170000AAA1280000AAAAACAADAAE......AAY......YYY↓↓↓↓↓↓6.678.3300......3.670 MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=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@9240@ D3x31: 1:6.67 2:8.33 6:2.67 16: 3.00 17:2.67 18:2.67 21: 3.67 22:6.33 23:8.00 24:3.33 25:3.67 26:5.33 27:3.33 28:5.00 29:4.00 30:3.67 ... D3x32: 1:2.67 2:10.00 22:4.33 23:11.67 24:3.33 25:3.00 26:7.67 27:3.33 28:7.00 ... From the list it is seen that the transformed vectors share more common indices, such as 1, 2, 22–28 etc. Therefore, the similarity between the two sequences is more likely to be captured by the new methods even they do not share explicitly those tri-peptides. The mismatch string kernels proposed in Leslie et al. [32] also consider the similarity between mismatch k-peptides. For example, compared with the conventional tri-peptide encoding, the two sequences share several more common tri-peptides, such as AAA and AAC, AAC and ACC, ACY and CCY, if one mismatch is allowed in two peptides. Therefore, our method is related to the mismatch string kernel but it is different. Multi-class classification system The efficient extension of SVMs to the handling of multiple classes has been achieved for applications to protein fold prediction [30] and the prediction of subcellular localization [7,16]. The one-versus-one [37] framework was used here for the assembly of the multi-class classifier from binary classifiers. For a classification problem of N class, it trains every pair-wise binary classifier. This gives a total of 1/2 N (N - 1) classifiers. The prediction of the label of a testing protein follows the jury voting; specifically, sum the predictions for each classifier and take the label with the highest votes. When ties arise, the class label is assigned to the class with the maximum value of the sum of the function margins. This jury voting scheme is very flexible for the assembly of the predictions obtained from various SVM models. It can integrate not only the outcome from binary predictors with one encoding scheme, but also those obtained from alternative encoding methods. Accordingly, the class label of the testing protein is assigned to the class with the maximum votes. Cross-validation and final prediction system The generalization performance of an SVM is controlled by the following parameters: (1) C: the trade-off between the training error and class separation; (2) γ: the parameter in the radial basis functions exp(-γ \|\| xi - xj \|\|2) or exp(-γ \|\| Dkxki - Dkxkj \|\|2); (3) J: the biased penalty for errors from positive and negative training points. The leave-one-out cross-validation (LOOCV) was employed for the evaluation. The LOOCV is also referred as jackknife test, which is considered to be more rigorous and reliable compared with other testing techniques. A justification of the rigorousness and reliability of the LOOCV can be found, e.g., in Chou and Zhang [38]. Assume that there are overall m proteins. Each protein was in turn considered as a testing protein and the parameters associated with the SVM model were optimized based on a 5-fold cross-validation by using the remaining m - 1 proteins. The criterion of the optimization is the sum of the Matthew's correlation coefficients over all classes [36]. The final LOOCV classifiers were determined by using the optimized parameters to train the set of the m - 1 proteins. The search ranges corresponding to the parameters in the 5-fold cross validation optimization are the following: (1) C: 2-2, 2-1, 1, ..., 29, 210; (2) γ: 2-15, 2-14, 2-13, ..., 214, 215; (3) J: 1, 2, 3, ..., 8, 9. The labels of the training sets were arranged in a way that the size of the negative set is always larger than that of the positive set in our experiment. Here, the penalty term C∑iξi MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciGacaGaaeqabaqabeGadaaakeaacqWGdbWqdaaeqbqaaiabe67a4naaBaaaleaacqWGPbqAaeqaaaqaaiabdMgaPbqab0GaeyyeIuoaaaa@347A@ in the SVM is split into two terms: C∑iξi=C∑{i:yi=1}ξi+CJ∑{i:y=−1}ξi MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciGacaGaaeqabaqabeGadaaakeaacqWGdbWqdaaeqbqaaiabe67a4naaBaaaleaacqWGPbqAaeqaaaqaaiabdMgaPbqab0GaeyyeIuoakiabg2da9iabdoeadnaaqafabaGaeqOVdG3aaSbaaSqaaiabdMgaPbqabaaabaGaei4EaSNaemyAaKMaeiOoaOJaemyEaK3aaSbaaWqaaiabdMgaPbqabaWccqGH9aqpcqaIXaqmcqGG9bqFaeqaniabggHiLdGccqGHRaWkcqWGdbWqcqWGkbGsdaaeqbqaaiabe67a4naaBaaaleaacqWGPbqAaeqaaaqaaiabcUha7jabdMgaPjabcQda6iabdMha5jabg2da9iabgkHiTiabigdaXiabc2ha9bqab0GaeyyeIuoaaaa@5885@. The heavier weight CJ imposed on the errors originating from the negative points enforces a low false positive rate for unbalanced training sets [39]. The final prediction system was constructed as follows. The entire set of proteins with single-localization was used as a training set; and the optimal value for each parameter of the SVMs for the training set was taken as the average value of the optimal parameters obtained from the LOOCV procedure. Using these optimized parameters, final binary classifies were learned from the training set. The evaluation for the set of multi-localization proteins was based on this final prediction system. The framework for the overall training and testing procedures is illustrated in Figure S1 in the supplementary file [see Additional file 2]. Availability and requirements Project name: Subnuclear Compartments Prediction System (Version 1.0) Project home page: Operating system(s): Linux Programming language: Perl License: None Any restrictions to use by non-academics: None Authors' contributions ZL designed the methodology and developed the programs. YD contributed with ideas on overall design, implementation, and assisted with drafting the manuscript. Supplementary Material Additional File 1 This file includes Table S1 – Prediction for multi-localization proteins. A correct prediction is counted if one of the localizations is predicted. Click here for file Additional File 2 This file includes Figure S1 – Diagrammatic view of our SVM-based system for the prediction of protein subnuclear localizations. Click here for file Acknowledgements This research was supported in part by National Science Foundation (EIA-022-0301) and Naval Research Laboratory (N00173-03-1-G016). The authors are grateful to Deepa Vijayraghaven for her assistance with the computing environment. We thank anonymous referees for their valuable suggestions. ==== Refs Heidi GES Gail KM Kathryn N Lisa VF Rachel F Graham D Javier FC Wendy AB Large-scale identification of mammalian proteins localized to nuclear sub-compartments Human Molecular Genetics 2001 10 1995 2011 11555636 10.1093/hmg/10.18.1995 Joanna MB Wendy AB Putting the genome on the map Trends Genet 1998 14 403 409 9820029 10.1016/S0168-9525(98)01572-8 Phair RD Misteli T High mobility of proteins in the mammalian cell nucleus Nature 2000 404 604 609 10766243 10.1038/35007077 Nakai K Horton P PSORT: a program for detecting sorting signals in proteins and predicting their subcellular localization Trends in Biochem Sci 1999 24 34 35 10087920 10.1016/S0968-0004(98)01336-X Chou K-C Elrod DW Protein subcellular location prediction Protein Eng 1999 12 107 118 10195282 10.1093/protein/12.2.107 Emanuelsson O Nielsen H Brunak S von Heijne G Predicting subcellular localization of proteins based on their N-terminal amino acid sequence Journal of Molecular Biology 2000 300 1005 10891285 10.1006/jmbi.2000.3903 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 Chou KC Prediction of protein cellular attributes using pseudo amino acid composition PROTEINS: Structure, Function, and Genetics 2001 43 246 255 10.1002/prot.1035 Chou KC Cai YD Using functional domain composition and support vector machines for prediction of protein subcellular location Journal of Biological Chemistry 2002 277 45765 45769 12186861 10.1074/jbc.M204161200 Nair R Rost B Better prediction of sub-cellular localization by combining evolutionary and structural information Proteins: Structure, Function, and Genetics 2003 53 917 930 10.1002/prot.10507 Pan YX Zhang ZZ Guo ZM Feng GY Huang ZD He L Application of pseudo amino acid composition for predicting protein subcellular location: stochastic signal processing approach Journal of Protein Chemistry 2003 22 395 402 13678304 10.1023/A:1025350409648 Zhou GP Doctor K Subcellular location prediction of apoptosis proteins PROTEINS: Structure, Function, and Genetics 2003 50 44 48 10.1002/prot.10251 Chou CK Cai YD A new hybrid approach to predict subcellular localization of proteins by incorporating gene ontology Biochem Biophys Res Comm 2003 311 743 747 14623335 10.1016/j.bbrc.2003.10.062 Cai YD Chou CK Predicting 22 protein localizations in budding yeast Biochem Biophys Res Comm 2004 323 425 428 15369769 10.1016/j.bbrc.2004.08.113 Szafron D Lu P Greiner R Wishart DS Poulin B Eisner R Lu Z Anvik J Macdonell C Fyshe A Proteome Analyst: custom predictions with explanations in a web-based tool for high-throughput proteome annotations Nucleic Acids Res 2004 32 W365 371 15215412 Yu CS Lin CJ Hwang JK Predicting subcellular localization of proteins for Gram-negative bacteria by support vector machines based on n-peptide compositions Protein Sci 2004 13 1402 1406 15096640 10.1110/ps.03479604 Gardy JL Laird MR Chen F Rey S Walsh CJ Ester M Brinkman FS PSORTb v.2.0: expanded prediction of bacterial protein subcellular localization and insights gained from comparative proteome analysis Bioinformatics 2005 21 617 623 15501914 10.1093/bioinformatics/bti057 Chou KC Cai YD Predicting protein localization in budding yeast Bioinformatics 2005 21 944 950 15513989 10.1093/bioinformatics/bti104 Xiao X Shao S Ding Y Huang Z Huang Y Chou KC Using complexity measure factor to predict protein subcellular location Amino Acids 2005 28 57 61 15611847 10.1007/s00726-004-0148-7 Gao Y Shao S Xiao X Ding Y Huang Y Huang Z Chou CK Using pseudo amino acid composition to predict protein subcellular location: approached with Lyapunov index, Bessel function, and Chebyshev filter Amino Acids 2005 28 373 376 15889221 10.1007/s00726-005-0206-9 Nair R Rost B Mimicking Cellular Sorting Improves Prediction of Subcellular Localization Journal of Molecular Biology 2005 348 85 100 15808855 10.1016/j.jmb.2005.02.025 Lei Z Dai Y A new kernel based on high-scored pairs of tri-peptides and its application in prediction of protein subcellular localization Proceedings of International Workshop on Bioinformatics Research and Applications 2005 3515 Lecture Notes in Computer Science (LNCS), Springer-Verlag, Berlin 903 910 Lei Z Dai Y A class of new kernels based on high-scored pairs of k-peptides and its application in prediction of protein subcellular localization LNCS Transactions on Computational Systems Biology 2005 Carmo-Fonseca M The contribution of nuclear compartmentalization to gene regulation Cell 2002 108 513 521 11909522 10.1016/S0092-8674(02)00650-5 Hancock R Internal organisation of the nucleus: assembly of compartments by macromolecular crowding and the nuclear matrix model Biology of the Cell 2004 96 595 601 15519694 10.1016/j.biolcel.2004.05.003 Dellaire G Farrall R Bickmore WA The Nuclear Protein Database (NPD): subnuclear localisation and functional annotation of the nuclear proteome Nucl Acids Res 2003 31 328 330 12520015 10.1093/nar/gkg018 Bickmore WA Sutherland HGE NEW EMBO MEMBER'S REVIEW: Addressing protein localization within the nucleus EMBO J 2002 21 1248 1254 11889031 10.1093/emboj/21.6.1248 Vapnik VN Statistical learning theory 1998 Wiley, New York Cristianini N Shawe-Taylor J An Introduction to Support Vector Machines 2000 Cambridge University Press Ding CHQ Dubchak I Multi-class protein fold recognition using support vector machines and neural networks Bioinformatics 2001 17 349 358 11301304 10.1093/bioinformatics/17.4.349 Yu CS Wang JY Yang JM Lyu PC Lin CJ Hwang JK Fine-grained protein fold assignment by support vector machines using generalized n-peptide coding schemes and jury voting from multiple-parameters sets PROTEINS: Structure, Function, and Genetics 2003 50 531 536 10.1002/prot.10313 Leslie CS Eskin E Cohen A Weston J Noble WS Mismatch string kernels for discriminative protein classification Bioinformatics 2004 20 467 476 14990442 10.1093/bioinformatics/btg431 Brendel V PROSET – a fast procedure to create non-redundant sets of protein sequences Mathl Comput Modelling 1992 16 37 43 10.1016/0895-7177(92)90150-J SVMLight Rost B Sander C Prediction of protein secondary structure at better than 70% accuracy Journal of Molecular Biology 1993 232 584 599 8345525 10.1006/jmbi.1993.1413 Matthews BW Comparison of the predicted and observed secondary structure of T4 phage lysozyme Biochim Biophys Acta 1975 405 442 451 1180967 Platt JC Cristianini N Shawe-Taylor J Large margin DAGs for multiclass classification Advances in Neural Information Processing Systems 2000 12 MIT Press 547 553 Chou KC Zhang CT Review: Prediction of protein structural classes Critical Reviews in Biochemistry and Molecular Biology 1995 30 275 349 7587280 Morik K Brockhausen P Joachims T Combining statistical learning with a knowledge-based approach – A case study in intensive care monitoring Proceedings of the Sixteenth International Conference on Machine Learning 1999 268 277	16336650	PMC1325059	CC BY	2021-01-04 16:27:47	no	BMC Bioinformatics. 2005 Dec 7; 6:291	utf-8	BMC Bioinformatics	2,005	10.1186/1471-2105-6-291	oa_comm
==== Front BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2831631863510.1186/1471-2105-6-283DatabaseDSD – An integrated, web-accessible database of Dehydrogenase Enzyme Stereospecificities Toseland Christopher P [email protected] Helen M [email protected] Darren R [email protected] The Jenner Institute, University of Oxford, Compton, Berkshire, RG20 7NN, UK2005 30 11 2005 6 283 283 27 6 2005 30 11 2005 Copyright © 2005 Toseland et al; licensee BioMed Central Ltd.2005Toseland 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 Dehydrogenase enzymes belong to the oxidoreductase class and utilise the coenzymes NAD and NADP. Stereo-selectivity is focused on the C4 hydrogen atoms of the nicotinamide ring of NAD(P). Depending upon which hydrogen is transferred at the C4 location, the enzyme is designated as A or B stereospecific. Description The Dehydrogenase Stereospecificity Database v1.0 (DSD) provides a compilation of enzyme stereochemical data, as sourced from the primary literature, in the form of a web-accessible database. There are two search engines, a menu driven search and a BLAST search. The entries are also linked to several external databases, including the NCBI and the Protein Data Bank, providing wide background information. The database is freely available online at: Conclusion DSD is a unique compilation available on-line for the first time which provides a key resource for the comparative analysis of reductase hydrogen transfer stereospecificity. As databases increasingly form the backbone of science, largely complete databases such as DSD, are a vital addition. ==== Body Background Dehydrogenase enzymes exhibit stereo-selectivity for diastereotopic hydrogen atoms, situated at the C4 position of the dihydronicotinamide ring of the pyridine nucleotide coenzymes, Nicotinamide Adenine Dinucleotide (NAD) and Nicotinamide Adenine Dinucleotide Phosphate (NADP). The dehydrogenases belong to the oxidoreductase group of enzymes (E.C. Class 1), which oxidise their substrates by transferring hydrogen from NAD or NADP. These molecules act as high energy electron donors, plus low energy acceptors and are derived from the vitamin niacin; the molecules consist of two ribose groups, linked by two phosphates, which are in turn bonded to two bases, adenine and nicotinamide (Figure 1a). NADP is analogous to NAD, but NADP contains an additional phosphate group which forms an adenosine monophosphate group. The whole of the coenzyme is involved in interactions with the enzyme, but hydrogen transfer occurs only between substrate and the nicotinamide ring; other parts of the coenzyme are solely involved in binding to the enzyme and not in hydrogen transfer. NAD(P) are ubiquitous in all living systems and are implicated in a vast number of reactions involving almost 20% of classified enzymes [1]. Figure 1 The structures of pyridine nucleotide coenzymes, NAD and NADP. The nicotinamide rings which are pivotal to the hydride transfer, are highlighted. The rings are bound to ribose, which with respect to NAD, is linked to another ribose sugar and adenine through two phosphate groups. With regard to NADP, the ribose and nicotinamide ring are bound by two phosphate groups to adenosine monophosphate. The structure of NAD(P) allows presentation of either hydrogen depending on the orientation of the nicotinamide ring, which can be bound in either syn- or anti- conformations, which are related by a 180 degree rotation about the glycosidic bond linking the nicotinamide base to the C1 of ribose [2] (Figure 1b). There have been several designations for the C4 hydrogens since the 1950's, however the approved systems are A/B [3] and pro R/pro S [1]. The pro R/pro S system is superior to A/B as it denotes the absolute configuration of the atom. The A/B system works by comparing enzyme stereo-selectivity with that of yeast alcohol dehydrogenase, which is an A stereospecific enzyme. Enzymes with the same stereospecificity are designated as A, while other enzymes are denoted as B stereospecific [3]. For comparison between the two systems, A is essentially the same as pro R and B is the same as pro S [4]. Fisher [5] was the first to observe these stereospecific reactions. It was shown that yeast alcohol dehydrogenase catalysed the hydride transfer between the substrate and NAD. Further studies were able to show that such stereospecificity was not unique to alcohol dehydrogenases; the same hydride transfer was seen in malate [6] and lactate [7] dehydrogenases. However, it was not until 1955 that B stereospecific enzymes were discovered: β-hydroxysteroid dehydrogenase [8] and transhydrogenase [9]. These initial investigations were conducted using mass spectroscopy, with deuterium-labelled coenzymes [10]. Tritium-labelled coenzymes were later used with scintillation counting [11,12]. The C4 location of the nicotinamide ring is labelled and the isotope level determined [13]. Both of these techniques were not adequate for large-scale studies, as they are laborious, time-consuming and require multiple purification steps [12]. In 1976, Arnold [14] introduced a technique based on 1H Nuclear Magnetic Resonance (NMR) spectroscopy. This method was accurate and allowed direct measurement of deuterium content at the reaction site, as well as being safer and more facile than previous techniques. The determination of stereospecificity is essential for gaining an understanding of the reaction mechanism of an enzyme. Most important of all, is the fact that knowledge of the stereochemistry based on hydride transfer can provide an insight into the relative substrate and coenzyme binding orientations; this can be instrumental in designing strategies to control and regulate reactions [13]. We have therefore accumulated a comprehensive compendium of hydride transfer stereospecificity data, and placed the information in a searchable database. The Dehydrogenase Stereospecificity Database (DSD) v1.0 database designates enzymes as either A or B stereospecific, for a given coenzyme. Two decades ago, You [1] compiled available stereospecificity data in a landmark review; we have used this and added experimental data determined during the past 20 years, to bring the compilation up to date. This provides a unique and accurate collection of dehydrogenase enzyme stereospecificities. Construction and content Database design and development DSD v1.0 has been implemented as a postgreSQL relational database. This establishes a flexible infrastructure capable of addressing all foreseeable developments of the archive. The data was initially compiled in a Microsoft ACCESS database after addition of data from previous studies and exhaustive searching of the primary literature. The postgreSQL database is structured into seven normalised tables, populated from a flat-file export of the ACCESS database using PERL scripts integrated with SQL. See Figure 2. As data are continually accumulating, archiving data is an on-going process: automatic, periodic updates will be made to the postgreSQL database. The DSD user interface is a series of HTML pages. Searchable HTML forms are available within the DSD site, and provide either a detailed DSD search or a DSD BLAST search. These forms target either a PERL/SQL script or a CGI script. These will in turn query the database. The bespoke search engine facilitates fast, efficient and flexible data retrieval (see 'Searching the Database'). DSD is freely available on the web . Figure 2 The anti- and syn- conformations of the nicotinamide ring. A stereospecific enzymes will bind NAD(P) in the anti- conformation, which B stereospecific enzymes bind NAD(P) in the syn- conformation. Implementing the database The data was sourced from a previous compilation [1] and exhaustive searching of the primary literature. The archive consists of entries containing the enzyme name, the species from which it was derived, experimental technique, coenzymes and the stereospecificity. For simplicity, DSD utilises the A/B naming system [3], which was adopted by You [1]. Cross-references to key external databases are also made. These provide links to the protein sequence, using NCBI Entrez-Protein, and the relevant protein structure in the Protein Data Bank [15]. If applicable, the Enzyme Classification is given with links to the Enzyme Nomenclature and Classification database, developed in line with the International Union of Biochemistry and Molecular Biology, providing details of the enzyme reactions. In addition, a full literature reference is given with links to the NCBI PubMed journals database. These links provide the user with necessary background knowledge for the archived enzyme. A full description of the entry fields is given in Table 1. Table 1 Contents of the database entries. Entry Field Description Enzyme The relevant protein and provides a link to the NCBI Entrez-Protein sequence PDB PDB identification code, plus a link to the equivalent structure EC The Enzymes Commissions identification number, plus a link to the external database Source Species Species in which the protein is found Coenzyme The coenzyme used in the experimental determination e.g. NAD(P) Stereospecificity Given according to the A/B nomenclature Method Experiment techniques used to obtain data e.g. NMR Temperature Temperature at which the experiment was carried out pH Range or fixed pH at which the experiment were carried out Conditions Concentrations, etc used in the experiment Reference Full literature reference with link to the PubMed database The usefulness of a database is governed by the accuracy of the data it contains. The data contained in DSD v1.0 was compiled manually from previously published, peer-reviewed articles, and verified, where possible, from the original literature. This suggests that, compared to some other databases, DSD will be accurate and reliable. Moreover, to maintain this accuracy, data was only added from the primary literature for widely accepted experimental techniques, such as NMR spectroscopy. Experimental conditions have also been included, with 113 temperature measurements, and 117 pH values, and 135 concentration measurements. As logistical considerations preclude us from undertaking independent verification of the data, we are obliged to trust the data reported in the literature. Searching the database Two search methods for DSD are available: a menu-based interface and an interface based on Basic Local Alignment Search Tool (BLAST) [16]. The implementation of a menu-based bespoke search system allows the user to perform either a broad or a detailed search from one simple search interface. The database is searched by selecting enzyme, source or experimental method, or a combination thereof. The HTML form passes the search criteria to SQL database queries. The data returned is formatted using HTML before being printed to the screen. The initial results page summarises the data fields 'PDB', 'Enzyme', 'Source', 'Method' and 'EC Number' for all hits. A 'View Data' button is available to obtain experimental detail for the selected data set. See Figure 3. Figure 3 DSD Table Schema. Figure 4 Overview of the DSD search. The search menus (A) provide an easy method of selecting for one, or all of, the three categories. The search will display a list of matches for the criteria (B), the DSD entry can then be viewed for the selected enzyme (C). The alternative search interface is based on BLAST [16]. A local database of protein sequences found in DSD was compiled from SWISS-PROT [17] and an additional postgreSQL table was created to hold this data. The local database is searched using the NCBI BLASTP and BLASTX programs [16], allowing for input of protein or nucleotide sequences. The HTML Front-End connects to a web server-based PL/CGI scripts which interacts with the BLASTP or BLASTX programs. In the output, DSD entries are linked to SWISS-PROT [17] using accession codes. Utility and discussion Implementation and initial analysis DSD is a relatively small database but is, nonetheless, essentially complete. The current size of the database is 461 entries, which corresponds to 185 distinct enzymes from 121 different source species. From the current data, we can conclude that NAD is the more frequently used coenzyme, since there are 272 entries for enzymes using NAD and 189 using NADP, which represents a 60:40 divide between the two main pyridine nucleotide coenzymes NAD and NADP, respectively. This may simply reflect experimental bias and not any natural selection for a particular coenzyme. It has been reported several times before that there is an approximately equal distribution of A and B stereospecific dehydrogenase enzymes [1,13] and that none are stereo-random; the up-to-date data contained within DSD v1.0 is consistent with this view. Our data displays an approximately even distribution, with 236 occurrences of A stereospecific enzymes and 225 occurrences of B stereospecific enzymes (Table 2). Table 2 Overview of the DSD data. Coenzyme Frequency NAD 271 NADP 181 Stereospecificity Occurrence A 242 B 232 The distribution of coenzymes can be seen above: a 60:40 division is present between the NAD and NADP, respectively. The overall distribution between the stereospecificities of dehydrogenase enzymes is as observed previously. Our data is analogous to prior studies (You, 1985; Hallis et al., 1998) which indicated an approximately equal distribution. A-stereospecific enzymes may have evolved before B-stereospecific enzymes, since A-stereospecific enzyme substrates are smaller and less complex compounds than those of B-stereospecific enzymes [2]. Prebiotic chemical processes would favour simpler molecules, while primitive organisms would have had rudimentary catabolic systems able to derive energy from small molecules. Due to structural conservation, it is also possible to deduce evolutionary correlations between diverse catalysts by comparing the binding orientations of the coenzymes, substrate and enzyme active site [18]. On this basis, Benner [19,20] postulated that A stereospecific enzymes will bind NAD(P) with the ring in the anti-conformation, while B stereospecific enzymes will bind NAD(P) with the ring in the syn-conformation (Figure 1b). With respect to dehydrogenase enzymes, which have divergently evolved from a common precursor, the stereospecificity will be maintained for as long as the protein fold of the catalytic domain is conserved, for a given substrate [21]. An A to B transformation in stereospecificity, or vice versa, would require a total rearrangement of the essential functional residues, to allow the correct binding orientation. This is based on the structural constraints mentioned above and therefore such an exchange is inconceivable during divergent evolution [18]. Binding constraints may not be an overall determinant, as several groups of enzymes from different sources having contrasting stereospecificities [21]. A relationship between the stereospecificity and equilibrium constant of a hydride transfer has also been proposed: reactions with pKeq > 11.3 correlate with A stereospecificity and reactions with pKeq <11.3 correlate with B stereospecificity. However, as with many postulated generalisations, there are exceptions, such as 3α-hydroxysteroid dehydrogenase. The enzyme reaction has a pKeq ~ 8, within the B range yet it is A stereospecific [18]. This highlights the problems associated with forming generalisations about these reactions. Future work With respect to future work, the database needs to be maintained and developed further, ensuring our links to external databases remain current and newly published data is added. Initially, as with all databases, random errors will have occurred due to human error during the data accumulation or will be extant within the original experimental data. The database will be assessed for errors and inconsistencies, thus maintaining, as far as possible, the overall veracity of our data. Moreover, feedback on the search interfaces and the general infrastructure will allow us to address issues raised and revise the database accordingly. You [1] included data on several other pyridine nucleotide coenzymes, such as 3-acetylpyridine adenine Dinucleotide ((3AcPy)AD), 3-cyanopyridine adenine Dinucleotide ((CNPy)AD), thionicotinamide adenine dinucleotide ((TN)AD) and flavin adenine dinucleotide (FAD). We will seek to incorporate data on these and similar coenzymes in future versions of the database. Likewise, we will seek to enhance search capabilities within the database by including the ability to search for types of interaction sets of amino acids within user-defined distances of NAD(P) coenzymes. We will complement this with a simple visual summary. Conclusion The DSD v1.0 database is a unique, up to date compilation of dehydrogenase stereospecificities, which has advanced on reviews put together over 20 years ago. We see the database as being relatively small, due to the constraints of information, but largely complete. As new data becomes available, the database will increase in size. The ease of access to the data is of great importance and the bespoke search system and the inclusion of a BLAST search greatly facilitates this. The addition of the cross-references to several external databases provides expansive background information. We hope the database will provide an important resource which will help enhance our understanding of enzyme mechanisms. In an age when databases are increasingly forming the backbone of science, largely complete databases, such as DSD, are an important addition. Availability and requirements The database is available at suitable for most graphical web browser. Authors' contributions CPT compiled the database and developed the BLAST search tool. HM was responsible for developing all postGRESQL and perl code and also designed the HTML web front end. DRF originated the concept, designed the database, and identified relevant data sources. All authors have read and approved the final manuscript. Acknowledgements We should like to thank Andrew Worth for his technical assistance and Martin Blythe for help with programming. The Edward Jenner Institute for Vaccine Research wishes to thank its sponsors: GlaxoSmithKline, the Medical Research Council, the Biotechnology and Biological Sciences Research Council, and the UK Department of Health. ==== Refs You KS Stereospecificity for nicotinamide nucleotides in enzymatic and chemical hydride transfer reactions CRC Crit Review Biochem 1985 17 313 451 You K Arnold LJ Allison WS Kaplan NO Enzyme stereospecificities for nicotinamide nucleotides Trends Biochem Sci 1978 3 265 10.1016/S0968-0004(78)95849-8 Levy HR Talalay P Vennesland B The steric course of enzymatic reactions at the meso carbon atoms: Application of hydrogen isotopes Progress in Stereochemistry 1962 3 299 Cornforth JW Cornforth RH Donniger C Popják G Ryback G Schroepfer GJ B 163 Proc R Soc London 1965 B 163 436 Fisher HF Ofner P Conn EE Vennesland B Westheimer FH Direct enzymatic transfer of hydrogen atoms between substrates and DPN Fed Proc 1953 11 211 Loewus FA Tchen TT Vennesland B The enzymatic transfer of hydrogen. III. The reaction catalyzed by malic dehydrogenase J Biol Chem 1955 212 787 800 14353881 Loewus FA Ofner P Fisher HF Westheimer FH Vennesland B The enzymatic transfer of hydrogen. II. The reaction catalyzed by lactic dehydrogenase J Biol Chem 1953 202 699 704 13061493 Talalay P Loewus FA Vennesland B The enzymatic transfer of hydrogen. IV. The reaction catalyzed by a beta-hydroxysteroid dehydrogenase J Biol Chem 1955 212 801 809 14353882 San Pietro A Kaplan NO Colowick SP Pyridine nucleotide transhydrogenase. VI. Mechanism and stereospecificity of the reaction in Pseudomonas fluorescens J Biol Chem 1955 212 941 952 14353895 Levy HR Vennesland B The stereospecificity of enzymatic hydrogen transfer from diphosphopyridine nucleotide J Biol Chem 1957 228 85 96 13475298 Jarabak J Talalay P Stereospecificity of hydrogen transfer by pyridine nucleotide-linked hydroxysteroid dehydrogenases J Biol Chem 1960 235 2147 2151 14406805 Trincone A Lama L Rella R D'Auria S Raia CA Nicolaus B Determination of hydride transfer stereospecificity of NADH-dependent alcohol-aldehyde/ketone oxidoreductase from Sulfolobus solfataricus Biochim Biophys Acta 1990 1041 94 96 2121281 Hallis TM Liu HW Stereospecificity of hydride transfer for the catalytically recycled NAD+ in CDP-D-glucose 4,6-Dehydrogenase Tetrahedron 1998 54 15975 15982 10.1016/S0040-4020(98)01005-9 Arnold LJ JrYou K Allison WS Kaplan NO Determination of the hydride transfer stereospecificity of nicotinamide adenine dinucleotide linked oxidoreductases by proton magnetic resonance Biochemistry 1976 15 4844 4849 186097 10.1021/bi00667a014 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 10592235 10.1093/nar/28.1.235 Altschul SF Madden TL Schaffer AA Zhang J Zhang Z Miller W Lipman DJ Gapped BLAST and PSI-BLAST: a new generation of protein database search programs Nucleic Acids Res 1997 25 3389 3402 9254694 10.1093/nar/25.17.3389 Boeckmann B Bairoch A Apweiler R Blatter MC Estreicher A Gasteiger E Martin MJ Michoud K O'Donovan C Phan I Pilbout S Schneider M The SWISS-PROT protein knowledgebase and its supplement TrEMBL in 2003 Nucleic Acids Res 2003 31 365 370 12520024 10.1093/nar/gkg095 Garavito RM Rossmann MG Argos P Eventoff W Convergence of active center geometries Biochemistry 1977 16 5065 5071 143959 10.1021/bi00642a019 Benner SA The stereoselectivity of alcohol dehydrogenases: a stereochemical imperative? Experientia 1982 38 633 636 7047206 Nambiar KP Stauffer DM Kolodziej PA Benner SA A mechanistic basis for the stereoselectivity of enzymatic transfer of hydrogen from nicotinamide cofactors J Am Chem Soc 1983 105 5886 10.1021/ja00356a028 Davies DD Teixeira A Kenworthy P The stereospecificity of nicotinamide-adenine dinucleotide-dependent oxidoreductases from plants Biochem J 1972 127 335 343 4403953	16318635	PMC1325060	CC BY	2021-01-04 16:27:46	no	BMC Bioinformatics. 2005 Nov 30; 6:283	utf-8	BMC Bioinformatics	2,005	10.1186/1471-2105-6-283	oa_comm
==== Front PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1637949610.1371/journal.pbio.0040011Research ArticleMolecular Biology/Structural BiologyArchaeaStepwise Translocation of Dpo4 Polymerase during Error-Free Bypass of an oxoG Lesion Translocation during oxoG Lesion BypassRechkoblit Olga 1 Malinina Lucy 1 Cheng Yuan 1 Kuryavyi Vitaly 1 Broyde Suse 2 Geacintov Nicholas E 3 Patel Dinshaw J [email protected] 1 1Structural Biology Program, Memorial Sloan-Kettering Cancer Center, New York, New York, United States of America2Biology Department, New York University, New York, New York, United States of America3Chemistry Department, New York University, New York, New York, United States of AmericaHerschlag Daniel Academic EditorStanford UniversityUnited States of America1 2006 3 1 2006 3 1 2006 4 1 e1112 8 2005 1 11 2005 Copyright: © 2006 Rechkoblit et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Righting the Wrongs: Structural Insights into Replicating Damaged DNA 7,8-dihydro-8-oxoguanine (oxoG), the predominant lesion formed following oxidative damage of DNA by reactive oxygen species, is processed differently by replicative and bypass polymerases. Our kinetic primer extension studies demonstrate that the bypass polymerase Dpo4 preferentially inserts C opposite oxoG, and also preferentially extends from the oxoG•C base pair, thus achieving error-free bypass of this lesion. We have determined the crystal structures of preinsertion binary, insertion ternary, and postinsertion binary complexes of oxoG-modified template-primer DNA and Dpo4. These structures provide insights into the translocation mechanics of the bypass polymerase during a complete cycle of nucleotide incorporation. Specifically, during noncovalent dCTP insertion opposite oxoG (or G), the little-finger domain–DNA phosphate contacts translocate by one nucleotide step, while the thumb domain–DNA phosphate contacts remain fixed. By contrast, during the nucleotidyl transfer reaction that covalently incorporates C opposite oxoG, the thumb-domain–phosphate contacts are translocated by one nucleotide step, while the little-finger contacts with phosphate groups remain fixed. These stepwise conformational transitions accompanying nucleoside triphosphate binding and covalent nucleobase incorporation during a full replication cycle of Dpo4-catalyzed bypass of the oxoG lesion are distinct from the translocation events in replicative polymerases. Crystal structures of the bypass polymerase Dpo4 at different stages of lesion-bypass reveal how the cell copes with oxidative DNA damage. ==== Body Introduction Y-family polymerases are able to bypass a variety of DNA lesions that impede high-fidelity replicative DNA polymerases. Such bypass polymerases exhibit a higher error rate and lower processivity on undamaged DNA templates, and can extend from mismatched base pairs (reviewed in [1,2]). Studies suggest that translesion Y-family DNA polymerases are temporarily recruited to overcome blocks to replicative polymerases [3,4]. Y-family polymerases have more spacious and solvent-accessible active sites as observed for archaeal DNA polymerase IV (Dpo4) [5] and Dbh [6], yeast pol η [7], human pol ι [8], and pol κ [9], crystallized in the apo form (pol η, Dbh, pol κ), and as ternary complexes with an incoming deoxyribonucleotide triphosphate (dNTP) (Dpo4, pol ι). The solvent-accessible nature of the active site and the smaller number of contacts of the template-primer DNA with the polymerase enable Dpo4 to accommodate unusual DNA structures in its active site. These include frameshift-template misaligned sequences [5], the cis–syn thymine photodimer [10], a bulky benzo[a]pyrene-diol–epoxide-adenine lesion [11], an abasic site [12], a reverse wobble G•T mismatch [13], and an ethenoguanine lesion [14]. Structural studies have elucidated the effect of metal ions, nucleotide selection, and pyrophosphorolysis on Dpo4 fidelity [15]. By contrast, replicative polymerases produce tight-fitting, solvent-excluding active sites upon binding of the correct nucleotide; the O helix in the finger domain undergoes a large movement to position itself on the flat surface of a complementary, Watson-Crick nascent base pair [3,16]. This may represent a kinetic rate–limiting step that occurs prior to covalent nucleotide incorporation which has been interpreted in terms of an induced-fit mechanism [17]. If an unusual DNA alignment or a damaged base is encountered at the active site, the O helix often remains in the “open,” inactive conformation [18–21]. In striking contrast to the replicative polymerases, the finger domains of Y-family Dpo4, Dbh, pol η, pol ι, and pol κ polymerases are missing the equivalent of the O helix [8,9,22]. Instead, the replicating base pair is contacted by a β-sheet (Dpo4 residues 41–46) of the finger domain, which forms the rigid roof of the active site, and by an adjacent extended loop (Dpo4 residues 53–59). Kinetic studies indicate that the Y-family pol η and Dpo4 employ a rate-limiting protein conformational change before covalent nucleotide incorporation occurs at the active site [23,24]. However, examinations of currently available crystal structures of Y-family polymerases do not reveal any obvious conformational changes, and it has been suggested that bypass polymerases, such as Dpo4, are always in the “closed,” ready-for-catalysis conformation [22,25]. The absence of an open-to-closed conformational transition was also observed in the case of the repair gap–filling pol λ [26]; the dNTP is accommodated in the space that was, in the binary complex, occupied by the side chain of a tyrosine residue, and the template strand is repositioned toward the active site. 7,8-dihydro-8-oxoguanine (oxoG) is the major known product of oxidation of DNA by reactive oxygen species induced by either ionizing radiation, photochemical mechanisms, or normal cellular metabolic activity [27]. An increased risk for developing cancer has been linked to oxidative stress due to the overproduction of reactive oxygen species resulting from the response of cells to inflammation and infection [28]. Replicative polymerases in vitro readily insert C or A opposite the oxoG lesion in varying proportions that depend on the polymerase, with extension occurring preferentially from oxoG•A mispairs [29–31]. In contrast, the Y-family polymerases yeast and human pol η [32], and Dpo4 (this study), preferentially insert C opposite oxoG, and also preferentially extend from the oxoG•C base pair, thus achieving error-free bypass of this lesion. To date, a number of crystal structures have been reported for replicative polymerases with oxoG-modified template-primers and dNTPs positioned opposite the lesion or the adjacent 5′-template base, corresponding to insertion and extension ternary complexes, respectively. These include insertion ternary complexes, with oxoG positioned opposite 2′-deoxycytidine 5′-triphosphate (dCTP) and 2′-deoxyadenosine 5′-triphosphate (dATP), with the repair gap–filling pol β [33], and opposite dCTP for the replicative polymerases Rb69 [18] and pol T7 [34], as well as extension complexes past oxoG•C and oxoG•A base pairs by pol T7 [34] and Bacillus pol I [35]. The objectives of this study are to address a major gap in our understanding of the conformational changes associated with the polymerase translocation steps that occur in Y-family polymerase–DNA substrate complexes that accompany dNTP insertion and the nucleotidyl transfer reaction. We have chosen Dpo4 due to its high homology to human pol η and similar lesion bypass properties, thereby making it a useful model system for studying Y-family polymerases [36]. We report on crystal structures of three complexes of bypass polymerase Dpo4 and oxoG-modified template-primer DNA, thus capturing two crucial steps of the polymerization cycle. Our results highlight novel and unanticipated conformational changes associated with the Dpo4-DNA interface upon dCTP binding opposite oxoG and upon covalent incorporation of this nucleobase into the primer strand, thereby providing snapshots of the intermediate translocation mechanics for a bypass polymerase. Results The objectives of this work were to investigate the conformational changes that occur when a dNTP is inserted opposite an oxoG template base in the active site of Dpo4, and the subsequent incorporation of this nucleobase into the nascent DNA strand. Since translesion bypass of oxidatively damaged DNA templates by Dpo4 has not been previously investigated to our knowledge, we first surveyed the kinetics of translesion bypass of oxoG in order to select the most appropriate dNTP to be inserted opposite this lesion in our subsequent co-crystal structural studies. Fidelity of Translesion Synthesis and Primer Extension Kinetics The results of steady-state, one-step primer extension experiments are summarized in Table 1. Primer extension studies in vitro utilizing all four dNTPs show that Dpo4 readily elongates past guanine (G) or oxoG lesions in the template strand to produce full-length, 19-mer extension products (Figure 1A). Michaelis-Menten parameters Vmax and Km (Table 1), were determined for the dNTP insertion step opposite oxoG (or G at the same site) (Figure 1B), and for the one-step extension step beyond an oxoG•C or an oxoG•A base pair (Figure 1C). The insertion frequency, fins, of dCTP opposite oxoG is only 2-fold smaller than it is opposite G, with the fins values opposite oxoG decreasing in the order dCTP > dATP, dTTP > dGTP (Table 1). Interestingly, the fext value is significantly greater for the one-step extension beyond an oxoG•C base pair than beyond a normal G•C base pair at the same site. Also, extension beyond an oxoG•A mispair is ~30 times smaller than extension from an oxoG•C base pair (Table 1). Taken together, these kinetic primer extension data suggest that the oxoG lesion is readily bypassed by Dpo4 in a predominantly error-free manner. In view of our primer extension results catalyzed by Dpo4, we have focused our structural studies on complexes of oxoG-modified template-primer DNA and Dpo4, in the absence and presence of incoming dCTP as well as covalently incorporated C opposite the oxoG lesion site. Figure 1 Fidelity and Efficiency of Base Incorporation on Unmodified G and oxoG-Modified Templates by Dpo4 (A) Time course of 32P 5′-end-labeled 13-mer primer extension on 19-mer G (left panel) or oxoG (right panel) templates in the presence of all four dNTPs by Dpo4. The labeled (−1) position or 13-mer represents the 13-mer primer with the 3′ end positioned one base before the G or oxoG; the 14-mer extended to position (0) opposite G or oxoG; the 19-mer extended to position (+5) corresponds to full extension. (B) Kinetics of single nucleotide insertion on G (left panel) and oxoG (right panel) templates by dCTP, dATP, dTTP and dGTP. The gels (top panels) show data as a function of dNTP concentration, while the plots (bottom panels) measure the log of the frequency of nucleotide insertion for each dNTP. (C) Kinetics of single nucleotide extension from C and A opposite G (left panel) and oxoG (right panel) templates by the next correct nucleotide (dGTP). The gels (top panels) show data as a function of dGTP concentration, while the plots (bottom panels) measure the log of the frequency of nucleotide extension from G/oxoG•C and G/oxoG•A pairs. The primer extension data recorded under single-hit polymerization conditions (less than 20% of primer extended) were used for estimation of Michaelis-Menten parameters listed in Table 1. Table 1 Kinetic Parameters of Insertion and Extension Catalyzed by Dpo4 DNA Substrate Design and Crystal Structure Determination In order to capture Dpo4 conformational changes during DNA synthesis, we crystallized this bypass polymerase in three types of complexes: the preinsertion binary complex containing oxoG (or G) 19-mer template and 13-mer primer strand (Figure 2A), the insertion ternary complex with a dCTP at the active site paired with oxoG (or G) in the template strand (Figure 2B), and the postinsertion binary complex with the incorporated C residue at the 3′ end of the primer paired with oxoG in the template strand (Figure 2C). All three types of complexes (Figure 2) belong to the P21 space group with distinct unit cell parameters and with two complexes per asymmetric unit (Table 2). Figure 2 Structures and Intermolecular Contact Details of Dpo4 Complexes with oxoG-Modified Template-Primer DNA (A) Schematic of the pairing of the oxoG-containing 19-mer template strand with the 13-mer primer strand ending in a 2′,3′-dideoxynucleotide in the preinsertion binary complex with Dpo4. (B) Schematic of the pairing of the oxoG-containing 19-mer template strand with the 13-mer primer strand ending in a 2′,3′-dideoxynucleotide plus incoming dCTP in the insertion ternary complex with Dpo4. dCTP is positioned opposite oxoG. (C) Schematic of the pairing of the oxoG-containing 19-mer template strand with the 14-mer primer strand ending in a 2′,3′-dideoxynucleotide in the postinsertion binary complex with Dpo4. Covalently incorporated C is positioned opposite oxoG. (D) Structure of the active site of the preinsertion binary complex. The DNA and protein are in stick and ribbon representations, respectively. Template residues 1–6 (including oxoG) of single-stranded overhang are disordered in the electron density maps. (E) Structure of the active site of the insertion ternary complex, with incoming dCTP paired with oxoG(anti) at the active site. (F) Structure of the active site of the postinsertion binary complex, with covalently incorporated C paired with oxoG(anti). (G) Details of the intermolecular contacts in the structure of the preinsertion binary complex. (H) Details of the intermolecular contacts in the structure of the insertion ternary complex. (I) Details of the intermolecular contacts in the structure of the postinsertion binary complex. Color-coding for panels G–I. DNA-backbone phosphate groups interacting with the palm domain are shown in light green, the phosphate groups interacting with the thumb domain within the minor groove are shown in red, and the phosphate groups interacting with the little-finger domain within the major groove are shown in purple. The conserved interactions are boxed for the thumb and little-finger contacts. Dpo4 amino acids interacting via the peptide backbone are labeled black, via their side chains, green, and via the backbone and side chains, orange. Black arrows indicate hydrogen bonding, blue arrows water-bridged contacts, and pink arrows depict van-der-Waals contacts. Table 2 Crystal Data, Data Collection, and Refinement Statistics The structure of the Dpo4 oxoG-modified insertion ternary complex was solved by molecular replacement employing the published Dpo4 ternary complex with an 18-mer unmodified template–12-mer primer junction and incoming ddADP [5] as a search model, and refined to 1.95-Å resolution. The molecular replacement method was employed to solve the structures of the preinsertion and postinsertion binary complexes using the oxoG-modified insertion ternary complex as a search model. The crystal data, together with the data collection and refinement statistics for all structures, are summarized in Table 2. Structure of the oxoG-Modified Preinsertion Binary Complex We have determined the crystal structure of the Dpo4 preinsertion binary complex with oxoG-modified 19-mer template/dideoxy-terminated, 13-mer primer at 2.35-Å resolution (Figure S1A). The corresponding structure of the preinsertion binary complex with unmodified G was solved at 2.7-Å resolution (Figure S2A). The 5′-C-T-A-A-C-oxoG (or G) single-stranded template overhang is disordered in both crystals. The structures of the oxoG-modified and unmodified preinsertion binary complexes are similar, with root mean square deviation (RMSD) = 0.68 Å; therefore we focus on the description of the oxoG-modified complex. The alignment of residues in the Dpo4 polymerase active site pocket is outlined in Figure 2D, with the C7•G13 base pair (the numbering scheme is defined in Figure 2A), adjacent to the oxoG lesion, enclosed in a box for reference. The roof of the active site, formed by the finger domain (in blue), is positioned directly over this C•G pair in the preinsertion binary complex (The Dpo4 bears only one Ca2+ cation, coordinated by invariant D7, D105, and E106 residues Figure 2D). The Dpo4 polymerase embraces the 19-mer template/13-mer primer DNA by its four domains: palm (residues 1–10 and 78–166), finger (residues 11–77), thumb (residues 167–233), and little-finger (residues 244–341) (see Figures S1A and S2A). The thumb is joined to the little-finger domain by a 10-amino-acid-long tether (residues 234–243) that allows positioning of the little finger on the other side of the DNA duplex. The contacts between Dpo4 amino acid residues and the DNA in the oxoG-modified preinsertion binary complex (see Figure S1A) are depicted in Figure 2G. The phosphate groups interacting with the thumb domain within the minor groove are shown in red, and the phosphate groups interacting with the little-finger domain within the major groove are shown in purple, with the conserved interactions boxed for both domains. Structure of oxoG-Modified Insertion Ternary Complex with Incoming dCTP We have determined the 1.95-Å crystal structure of the Dpo4 insertion ternary complex with oxoG-modified, 19-mer template/dideoxy-terminated 13-mer primer, and dCTP positioned opposite oxoG (see Figure S1B). The corresponding structure of the insertion ternary complex with unmodified G was solved at 2.80-Å resolution (Figure S2B). Residues 3–19 of the template strand in the oxoG-modified complex and the entire template strand in the unmodified complex were successfully traced in the electron density map. The structures of the oxoG-containing and unmodified insertion ternary complexes are similar with RMSD = 0.60 Å; therefore we focus on the description of the oxoG-modified complex. The finger domain, which is on top of the C7•G13 base pair in the active site of the preinsertion binary complex (Figure 2D), is now positioned over the replicating oxoG(anti)•dCTP base pair (Figure 3A) in the active site of the insertion ternary complex (see Figure 2E). The side chains of the hydrophobic residues V32 and A42 and the backbone of G58 are packed against the purine ring of oxoG, and the hydrophobic residues A44, A57, and Y12 are packed against the pyrimidine and sugar rings of incoming dCTP. The positioning of the triphosphate backbone of dCTP within the catalytic pocket is shown in Figure 3B. Details of the intermolecular contacts between Dpo4 and the oxoG-containing template-primer junction and the dCTP in the insertion ternary complex (see Figure S1B) are shown schematically in Figure 2H. Figure 3 Comparison of Pairing Alignments and Intermolecular Interactions Involving oxoG-Modified and Unmodified G Residues in Insertion Ternary Complexes with Incoming dCTP (A) oxoG(anti) paired with dCTP(anti) in the oxoG-modified insertion ternary complex, with a coordinated Ca2+ cation. 2Fo-Fc 1.95 Å electron density map contoured at 1σ level (1.95 Å resolution). (B) Shape-complementarity between incoming dNTP and the Dpo4 active site pocket containing two chelated Ca2+ cations (pink spheres). The dCTP (in silver) is aligned opposite oxoG lesion (in gold). (C) Distinguishing Ca2+ from Mg2+ ions in the Dpo4 oxoG-modified insertion ternary complex. Strong peaks, contoured in pink at 5 σ level, are found in the 2.0 Å anomalous map plotted for a dataset collected at 1.5418 Å wavelength (Mg2+ cation has a much weaker anomalous scattering at this wavelength than Ca2+ cation). 2Fo-Fc electron density map for Ca2+ cations contoured in blue at 3 σ level are shown by small pink spheres. (D) Interactions between oxoG nucleotide and amino acid side chains of Dpo4 in the oxoG-modified insertion ternary complex. (E) Interactions between G nucleotide and amino acid side chains of Dpo4 in the unmodified G insertion ternary complex. (F) Comparison of the phosphate backbone conformation of C5-oxoG6-C7 (gold) and unmodified C5-G6-C7 (gray) segments within the Dpo4 active site of their respective insertion ternary complexes. Four divalent cation sites were identified in the first molecule, and three sites were found in the second molecule within the asymmetric unit of the oxoG-modified insertion ternary complex. Since the crystallization conditions for the Dpo4-DNA complexes included 100 mM calcium and 5 mM magnesium ions, we used the anomalous signal of the calcium ion at 1.5418 Å (Cu Kα radiation) to distinguish it from the much weaker anomalous scattering by the magnesium ion at this wavelength. Seven strongest peaks on the anomalous map allowed us to assign all ions to calcium. We find one Ca2+ ion coordinated by the catalytic triad, the second chelated by the phosphate groups of the incoming dCTP, with the third ion coordinated by the loop of the thumb domain (residues 181 and 186), adjacent to the tip of helix H (Figure 3C). The fourth calcium ion (absent in the second molecule of the asymmetric unit) is bound between adjacent C•G and A•T base pairs in the free portion of the double-stranded DNA region. Comparison of oxoG versus G Alignments in Insertion Ternary Complexes Both oxoG(anti)•dCTP (Figure 3A) and G(anti)•dCTP pairing alignments are of the Watson-Crick type in their respective insertion ternary complexes. The sugar-phosphate backbone of the oxoG and G residues and their interactions with amino-acid side chains in their respective insertion ternary complexes are shown in Figure 3D and 3E, respectively. Accommodation of the oxoG carbonyl functionality within the Dpo4 polymerase active site results in a shift of the phosphate group of the oxoG base by 3.5 Å and a change in the backbone torsion angle α (O3′-P-O5′-C5′) by 176° as compared with the unmodified structure (superimposed segments compared in Figure 3F), with a similar observation reported previously for the pol β complex [33]. The oxygen atoms of the relocated phosphate group form two hydrogen bonds with the guanidine group of Arg331, one hydrogen bond with the guanidine group of Arg332, and two hydrogen bonds with the hydroxyl group of Ser34 (Figure 3D). The carbonyl oxygen at C8 forms a water-mediated hydrogen bond with the side chain of Arg332, thus helping to lock the oxoG residue in the anti conformation (Figure 3D). By contrast, the phosphate group of unmodified G forms only two hydrogen bonds with the guanidine group of Arg332 (Figure 3E); thus oxoG(anti) forms four more hydrogen bonds with Dpo4 (Figure 3D) than unmodified G does (Figure 3E). The dCTP-Binding Step: Conformational Transitions from Preinsertion Binary to Insertion Ternary Complex We now compare the structures of the preinsertion binary complex with the insertion ternary complex that contains a dCTP molecule opposite oxoG by superpositioning their DNA duplexes in order to identify conformational changes associated with translocation of the polymerase during the dCTP-binding step. This superposition for the entire complex is shown in Figure 4A and 4B, with little-finger domain–DNA interactions shown in Figure 4C and thumb domain–DNA interactions shown in Figure 4D. The key observation is that the finger, palm, and little-finger domains move up by one nucleotide to allow for the noncovalent insertion of dCTP into the active site (Figure 4C, with details shown in Figure 5), while the thumb maintains its contacts with the DNA duplex (shown by arrows in Figure 4D, with details shown in Figure 6). Figure 4 Conformational Transitions of Dpo4 and oxoG-Modified Template-Primer DNA Associated with the Nucleotide-Binding Step Following Superimposition of DNA Duplexes (A) Overall comparative views normal to the DNA helix axis (following superimposition of the DNA duplexes) of the structures of the preinsertion binary complex (in silver) and the insertion ternary complex with incoming dCTP (in color). The DNA-backbone phosphate groups in contact with the Dpo4 little-finger and thumb domains in the ternary complex are shown by colored spheres. (B) Overall comparative view looking down the DNA helix axis. (C) Comparative views of contacts between the little-finger domain and the DNA backbone, with contacts shown by CPK spheres. The β-sheets of the little-finger domain move by a one nucleotide step upon proceeding from preinsertion binary complex (silver) to insertion ternary complex (purple). (D) Comparative views of contacts between the thumb domain and the DNA backbone, with contacts shown by CPK spheres. The α-helical segments of the thumb domain (helixes H, K, J) retain the same contacts with the backbone phosphates of the primer and template strands (shown by arrows) upon proceeding from preinsertion binary complex (silver) to insertion ternary complex (red). Figure 5 Details of Changes in Dpo4 Little-Finger Domain Contacts with Backbone Phosphates on the oxoG-Modified Template-Primer DNA Associated with the Nucleotide Binding and Incorporation Steps (A) Intermolecular contacts between the little-finger domain (purple) and the template strand (cyan) in the preinsertion binary complex. The phosphate groups interacting with Dpo4 amino acids are highlighted by shaded spheres. The phosphate group of A9 of the template strand is circled for reference. (B) Intermolecular contacts between the little-finger domain (purple) and the primer strand (green) in the preinsertion binary complex. The phosphate groups interacting with Dpo4 amino acids are highlighted by shaded spheres. The phosphate group of G6 of the primer strand is circled for reference. (C) Intermolecular contacts between the little-finger domain (purple) and the template strand (cyan) in the insertion ternary complex. The intermolecular contacts involving the little finger are translocated by one nucleotide step on proceeding from the preinsertion binary complex (in A) to the insertion ternary complex (in C) upon dCTP binding. (D) Intermolecular contacts between the little-finger domain (purple) and the primer strand (green) in the insertion ternary complex. The intermolecular contacts involving the little finger are translocated by one nucleotide step on proceeding from the preinsertion binary complex (in B) to the insertion ternary complex (in D) upon dCTP binding. (E) Intermolecular contacts between the little-finger domain (purple) and the template strand (cyan) in the postinsertion binary complex. The intermolecular contacts involving the little finger are retained on proceeding from the insertion ternary complex (in C) to postinsertion binary complex (in E) on dCTP incorporation. (F) Intermolecular contacts between the little-finger domain (purple) and the primer strand (green) in the postinsertion binary complex. The intermolecular contacts involving the little finger are retained on proceeding from the insertion ternary complex (in D) to postinsertion binary complex (in F) on dCTP incorporation. Figure 6 Details of Changes in Dpo4 Thumb Domain Contacts with Backbone Phosphates on the oxoG-Modified Template-Primer DNA Associated with the Nucleotide Binding and Incorporation Steps (A) Intermolecular contacts between the thumb domain (red) and the template strand (cyan) in the preinsertion binary complex. The phosphate groups interacting with Dpo4 amino acids are highlighted by shaded spheres. The phosphate group of T13 of the template strand is circled for reference. (B) Intermolecular contacts between the thumb domain (red) and the primer strand (green) in the preinsertion binary complex. The phosphate groups interacting with Dpo4 amino acids are highlighted by shaded spheres. The phosphate group of A12 of the primer strand is circled for reference. (C) Intermolecular contacts between the thumb domain (red) and the template strand (cyan) in the insertion ternary complex. The intermolecular contacts involving the thumb are retained on proceeding from the preinsertion binary complex (in A) to the insertion ternary complex (in C) on dCTP binding. (D) Intermolecular contacts between the thumb domain (red) and the primer strand (green) in the insertion ternary complex. The intermolecular contacts involving the thumb are retained on proceeding from the preinsertion binary complex (in B) to the insertion ternary complex (in D) on dCTP binding. (E) Intermolecular contacts between the thumb domain (red) and the template strand (cyan) in the postinsertion binary complex. The intermolecular contacts involving the thumb are translocated by one nucleotide step on proceeding from the insertion ternary complex (in C) to the postinsertion binary complex (in E) upon dCTP incorporation. (F) Intermolecular contacts between the thumb domain (red) and the primer strand (green) in the postinsertion binary complex. The intermolecular contacts involving the thumb are translocated by one nucleotide step on proceeding from the insertion ternary complex (in D) to the postinsertion binary complex (in F) upon dCTP incorporation. The little finger slides and rotates counterclockwise (with respect to the 5′-to-3′ direction of the template strand) as a rigid entity along the DNA duplex, resulting in translocation of the polymerase by one step (Figure 4C), thereby allowing the next template base and incoming dCTP to enter the active site. Upon dCTP binding, the contacts of the little finger with the phosphate groups of residues T8, A9, C10, and C11 (Figure 5A) are relocated to the phosphate groups of C7, T8, A9, and C10 (Figure 5C) on the template strand. At the same time, related little-finger contacts on the primer strand are displaced from the phosphate groups of G5, G6, and A7 (Figure 5B) to the phosphate groups of G6, A7, and T8 (Figure 5D). It is interesting that the contacts between the DNA phosphate groups and the little finger, which is unique among Y-family bypass polymerases, involve electrostatic interactions between alternating arginine and peptide backbone nitrogen groups (Figure 5A and 5C). The little finger provides an interface along which the DNA can more easily slide during the translocation step in a manner that is independent of base sequence (see Figure 4C). The little finger is positioned closer to the primer strand in the preinsertion binary complex than in the insertion ternary complex, thus allowing for additional hydrogen-bonding contacts with the phosphate groups of T8 in the template strand (Figure 5A) and G5 in the primer strand (Figure 5B) in the former complex. By contrast, as a result of the dCTP binding, the α-helical segments of the thumb domain (helixes H, K, J) change their relative position with respect to the palm domain, and even though they rotate counterclockwise relative to the DNA, they maintain the same contacts with the phosphate of T13 of the template strand (Figure 6A and 6C) and with the backbone phosphates of T11 and A12 of the primer strand (Figure 6B and 6D). In quantitative terms, dCTP binding results in rotation of the little finger with respect to the DNA by 29°, almost its full-twist value, while the palm and finger domains, as well as the thumb domain, rotate by 18°, half of their full rotation cycle (Table 3). The little-finger domain, as well as the palm and finger domains, moves along the DNA axis by 3.2–3.5 Å, generating space for nascent base-pair formation. By contrast, the thumb domain, which maintains its contacts with DNA on bonding of dCTP, does not translate with respect to the DNA (Table 3). Table 3 Rotation and Translation of Dpo4 Domains with Respect to DNA for the Preinsertion Binary to Insertion Ternary Step (dCTP Binding) and for the Insertion Ternary to Postinsertion Binary Step (dCTP Incorporation) Structure of the oxoG-Modified Postinsertion Binary Complex with Covalently Incorporated Cytosine Opposite oxoG In the case of the postinsertion binary complex, a covalently incorporated cytosine is positioned opposite the oxoG site (see Figure 2C) and forms a Watson-Crick oxoG•C pair. Residues 3–19 of the template strand can be traced in the electron density map. The finger domain is positioned directly over the Watson-Crick oxoG•C pair with one Ca2+ cation coordinated by the D7, D105, and E106 at the active site of Dpo4 (see Figure 2F). Details of the intermolecular contacts between Dpo4 and the oxoG-containing template-primer junction in the postinsertion binary complex (Figure S1C) are shown schematically in Figure 2I. The Cytosine-Incorporation Step: Conformational Transitions from Insertion Ternary Complex to Postinsertion Binary Complex We compare the structures of the insertion ternary complex that contains the incoming dCTP molecule opposite oxoG with the postinsertion binary complex that contains covalently incorporated C opposite oxoG by superpositioning their DNA duplexes in order to identify the conformational changes associated with translocation of the polymerase during the dCTP incorporation step. This superposition for the entire complex is shown in Figure 7A and 7B, with little-finger-domain–DNA interactions shown in Figure 7C and thumb domain–DNA interactions in Figure 7D. The key observation is that the little-finger domain maintains its contacts with the DNA (Figure 7C), while the thumb domain moves by one nucleotide (shown by arrows in Figure 7D) during covalent incorporation of the cytosine opposite the oxoG lesion site. These conformational changes in the thumb domain on cytosine incorporation are accompanied by the release of the pyrophosphate together with a divalent cation. Figure 7 Conformational Transitions of Dpo4 and oxoG-Modified Template-Primer DNA Associated with the Covalent Incorporation Step Following Superimposition of DNA Duplexes (A) Overall comparative views (following superimposition of DNA duplexes) of the structures of the insertion ternary complex with incoming dCTP (in color) with the postinsertion binary complex after covalent cytosine incorporation (in beige). The DNA-backbone phosphate groups in contact with the Dpo4 little-finger and thumb domains in the ternary complex are shown by colored spheres. (B) Overall comparative view looking down the DNA helix axis. (C) Comparative views of contacts between the little-finger domain and the DNA backbone are maintained upon proceeding from the insertion ternary complex (purple) to the postinsertion binary complex (beige). (D) Comparative views of contacts between the thumb domain and the DNA backbone, which shift by a one nucleotide step (shown by arrows) upon proceeding from the insertion ternary complex (purple) to the postinsertion binary complex (beige). Upon covalent cytosine incorporation, there is retention of little-finger contacts with the phosphate groups of C7, T8, A9, and C10 on the template strand (see Figure 5C and 5E), and the phosphate groups of G6, A7, and T8 on the primer strand (see Figure 5D and 5F). By contrast, the contact of the thumb domain with the phosphate group of T13 (see Figure 6C) is relocated to the phosphate group of A12 (see Figure 6E) on the template strand. At the same time, related thumb-domain contacts on the primer strand are displaced from the phosphate groups of T11 and A12 (see Figure 6D) to the phosphate groups of A12 and G13 (see Figure 6F). In quantitative terms, covalent cytosine incorporation results in rotation of the thumb domain and palm and finger domains by 17 to 18 degrees, half of their full rotation cycle, while the little finger undergoes a minimal rotation of 3° (Table 3). The thumb domain moves along the DNA axis by 2.7 Å, while the little-finger domain, which maintains its contacts with DNA on cytosine incorporation, and palm and finger domains, undergo minimal translation (0.5 Å) with respect to the DNA (Table 3). dCTP Positioning in Insertion Ternary Complex The position of the substrate, dCTP, with respect to the 3′-end nucleotide, G13, on the primer strand, differs substantially from a B-DNA step, with a reduction in the twist angle and displacement of the substrate base toward the DNA major groove (Figure 8A). The view looking down the helix axis clearly demonstrates this shifted alignment of dCTP relative to the 3′ portion of the primer strand (Figure 8B), with the sugar ring of dCTP positioned significantly closer to the sugar ring of G13, when compared with the typical B-DNA separation between A12 and G13. Upon covalent base incorporation, the restoration of the B-DNA conformation between the nascent base pair and its adjacent base pair (Figure 8C) propagates toward the 5′ end of the primer strand. When viewed down the helix axis, the separations between the sugar rings of G13 and the newly incorporated base, C14, and between G13 and A12, are approximately equal (Figure 8D). Figure 8 Relative Positioning of the oxoG•dCTP Pair in the Insertion Ternary Complex and the oxoG•C Pair in the Postinsertion Binary Complex (A) View of the (oxoG6-C7-T8)•(A12-G13) segment plus dCTP and Y12 in the modified ternary complex. The aromatic ring of Y12 is packed against the sugar ring of the incoming dCTP. (B) View of (A) looking down the helix axis. (C) View of the (oxoG6-C7-T8)•(A12-G13-C14) segment and Y12 in the postinsertion binary complex. (D) View of (C) looking down the helix axis. (E) Superposition of the (oxoG6-C7-T8)•(A12-G13) segment plus dCTP and Y12 of the modified ternary complex (in color) with the (oxoG6-C7-T8)•(A12-G13-C14) segment and Y12 of the posinsertion binary complex (in beige) by the oxoG6-C7-T8 and G6-C7-T8 template strands segments. Note the relative positioning of the sugar ring of dCTP/C14 and the aromatic ring of Y12 in the two complexes. The movement of primer strand 3′-terminal residue with respect to Dpo4 palm and finger domains is defined by the restraining function of the aromatic ring of Y12, which is packed against the sugar ring of the dCTP in the insertion ternary complex and against the sugar ring of newly incorporated C14 base in the postinsertion binary complex (Figure 8E). The conformational transition accompanying a restoration of the B-DNA step results in the contacts between the α-helical segments of the thumb domain and DNA phosphate backbone shifting by one step. Base-Pair Progression during dCTP Binding and Cytosine-Incorporation Steps The DNA molecule retains a B-like conformation in the preinsertion binary, insertion ternary, and postinsertion binary complexes. However, there is a narrowing of the minor groove at the contact sites with the thumb domain and a widening of the major groove at the contact sites with the little finger, which is observed in all complexes, and which increases in the insertion ternary complex relative to both preinsertion binary and postinsertion binary complexes. The C7•G13 pair is adjacent to the oxoG lesion site, and its positioning within/adjacent to the Dpo4 active site is of interest in the three complexes. The preinsertion binary and insertion ternary complexes associated with dCTP binding can be compared by superpositioning their finger and palm domains, which constitute the Dpo4 active site (Figure 9A). The positioning of the C7•G13 pair relative to the active site for the two complexes is shown in Figure 10A. The template strand C7 of the preinsertion binary complex (in silver) is rotated and displaced along the helical axis to a new position in the insertion ternary complex (in color), while primer strand G13 is only displaced to maintain Watson-Crick hydrogen bonding, but is minimally rotated upon dCTP binding. Figure 9 Conformational Transitions of Dpo4 and oxoG-Modified Template-Primer DNA Associated with Nucleotide Binding and Covalent Incorporation Steps Following Superimposition of Dpo4 Palm and Finger Domains (A) Overall comparative views (following superimposition of Dpo4 palm and finger domains) of the structures of the preinsertion binary complex (in silver) and the insertion ternary complex with incoming dCTP (in color). The DNA-backbone phosphate groups in contact with the Dpo4 little-finger and thumb domains are shown as spheres. Upon dCTP binding, DNA slides one step along the little-finger domain, and the oxoG residue of the template strand of the insertion ternary complex takes the place of C7 in the preinsertion binary complex. (B) Overall comparative views (following superimposition of Dpo4 palm and finger domains) of the structures of the insertion ternary complex with incoming dCTP (in color) with the postinsertion binary complex after covalent cytosine incorporation (in beige). Figure 10 Comparison of the Relative Progression (Twist Angle Viewed down the Helix Axis) of C7•G13 Pairs between Complexes Superpositioned on their Finger and Palm Domains (A) C7•G13 pairs in the preinsertion binary (silver) and insertion ternary (color) complexes. (B) C7•G13 pairs in the insertion ternary (color) and postinsertion binary (beige) complexes. (C) C7•G13 pairs in the preinsertion binary (silver) and postinsertion binary (beige) complexes. The insertion ternary and postinsertion binary complexes associated with dCTP incorporation can be compared by superpositioning their Dpo4 finger and palm domains (see Figure 9B). The relative positioning of the C7•G13 pair for the two complexes is shown in Figure 10B. Primer strand G13 of the insertion ternary complex (in color) is rotated and displaced to a new position in the postinsertion binary complex (in beige), while template strand C7 is only minimally displaced and rotated to maintain Watson-Crick hydrogen bonding. Superposition of the preinsertion binary with the postinsertion binary complexes by their Dpo4 palm and finger domains (Figure S3A) results in almost ideal overlap between the DNA duplexes, with rotation and displacement with respect to the DNA helical axis by one full nucleotide step (Figure S3B). This can be best visualized by viewing the positioning of the C7•G13 pairs of the preinsertion binary (in silver) and postinsertion binary (in beige) complexes (Figure 10C). Discussion The structures of the active sites of the preinsertion binary complex with no base opposite oxoG (see Figure 2D), the insertion ternary complex with incoming dCTP opposite oxoG (see Figure 2E), and the postinsertion binary complex following covalent C incorporation opposite oxoG (see Figure 2F) provide snapshots of the unique conformational changes of the Y-family Dpo4 polymerase as it translocates along the DNA molecule during one cycle of replication opposite the oxoG lesion (Figure 11). Figure 11 Schematic of Translocation Mechanism of Dpo4 Complexes with oxoG-Modified Template-Primer DNA (A) oxoG-modified preinsertion binary complex, where the overhang segment of the template strand, including oxoG, is disordered in the crystal (shown by light coloration). (B) oxoG-modified insertion ternary complex with incoming dNTP. (C) oxoG-modified postinsertion complex following covalent nucleotide incorporation. The arrows represent one nucleotide translocation of individual color-coded domains in the 3′-to-5′ direction of the template strand. Color-coding. DNA: template strand, cyan; primer strand, green; dCTP, light gray; oxoG residue, orange. Dpo4 domains: palm, light green; finger, blue; thumb, red; little-finger, purple; tether-connecting-thumb and little-finger, gray; Ca2+ cations, pink spheres. The dCTP-Binding Step: General Overview The catalytic mechanism of Dpo4 involves movement of the bypass polymerase relative to DNA during both nucleotide binding (Figure 11A and 11B) and incorporation (Figure 11B and 11C) steps. Screw-like rotation of the polymerase counterclockwise with respect to the 5′-to-3′ direction of the template strand opens space for the next template residue to enter under the finger domain (movement of finger, palm, and little-finger domains shown schematically by arrows on blue, green, and purple backgrounds, respectively, in Figure 11B) and for the complementary incoming nucleotide to form a base pair with it. The incoming dNTP in the insertion ternary complex replaces the 3′ nucleotide in the preinsertion binary complex, since there is no open space to accommodate the entire incoming nucleotide, as in processive polymerases. The binding of dNTP could occur through initial positioning of its triphosphate, given that its binding site is the only part of the catalytic binding surface accessible prior to nascent base-pair formation. At this stage, the dimensions of the binding pocket discriminate against mispairs, while nascent base-pair formation involving complementary hydrogen-bond formation locks the template base and prevents it from sliding back out of the catalytic position. The thumb and little-finger domains hold the DNA helix from opposite sides in bypass polymerase–DNA complexes and are the key determinants whose contacts most affect the relative rotation of the enzyme around DNA during dNTP-binding and nucleotide-incorporation steps. During the dNTP-binding step associated with nascent base-pair formation, the DNA is anchored to the thumb domain, and slides by one step along the little-finger domain (movement of little-finger domain is shown schematically by an arrow on purple background, Figure 11B). Since there are changes in the relative interdomain positions during this step, the helical axis of DNA is not a global axis for rotation of the complex. The bending in the DNA helical axis, due to additional narrowing of the DNA minor groove facing the thumb domain, and widening of the major groove facing the little-finger domain, help to accommodate the incoming nucleotide between the thumb and finger domains, while retaining thumb-domain–DNA contacts. The above outline describes the translocation and conformational changes associated with the first step and formation of a nascent base pair in a two-step sliding mechanism. Mechanisms of Nucleotide Selection by the Dpo4 Polymerase Dpo4 imposes geometrical constraints on the length of the nascent base pair. From one side of the nascent base pair, the palm and finger domains define the position of the incoming nucleotide. There is shape-complementarity within the dNTP-binding pocket, because the dNTP sugar is packed against the aromatic side chain of tyrosine 12 and the triphosphate backbone with the chelated divalent cation precisely positioned within a channel lined by basic and acidic polar amino acids (see Figure 3B). From the other side of the nascent base pair, the little-finger domain holds the phosphate group of the template base. As a result, a longer purine–purine mispair would place the template base and incoming nucleotide phosphates farther apart, while a pyrimidine–pyrimidine mispair would bring them closer together. In both cases, the alignment of the active site in the ternary complex would be disrupted, including the placement of the second metal ion that accompanies the arrival of the dNTP at the binding site. The correct Watson-Crick base pair fits properly within the active site that is defined by the steric constraints that yield a ~10−4 error rate when an unmodified template strand is replicated. This selectivity is achieved despite the fact that the contour of the nascent base pair is such that its major and minor groove edges have minimal contacts with the surface of the bypass polymerase. The dCTP-Binding Step: Mechanistic Differences between Replicative and Bypass Polymerases In replicative polymerases (Figure S4A), a dNTP binds to the open and accessible active site of a binary complex (Figure S4B) to form a base pair with the next available, unpaired template base n, with the O helix of the finger domain closing on the flat surface of the nascent base pair to form a tight-fitting, solvent-excluding, reaction-ready, ternary complex (Figure S4C) [16]. This conformational change is accompanied by a displacement of the template base n from a polymerase preinsertion pocket to a stacked position within the DNA helix [37]. A conserved tyrosine residue located at the base of the O helix facilitates this transition by vacating its stacking alignment with the (n−1) template base in the binary complex and occupying the polymerase preinsertion pocket in the ternary complex, thereby preventing a slippage of the next upstream template base into the active site. There are only minimal changes in the protein–DNA contacts beyond the base-pair recognition site during dNTP incorporation catalyzed by replicative polymerases. By contrast, there is no open and accessible binding site for dNTP binding in the preinsertion binary complex of the Dpo4 bypass polymerase. Instead, the incoming dNTP must be inserted into the site previously occupied by the 3′-end primer residue. This position becomes accessible as a result of a rotating and sliding motion of the DNA along the little-finger domain by one nucleotide (see Figure 4C), while retaining phosphate contacts with the thumb domain (see Figure 4D), thereby placing the next template base (oxoG or unmodified G) in a stacked conformation under the finger domain. The binding of dCTP locks this position, orienting it for the subsequent nucleotidyl transfer reaction. There are no local conformational changes within the finger domain when dCTP binding occurs, and, therefore, there is no transition from an open to a closed complex in this bypass polymerase, since the finger domain does not undergo any conformational changes. The Cytosine-Incorporation Step: General Overview The incoming dNTP, whose base is displaced toward the major groove and exhibits a substantially reduced twist angle, is inclined such that its catalytically competent phosphate oxygen is poised for coordination with both the 3′-end nucleotide of the primer strand and divalent cations (see Figure 8A and 8B). The conformational change associated with the transition from the underwound position of the bound dNTP to the restored B-DNA geometry on phosphodiester bond formation (see Figure 8C and 8D) is propagated bidirectionally, perhaps analogous to the release of a compressed spring. There is an increased separation between the sugar ring of dNTP, which is packed against and restrained by the aromatic ring of Tyr12, and the sugar ring of the 3′-end nucleotide of the primer strand, on phosphodiester bond formation (see Figure 8E), thereby initiating the translocation of the α-helical contacts of the thumb domain by one step along the phosphodiester backbones of both primer and template strands (movement of thumb domain is shown schematically by an arrow on red background, Figure 11C). The above outline describes the translocation and conformational changes associated with the second step, covalent bond formation, in a two-step sliding mechanism. The Cytosine-Incorporation Step: Mechanistic Differences between Replicative and Bypass Polymerases Phosphodiester bond formation in replicative polymerases releases pyrophosphate, thereby breaking apart the triphosphate backbone-mediated interactions between the palm and finger domains. This results in the movement of the O helix of the finger domain back to the open state, while positioning the aromatic ring of a tyrosine for stacking over and pushing against the newly formed base pair, thus triggering translocation. The polymerase translocates upward by one base-pair step in the 3′ to 5′ direction of the template strand, and the minor groove edge of the nascent base pair is examined for correct Watson-Crick pairing alignment by the proofreading apparatus of the polymerase [16,37]. In the case of Dpo4, release of the pyrophosphate is not accompanied by breaking apart the triphosphate backbone-mediated interactions between the palm and finger domains. The formation of the phosphodiester bond instead triggers rotation of the entire DNA molecule relative to the polymerase active site, formed by the finger and palm domains, in order to create a greater distance and thus diminish repulsive interactions involving phosphate groups between the newly incorporated C residue and the adjacent nucleotide on the primer strand. As a result, the contacts between the thumb domain and the backbone DNA phosphate groups shift by one nucleotide step (see Figure 7D). The nucleotidyl transfer reaction and release of the pyrophosphate provide the driving force for this transition. The nascent base pair retains its contacts with the finger domain (see Figure 7C), and there are no proofreading interactions between this base pair and the Dpo4 polymerase. The structures of the unmodified preinsertion binary and insertion ternary Dpo4 complexes are identical to their oxoG-modified counterparts with the exception of a different location of the oxoG phosphate group and Arg 332 in the insertion ternary complex (see Figure 3F). Thus, the above-described mechanism of dNTP binding to the preinsertion binary complex to form the insertion ternary complex applies not only in the case of the oxoG-lesion but also to unmodified DNA templates. Structure-Based Mechanistic Insights into Dpo4 Insertion Accuracy Opposite the oxoG Lesion and Comparison with Replicative Polymerases Dpo4 and the Y-family yeast and human pol η [32] are able to achieve a 100-fold preference for inserting dCTP over dATP opposite oxoG. However, replicative T7 [38] and Rb69 [18] incorporate dCTP only two to seven times more frequently than dATP [18,34,38], the gap-filling pol β has a slight preference for inserting dATP [39], while replicative pol α, pol δ, Bacillus pol I, and HIV-1 reverse transciptase predominantly insert dATP [29,35,38,39]. The polymerases pol β, Rb69, T7, and Dpo4 avoid the steric clash of the O8 atom with the sugar-phosphate backbone that arises in the oxoG(anti) •dCTP base pair either by a sharp kink in the single-stranded template overhang within Rb69 [18] and T7 polymerase [34] active sites, or, in the case of Dpo4 (this study) and pol β [33], by a flipping of the phosphate group of oxoG by 180°. However, Dpo4 forms six hydrogen bonds with oxoG(anti) in the active site of Dpo4 (five with the phosphate group of oxoG and one with the O8 atom through a water bridge, see Figure 3D), while pol β [33] and Rb69 [18] do not make any specific contacts with oxoG(anti), and T7 forms only one hydrogen bond with O8 of oxoG by the relocated side chain of a lysine residue [34]. Thus, the multiple and favorable contacts of oxoG(anti) with amino acids within the Dpo4 active site appear to be the most important factor responsible for the remarkable fidelity of replication of oxoG lesion catalyzed by Dpo4. The uniqueness of the Dpo4 base-insertion mechanism among the polymerases studied in detail provides a second factor for the preferred incorporation of dCTP opposite oxoG. We hypothesize that, prior to dNTP binding, the next available template base to be replicated interconverts between a looped-out conformation in the single-stranded overhang and a base–base stacking conformation involving the 3′ base in the template strand. Once the oxoG base becomes inserted into the helix in this manner, we hypothesize that the more favorable contacts of amino acid residues in Dpo4 with oxoG(anti) rather than with oxoG(syn) drive the (anti)–(syn) equilibrium toward the anti conformation. By contrast, the mechanism of replicative polymerases, as well as gap-filling pol β, do not allow any control over the conformation of oxoG before nucleotide binding. In Rb69, T7, and Bacillus pol I, the next to-be-replicated template base n is prevented from stacking into the helix by the conserved tyrosine at the base of the O helix [37]. This tyrosine is displaced out of its position by the template base n in the same timeframe as formation of a base pair with an incoming dNTP. An incoming dCTP favors base-pairing with oxoG in the anti conformation (Figure S5A), while an incoming dATP would facilitate pairing with oxoG in the syn conformation (Figure S5B), both occurring before the O helix is closed and the active site is fully assembled. Thus, the conformation of the oxoG residue would be selected by its base-pairing partner. In the pol β binary complex with a single-nucleotide gapped substrate [40], the next template base n is stacked into the helix. Thus, the oxoG conformation might be already fixed even before the dNTP is bound. However, neither the phosphate group of the oxoG residue [33], nor the unmodified G [40], form any contacts with pol β. Apparently, in the case of pol β, as well as T7, Rb69, and Bacillus pol I, there is no mechanism available that can influence potential (anti)–(syn) equilibrium of oxoG before dNTP is bound, as is the case in Dpo4. Dpo4 demonstrates the ability to achieve at least 10-fold higher accuracy of insertion of dCTP opposite oxoG than the high-fidelity T7 and Rb69 polymerases. This is achieved by imposing an anti conformation on the oxoG lesion that, in turn, favors pairing with dCTP, thus resulting in a non-mutagenic bypass of oxoG catalyzed by Dpo4. Furthermore, the binding of dCTP further stabilizes oxoG(anti) due to a higher thermodynamic stability of the oxoG(anti) •dCTP (Figure S5A) over the oxoG(syn) •dATP (Figure S5B) mispair at the template-primer junction [41]. Conclusions The outlined research allows us to draw three major conclusions that add to our current understanding of lesion bypass by Y-family Dpo4 polymerase. First, two arginine residues (R331 and R332) of Dpo4 play a critical role in accommodating the oxoG lesion and facilitating an anti conformation for Watson-Crick base pairing with dC. Second, in the stepwise nucleotide incorporation opposite an oxoG, reflecting transitions from preinsertion binary to insertion ternary, and subsequently to postinsertion binary complexes, the little-finger and thumb domains move one at a time, tracking the template and primer strand translocation separately. Third, the dCTP bound in the insertion stage is not properly base-stacked with the DNA duplex as in regular B-form DNA. This suggests that the ternary complex may need to undergo subtle conformational changes prior to the chemistry step of phosphoryl transfer. Materials and Methods Preparation and purification of template and primer DNA strands The DNA 19-mer template 5′- CTAAC[G] CTACCATCCAACC-3′ with [G] = oxoG or G, with a 3′-OH terminus was synthesized using an automated Applied Biosystems (Foster City, California, United States) 392 DNA synthesizer with phosphoramidite chemistry, cleaved from the support, deprotected with ammonium hydroxide for 5 h at 55 °C (with 0.25 M 2-mercaptoethanol for 17 h for the oxoG-modified oligonucliotide to avoid any oxidative degradation of the oxoG site), purified on 20% polyacrylamide gel in the presence of 8 M urea, electro-eluted, and desalted. 2′,3′-dideoxy-G at the 3′ end was introduced into the 13-mer primer 5′- GGTTGGATGGTAG-3′ by reverse 5′-to-3′ synthesis using 5′-CE phosphoramidites, the 14-mer primer with 2′,3′-dideoxy-C-3′ terminal was synthesized using 2′,3′-ddC columns by regular 3′-to-5′ synthesis, and purified as described. All phosphoramidites were purchased from Glen Research (Sterling, Virginia, United States). [γ−32P]ATP (specific activity, >3000 Ci/mmol) was obtained from NEN LifeScience Products (Perkin-Elmer, Boston, Massachusetts, United States), T4 polynucleotide kinase was purchased form Amersham Pharmacia Biotech (Piscataway, New Jersey, United States), and the reagents for crystallization were obtained from Hampton Research (Aliso Viejo, California, United States). Preparation and purification of Dpo4 The DNA fragments encoding 353 amino acid full-length Dpo4 (plus 5 glycines at the N-terminus) were obtained by PCR and inserted into pET-28a between NdeI and EcoRI cleavage sites. The His-tag version of Dpo4 was expressed in E. coli BL21–CodonPlus (DE3)-RIL strain (Stratagene, La Jolla, California, United States), and then subjected to a Ni-chelating column. The 6-His tag was next cleaved with thrombin, and the Dpo4 protein was then further purified by heparin column chromatography, followed by passage through a Superdex-200 column. Dpo4 was concentrated to 26 mg/ml in 5 mM DTT, 25 mM HEPES (pH 7.5), and 300 mM NaCl. Primer elongation standing-start assay The template-primer DNA complexes formed by annealing a 19-mer template containing oxoG (or G at the same site) with a 32P 5′-end-labeled 3′-OH terminated primer (13-mer ending one base before the oxoG site, or 14-mers with C or A opposite G or oxoG), were incubated with Dpo4 polymerase in the presence of all four dNTPs. Aliquots were withdrawn from the reaction mixture after 1, 3, 5, 10, 15, and 20 min incubation times and quenched by a gel-loading buffer (95% formamide with 20 mM EDTA, 45 mM Tris-borate, 0.1% bromophenol blue, 0.1% xylene cyanol). For a typical experiment, a 12-μl solution of all four dNTPs was added to the 12-μl polymerase plus template-primer DNA mixture, both solutions were in 100 mM HEPES (pH 8.0), 5 mM MgCl2, 60 mM NaCl, and 5 mM dithiothreitol. The polymerization reactions were conducted at 30 °C, final template-primer DNA concentration was 10 nM, Dpo4 was 1 nM, and each dNTP 100 μM. The reaction products were resolved on 20% polyacrylamide gels in the presence of 8 M of urea, and the gels were dried before radiography with Fuji image plate (Fuji Photo Film Company, Tokyo, Japan). The images were scanned on Fuji PhosphoImager (Fuji), and the bands were quantified using profile analysis mode in ImageGauge software (Fuji). Steady-state kinetic analysis of one-base insertion and extension Steady-state kinetics parameters were analyzed for incorporation of each deoxynucleotide opposite the G and oxoG and for extension of primers with C or A 3′-terminal opposite G and oxoG in the presence of the next correct nucleotide, 2′-deoxyguanosine 5′-triphosphate (dGTP), as described in [42]. In a typical experiment, 6-μl solutions with increasing concentrations of single dATP, dTTP, dGTP, or dCTP were added to 6 μl of polymerase/template-primer DNA mixture, and the reactions were stopped after 3 min by the addition of 12 μl of denaturing loading buffer. To ensure single-hit polymerization conditions (less than 20% primer extended), the nucleotide concentration interval was adjusted for every experiment. The template-primer DNA concentrations were 50 nM, and Dpo4 was 1 nM. The gel-band intensities of extended and unreacted primer strands were quantified as described. The reaction rates (ν, nM/min) were plotted as a function of the dNTP concentration, and the data were fit by nonlinear regression of the Michaelis-Menten equation, to calculate apparent KM and Vmax steady-state parameters [32,42]. The frequency of nucleotide incorporation (finc) and extension (fext) were defined as follows: Crystallization The crystals of the Dpo4 ternary complexes containing oxoG (or G) 19-mer templates and the 13-mer primer terminated by 2′,3′-dideoxyguanosine were grown in the presence of dCTP. The preinsertion binary complexes in the absence of dCTP, the postinsertion binary complex with the oxoG-modified template, and the 14-mer primer 3′-terminated with 2′,3′-dideoxycytosine were grown under conditions similar to those described in the literature [5]. Briefly, template-primer DNAs were annealed and mixed with Dpo4 in 1.2:1 molar ratio to a final concentration 0.15 mM in 20 mM HEPES (pH 7.0), 60 mM NaCl, 5 mM MgCl2, 1mM DTT, and dCTP (1 mM) was added to produce the ternary complexes. The protein-DNA complexes were then incubated at 37 °C for 5 min and centrifuged at 11,000 rpm for 7 min at 4 °C. The crystals were grown by the hanging-drop method against a reservoir solution containing 100 mM HEPES (pH 7.0), 100 mM calcium acetate, and 10% PEG 4000 at 20 °C. Several rounds of micro seeding (Hampton Research kit) were employed to produce the diffraction quality crystals of the short binary complexes. The crystals were transferred to a cryosolution containing the mother liquor with 15% PEG 4000 and 15% ethylene glycol and flash-frozen in liquid nitrogen for X-ray data collection. When compared with the type I structure determined by Yang and colleagues [5], the little-finger domain in both of our oxoG-damaged and unmodified ternary complexes is shifted by up to 0.5 Å and rotated by 5° toward the finger domain, leading to a concomitant shift of the DNA duplex. However, the contacts with the equivalent phosphate groups of the DNA backbone are maintained. This repositioning of the little finger is likely due to the more spacious crystal lattice with a 10% higher solvent content and the different DNA sequence in our case. Similar rearrangements of the little-finger domain were noticed for crystals of ternary complexes characterized by different lattices [12]. Thus, RMSD of polymerase Cα atoms between our oxoG-modified, and the unmodified ternary type I complex reported previously [5], is only 0.368 Å, excluding the little finger, and 0.667 Å for all 341 residues. Structure determination and refinement X-ray diffraction data were collected either at an Advanced Photon Source 14-ID beam line (Argonne National Laboratory, Chicago, United States) or with an in-house Rigaku R-axis IV detector (Rigaku, The Woodlands, Texas, United States) mounted on an RU 200 generator. The data were processed and scaled using HKL2000 suite (Structural Biology Center, Argonne National Laboratory). The crystal of the oxoG-modified insertion ternary complex with incoming dCTP belonged to the P21 space group, and exhibited unit cell parameters not observed previously for Dpo4-DNA complexes [5,10–13]. The structure was solved by molecular replacement (AMoRe [43]) using Dpo4-DNA-ddADP type I structure [5] as a search mode. The little-finger domain (residues 244–341) was split from the complex and its position refitted by AMoRe. The model building, including substitution of the DNA sequence, was manually finished in TURBO based on the electron density map calculated in X-PLOR, and the resulting model was refined in REFMAC [44] at 1.95 Å resolution to a final Rfactor/Rfree of 0.226/0.253 (see Table 2). The structure of the oxoG-modified preinsertion binary complex was solved by molecular replacement in the P21 space group (see Table 2), employing refined oxoG-modified insertion ternary complex as a search model. After the initial positions of both molecules in the asymmetric unit cell were defined, the model was split into Dpo4 domains (the palm-finger, little-finger, and thumb) and the DNA duplex, and the new positions of the thumb and little-finger domains were found. The template-primer DNA was retraced in TURBO based on the electron density map produced by CNS after rounds of refinement with the rigid group, rigid pair, and annealing routines in CNS. The crystal displayed pseudo-merohedral twinning [45] with twin fraction of 0.4, as estimated from the Britton plot generated by the DETWIN module of CCP4. The data were detwinned (Protocol S1), and the model was refined in REFMAC at 2.35 Å resolution to Rfactor/Rfree of 0.249/0.318 (see Table 2). Despite the relatively high R-factors (probably a consequence of twinning), the final electron-density map clearly showed molecular details. Molecular replacement with the oxoG-modified insertion ternary complex as a search model was used to solve the structure of the oxoG-modified postinsertion binary complex with C opposite the lesion site. The final structure was refined by REFMAC to 2.65 Å and Rfactor/Rfree of 0.238/0.287 (see Table 2). The structures of the unmodified insertion ternary complex with incoming dCTP and preinsertion binary complexes with primers terminated with dideoxy- and deoxyguanine were solved by molecular replacement employing the structures of oxoG-modified ternary and binary complexes as search models, and refined by REFMAC to values of R-factors indicated in Table 2. Detwinning in the case of an unmodified preinsertion complex did not improve refinement statistics. However, careful processing of the data set allowed us to refine the structure to Rfactor/Rfree 0.271/0.310 (see Table 2). Structure analysis Dpo4 conformations in the ternary and binary complexes were compared by the program DynDom to determine protein domains, hinge axes, and amino acid residues involved in the hinge bending [46]. Parameters of DNA duplexes were calculated by the program Curves. Supporting Information Figure S1 Structures of Complexes of Dpo4 with oxoG-Modified Template-Primer DNA (A) Overall structure of the preinsertion binary complex. Template residues 1–6 (including oxoG) of single-stranded overhang are disordered in the electron density maps. (B) Overall structure of the insertion ternary complex, with incoming dCTP paired with oxoG(anti) at the active site. (C) Overall structure of the postinsertion binary complex, with covalently incorporated C paired with oxoG(anti). The DNA and protein are in stick and ribbon representations, respectively. Note that there is a 2′,3′-dideoxy residue at the 3′ end of the primer strand in all three complexes. Color-coding for top lane. DNA: template strand, cyan; primer strand, green; dCTP, light gray; oxoG residue, orange. Dpo4 domains: palm, light green; finger, blue; thumb, red; little-finger, purple; tether-connecting-thumb and little-finger, gray; Ca2+ cations, pink spheres. (6.3 MB TIF). Click here for additional data file. Figure S2 Structures of Complexes of Dpo4 with Unmodified Template-Primer DNA (A) Overall structure of the preinsertion binary complex. The DNA and protein are in stick and ribbon representations, respectively. Template residues 1–6 (including G) of single-stranded overhang are disordered in the electron density maps. (B) Overall structure of the insertion ternary complex with incoming dCTP. The DNA and protein are in stick and ribbon representations, respectively. Note that there is a 2′,3′-dideoxy residue at the 3′ end of the primer strand in both complexes. Color-coding for top lane. DNA: template strand, cyan; primer strand, green; dCTP, light gray; oxoG residue, orange. Dpo4 domains: palm, light green; finger, blue; thumb, red; little-finger, purple; tether-connecting-thumb and little-finger, gray; Ca2+ cations, pink spheres. (C) Details of the active site of the preinsertion binary complex. (D) Details of the active site of the insertion ternary complex with incoming dCTP. Note that there is a 2′,3′-dideoxy residue at the 3′ end of the primer strand in both complexes. (6.6 MB TIF). Click here for additional data file. Figure S3 Structural Equivalence of Preinsertion and Postinsertion oxoG-Modified Binary Complexes (A) Overall comparative views (following superimposition of Dpo4 palm and finger domains) of the structures of the preinsertion binary complex (in silver) and the postinsertion binary complex (in beige). The DNA-backbone phosphate groups in contact with the Dpo4 little-finger and thumb domains are shown by spheres, in silver for preinsertion and in beige for postinsertion complexes. (B) Top five DNA duplex base pairs (following superimposition of Dpo4 palm and finger domains) of the preinsertion binary complex (in silver) and the postinsertion binary complex (in beige). Upon completion of a Dpo4 catalytic cycle, the DNA duplex is shifted one full base-pair step, so that the template strand residues from oxoG to C10 of the postinsertion complex take the place of C7 to C11 in the preinsertion binary complex, while primer strand C14 to G10 of the postinsertion complex takes the place of G13 to G9 of the preinsertion complex. The DNA-backbone phosphate groups in contact with the Dpo4 little-finger and thumb domains are shown by transparent spheres. (4.4 MB TIF). Click here for additional data file. Figure S4 Conformational Changes in pol I Family Klentaq Polymerase upon Nucleotide Binding Involving Transition from an Open Binary to Closed Ternary Complex (A) Overall comparative view (following superimposition of palm domains) of the structures of the open binary complex (in silver) (PDB ID: 4KTQ), and closed ternary complex with a bound dCTP (in color) (PDB ID: 3KTQ) [S2]. On dNTP binding, the finger domain (blue in ternary complex) rotates as a rigid body by 6° toward the palm domain (light green in the ternary complex), while the O helix (colored light yellow in the binary complex and bright yellow in the ternary complex) moves by 40° to form the closed ternary complex. (B) The Klentaq active site in open binary complex. (C) The Klentaq active site in closed ternary complex. (6.4 MB TIF). Click here for additional data file. Figure S5 Base-Pairing Alignments of 8-oxo-Guanine (oxoG) (A) Hydrogen-bonding alignments in oxoG(anti)•C(anti) pair. (B) Hydrogen-bonding alignments in oxoG(syn)•A(anti) pair. (477 KB TIF). Click here for additional data file. Protocol S1 Experimental Procedures: Full Description of the Twinning Problem for Preinsertion Binary Complexes (28 KB DOC). Click here for additional data file. Accession Numbers Coordinates for the oxoG-modified binary preinsertion (2ASJ), ternary insertion (2ASD), and binary postinsertion (2ASL) complexes; and unmodified binary preinsertion (2AU0); and ternary insertion (2ATL) complexes have been deposited in the Protein Data Bank (http://www.rcsb.org/pdb). The research was supported by National Institutes of Health grants CA46533 to DJP, CA99194 to NEG, CA75449 to SB, and Ruth L. Kirschstein National Research Service Award F32 GM069152 to OR. We would like to thank the staff at beamline 14-ID of the Advance Photon Source (APS) supported by the U.S. Department of Energy for assistance with data collection. Competing interests. The authors have declared that no competing interests exist. Author contributions. OR conceived and designed the experiments. OR performed the experiments. OR, YC, VK, SB, NEG, and DJP analyzed the data. OR and YC contributed reagents/materials/analysis tools. OR wrote the paper with the help of SB, NEG, and DJP. LM contributed in the crystallographic part: general supervision, molecular replacement, and refinement of several structures. SB proposed to study an oxoG lesion bypass by Dpo4 polymerase, discussed the results, and edited the paper. NEG discussed the results and their general significance and edited the paper. DJP conducted general supervision of the project. Citation: Rechkoblit O, Malinina L, Cheng Y, Kuryavyi V, Broyde S, et al. (2006) Stepwise translocation of Dpo4 polymerase during error-free bypass of an oxoG lesion. PLoS Biol 4(1): e11. Abbreviations dATP2′-deoxyadenosine 5′-triphosphate dCTP2′-deoxycytidine 5′-triphosphate dNTPdeoxyribonucleotide triphosphate Dpo4DNA polymerase IV Gguanine oxoG8-oxoguanine RMSDroot mean square deviation ==== Refs References Prakash S Johnson RE Prakash L Eukaryotic translesion synthesis DNA polymerases: Specificity of structure and function Annu Rev Biochem 2005 74 317 353 15952890 Bebenek K Kunkel TA Functions of DNA polymerases Adv Protein Chem 2004 69 137 165 15588842 Plosky BS Woodgate R Switching from high-fidelity replicases to low-fidelity lesion-bypass polymerases Curr Opin Genet Dev 2004 14 113 119 15196456 Friedberg EC Lehmann AR Fuchs RP Trading places: How do DNA polymerases switch during translesion DNA synthesis? Mol Cell 2005 18 499 505 15916957 Ling H Boudsocq F Woodgate R Yang W Crystal structure of a Y-family DNA polymerase in action: A mechanism for error-prone and lesion-bypass replication Cell 2001 107 91 102 11595188 Silvian LF Toth EA Pham P Goodman MF Ellenberger T Crystal structure of a DinB family error-prone DNA polymerase from Sulfolobus solfataricus Nat Struct Biol 2001 8 984 989 11685247 Trincao J Johnson RE Escalante CR Prakash S Prakash L Structure of the catalytic core of S. cerevisiae DNA polymerase eta: Implications for translesion DNA synthesis Mol Cell 2001 8 417 426 11545743 Nair DT Johnson RE Prakash S Prakash L Aggarwal AK Replication by human DNA polymerase-iota occurs by Hoogsteen base-pairing Nature 2004 430 377 380 15254543 Uljon SN Johnson RE Edwards TA Prakash S Prakash L Crystal structure of the catalytic core of human DNA polymerase kappa Structure (Camb) 2004 12 1395 1404 15296733 Ling H Boudsocq F Plosky BS Woodgate R Yang W Replication of a cis–syn thymine dimer at atomic resolution Nature 2003 424 1083 1087 12904819 Ling H Sayer JM Plosky BS Yagi H Boudsocq F Crystal structure of a benzo[a]pyrene diol epoxide adduct in a ternary complex with a DNA polymerase Proc Natl Acad Sci U S A 2004 101 2265 2269 14982998 Ling H Boudsocq F Woodgate R Yang W Snapshots of replication through an abasic lesion; structural basis for base substitutions and frameshifts Mol Cell 2004 13 751 762 15023344 Trincao J Johnson RE Wolfle WT Escalante CR Prakash S Dpo4 is hindered in extending a GT mismatch by a reverse wobble Nat Struct Mol Biol 2004 11 457 462 15077104 Zang H Goodenough AK Choi JY Irimia A Loukachevitch LV DNA adduct bypass polymerization by Sulfolobus solfataricus DNA polymerase Dpo4. Analysis and crystal structures of multiple base-pair substitution and frameshift products with the adduct 1,N2-ethenoguanine J Biol Chem 2005 280 29750 29764 15965231 Vaisman A Ling H Woodgate R Yang W Fidelity of Dpo4: Effect of metal ions, nucleotide selection and pyrophosphorolysis EMBO J 2005 24 2957 2967 16107880 Steitz TA Yin YW Accuracy, lesion bypass, strand displacement and translocation by DNA polymerases Philos Trans R Soc Lond B Biol Sci 2004 359 17 23 15065652 Johnson KA Conformational coupling in DNA polymerase fidelity Annu Rev Biochem 1993 62 685 713 7688945 Freisinger E Grollman AP Miller H Kisker C Lesion (in)tolerance reveals insights into DNA replication fidelity EMBO J 2004 23 1494 1505 15057282 Dutta S Li Y Johnson D Dzantiev L Richardson CC Crystal structures of 2-acetylaminofluorene and 2-aminofluorene in complex with T7 DNA polymerase reveal mechanisms of mutagenesis Proc Natl Acad Sci U S A 2004 101 16186 16191 15528277 Li Y Dutta S Doublie S Bdour HM Taylor JS Nucleotide insertion opposite a cis–syn thymine dimer by a replicative DNA polymerase from bacteriophage T7 Nat Struct Mol Biol 2004 11 784 790 15235589 Johnson SJ Beese LS Structures of mismatch replication errors observed in a DNA polymerase Cell 2004 116 803 816 15035983 Yang W Damage repair DNA polymerases Y Curr Opin Struct Biol 2003 13 23 30 12581656 Washington MT Prakash L Prakash S Yeast DNA polymerase eta utilizes an induced-fit mechanism of nucleotide incorporation Cell 2001 107 917 927 11779467 Fiala KA Suo Z Mechanism of DNA polymerization catalyzed by Sulfolobus solfataricus P2 DNA polymerase IV Biochemistry 2004 43 2116 2125 14967051 Zhou BL Pata JD Steitz TA Crystal structure of a DinB lesion bypass DNA polymerase catalytic fragment reveals a classic polymerase catalytic domain Mol Cell 2001 8 427 437 11545744 Garcia-Diaz M Bebenek K Krahn JM Kunkel TA Pedersen LC A closed conformation for the Pol lambda catalytic cycle Nat Struct Mol Biol 2005 12 97 98 15608652 Cadet J Douki T Gasparutto D Ravanat JL Oxidative damage to DNA: Formation, measurement and biochemical features Mutat Res 2003 531 5 23 14637244 Grisham MB Jourd'heuil D Wink DA Review article: Chronic inflammation and reactive oxygen and nitrogen metabolism—Implications in DNA damage and mutagenesis Aliment Pharmacol Ther 2000 14 Suppl 1 3 9 10807397 Shibutani S Takeshita M Grollman AP Insertion of specific bases during DNA synthesis past the oxidation-damaged base 8-oxodG Nature 1991 349 431 434 1992344 Einolf HJ Schnetz-Boutaud N Guengerich FP Steady-state and pre-steady-state kinetic analysis of 8-oxo-7,8-dihydroguanosine triphosphate incorporation and extension by replicative and repair DNA polymerases Biochemistry 1998 37 13300 13312 9748338 Einolf HJ Guengerich FP Fidelity of nucleotide insertion at 8-oxo-7,8-dihydroguanine by mammalian DNA polymerase delta. Steady-state and pre-steady-state kinetic analysis J Biol Chem 2001 276 3764 3771 11110788 Haracska L Yu SL Johnson RE Prakash L Prakash S Efficient and accurate replication in the presence of 7,8-dihydro-8-oxoguanine by DNA polymerase eta Nat Genet 2000 25 458 461 10932195 Krahn JM Beard WA Miller H Grollman AP Wilson SH Structure of DNA polymerase beta with the mutagenic DNA lesion 8-oxodeoxyguanine reveals structural insights into its coding potential Structure (Camb) 2003 11 121 127 12517346 Brieba LG Eichman BF Kokoska RJ Doublie S Kunkel TA Structural basis for the dual coding potential of 8-oxoguanosine by a high-fidelity DNA polymerase EMBO J 2004 23 3452 3461 15297882 Hsu GW Ober M Carell T Beese LS Error-prone replication of oxidatively damaged DNA by a high-fidelity DNA polymerase Nature 2004 431 217 221 15322558 Boudsocq F Iwai S Hanaoka F Woodgate R Sulfolobus solfataricus P2 DNA polymerase IV (Dpo4): An archaeal DinB-like DNA polymerase with lesion-bypass properties akin to eukaryotic pol eta Nucleic Acids Res 2001 29 4607 4616 11713310 Johnson SJ Taylor JS Beese LS Processive DNA synthesis observed in a polymerase crystal suggests a mechanism for the prevention of frameshift mutations Proc Natl Acad Sci U S A 2003 100 3895 3900 12649320 Furge LL Guengerich FP Analysis of nucleotide insertion and extension at 8-oxo-7,8-dihydroguanine by replicative T7 polymerase exo- and human immunodeficiency virus-1 reverse transcriptase using steady-state and pre-steady-state kinetics Biochemistry 1997 36 6475 6487 9174365 Efrati E Tocco G Eritja R Wilson SH Goodman MF “Action-at-a-distance” mutagenesis. 8-oxo-7, 8-dihydro-2′-deoxyguanosine causes base substitution errors at neighboring template sites when copied by DNA polymerase beta J Biol Chem 1999 274 15920 15926 10336498 Sawaya MR Prasad R Wilson SH Kraut J Pelletier H Crystal structures of human DNA polymerase beta complexed with gapped and nicked DNA: Evidence for an induced fit mechanism Biochemistry 1997 36 11205 11215 9287163 Kornyushyna O Berges AM Muller JG Burrows CJ In vitro nucleotide misinsertion opposite the oxidized guanosine lesions spiroiminodihydantoin and guanidinohydantoin and DNA synthesis past the lesions using Escherichia coli DNA polymerase I (Klenow fragment) Biochemistry 2002 41 15304 15314 12484769 Creighton S Bloom LB Goodman MF Gel fidelity assay measuring nucleotide misinsertion, exonucleolytic proofreading, and lesion bypass efficiencies Methods Enzymol 1995 262 232 256 8594351 Navaza J AMoRe: An Automated Package for Molecular Replacement Acta Crystallogr A 1994 50 157 163 Murshudov GN Vagin AA Dodson EJ Refinement of macromolecular structures by the maximum-likelihood method Acta Crystallogr D Biol Crystallogr 1997 53 240 255 15299926 Rudolph MG Wingren C Crowley MP Chien YH Wilson IA Combined pseudo-merohedral twinning, non-crystallographic symmetry and pseudo-translation in a monoclinic crystal form of the gammadelta T-cell ligand T10 Acta Crystallogr D Biol Crystallogr 2004 60 656 664 15039553 Hayward S Berendsen HJ Systematic analysis of domain motions in proteins from conformational change: New results on citrate synthase and T4 lysozyme Proteins 1998 30 144 154 9489922	16379496	PMC1325099	CC BY	2021-01-05 08:21:47	no	PLoS Biol. 2006 Jan 3; 4(1):e11	utf-8	PLoS Biol	2,006	10.1371/journal.pbio.0040011	oa_comm
==== Front BMC PsychiatryBMC Psychiatry1471-244XBioMed Central London 1471-244X-5-471635173410.1186/1471-244X-5-47Research ArticleLooking for pyromania: Characteristics of a consecutive sample of Finnish male criminals with histories of recidivist fire-setting between 1973 and 1993 Lindberg Nina [email protected] Matti M [email protected] Pekka [email protected] Matti [email protected] Department of Psychiatry, University of Helsinki, Finland2005 14 12 2005 5 47 47 10 8 2005 14 12 2005 Copyright © 2005 Lindberg et al; licensee BioMed Central Ltd.2005Lindberg 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 pyromania is a rare diagnosis with questionable validity, we aimed to describe a forensic psychiatric population of arson recidivists. Methods The medical records as well as the forensic psychiatric examination statements of 90 arson recidivists referred for pretrial psychiatric assessment in Helsinki University Hospital Department of Forensic Psychiatry between 1973 and 1993 were reviewed. Results The most important diagnostic categories of arson recidivists were personality disorders, psychosis and mental retardation, often with comorbid alcoholism. In all, 68% of arsonists were under alcohol intoxication during the index crime. Psychotic as well as mentally retarded persons with repeated fire-setting behaviour were mostly "pure arsonists"- persons guilty only of arsons during their criminal careers. Arson recidivists with personality disorder, in contrast, often exhibited various types of criminal behaviour and arson appeared to be only one expression of a wide range of criminal activity. Comorbid alcoholism was apparently a more rarely observed phenomenon among pure arsonists than in "nonpure arsonists". We found only three subjects fulfilling the present diagnostic criteria for pyromania. Conclusion Using the criteria of the DSM-IV-TR, pyromania must be regarded as an extremely rare phenomenon. Especially the question of substance intoxication as an exclusion criterion for pyromania should be reconsidered. ==== Body Background Arson is a major source of property damage, injury and death in many Western countries. The incidence of arson appears to have increased both in the United States and in Europe in recent years [1,2] and in Finland it has doubled between 1985 and 1995 [3]. Fire-setting recidivism rates have varied widely in the literature, from 4% to as much as 60%, depending on the study population [4]. Mentally disordered fire-setters have higher rates of recurrence of fire-setting than nonmentally disordered fire-setters and commit fewer common offences other than fire-setting [5]. In the study by Barnett et al. [6] arsonists who were only partly responsible and who committed no crimes other than arson showed the highest number of fire-setting incidents. Arson recidivists have rarely been thoroughly characterized, but their psychiatric disorders appear to be heterogeneous and include schizophrenia, bipolar disorder, substance abuse, personality disorders as well as mental retardation [7-9]. Pyromania – deliberate and purposeful fire- setting on more than one occasion- is a rare phenomenon that seldom explains repeated fire-setting behaviour [10,11], and the phenomenon as a valid diagnosis has even been questioned [12]. Taken together, further research is needed to examine the different subgroups of arsonists [4]. We aimed to characterize arson recidivists by describing some major psychiatric and demographic variables as well as criminal histories of a representative sample of Finnish male criminals with repeated fire-setting behaviour. We further aimed to characterize arsonists with "pyromanic" behavioural patterns by extracting the "pure arsonists" i.e. those with only arsons in their criminal histories from the study population and comparing them with "nonpure arsonists" i.e. those with also other types of offences in their previous criminal careers. We hypothesized that these two populations may differ in that the mental disorders of pure arsonists would somehow explain their repeating behaviour whereas arsons committed by nonpure arsonists may have more instrumental properties. Methods Subjects The hospital records and forensic psychiatric examination statements of arsonists who underwent a pretrial forensic psychiatric evaluation during a 21-year period (1973 – 1993) in Helsinki University Central hospital were reviewed. Ninety (22.4%) of the altogether 401 arsonists (all male over 16 years old) were recidivists i.e. had committed two or more separate arsons before the evaluation. Some of the participants were convicted in court for arson earlier during their criminal carrier, and others were in court process for the first time in their lives. Diagnostic categories Psychiatric classification according to the International Classification of Diseases- Eight Revision (ICD-8) [13] was used in clinical practice in Finland between 1969 and 1986 and it was replaced by the classification according to ICD-9[14] in January 1987. We pooled the principal diagnoses of 90 arson recidivists into 6 main diagnostic categories: mental retardation, psychosis, personality disorders, alcoholism, organic brain disorders and mood disorders. The diagnosis of mental retardation was based on an IQ of 70 or below [15,16]; 70% of the offenders in this category were proposed by the investigating forensic psychiatrist to be nonresponsible for the index crime while 30 % showed diminished responsibility. None of the mentally retarded persons guilty of repeated arsons were diagnosed as psychotic. Persons with the following ICD-8 diagnoses: schizophrenia paranoid type, schizophrenic residual state, schizophrenia catatonic type, schizophrenia NOS, psychosis NOS as well as the following ICD-9 diagnoses: unspecified schizophrenia, bipolar affective disorder, manic, were placed in psychosis category. All offenders in this category were proposed to be nonresponsible for their index arson. Subjects with the principal diagnosis of ICD-8 personality disorder (immature, antisocial, narsistic, explosive, NOS) or ICD-9 personality disorder (antisocial, borderline, dependent, mixed type) showed no Axis I diagnosis, except for possible comorbid abuse/dependence disorder, and were placed in the personality disorders category. The alcoholism category included persons with no Axis I or Axis II diagnoses other than substance dependence (ICD-8: episodic excessive drinking, alcoholic addiction, habitual excessive drinking; ICD-9: alcohol dependence syndrome). The organic brain disorder category included persons with severe head injury and personality change as well as those with dementia, while the mood disorders category included patients with nonpsychotic depressive disorders. Pure vs. nonpure arsonists The arson recidivists were then divided into two subgroups based on their previous criminal histories. Pure arsonists were offenders with only arsons in their previous criminal careers, while nonpure arsonists were those who had committed other crimes in addition to arsons (traffic, property and violent offences). Pyromania All 90 forensic psychiatric examination statements were carefully reviewed and pyromania was assessed using the diagnostic criteria of the Diagnostic and Statistical Manual of Mental disorders- Fourth Edition- Text Revision (DSM-IV-TR) [17]. These criteria included tension or affective arousal before the act, fascination with, interest in, curiosity about, or attraction to fire and its situational contexts, as well as pleasure, gratification, or relief when setting fires or when witnessing or participating in their aftermath [17]. Disorders as psychosis, mania, mental retardation, dementia and antisocial personality disorder were used as exclusion criteria according to the DSM-IV-TR. Statistics After description of the sample as a whole, the pure vs. nonpure arsonist subgroups were compared using the nonparametric Mann-Whitney independent samples test for the continuous numeric variables and the chi-square (χ2) test for the categorical variables. Results Characteristics of arson recidivists The mean age of the arson recidivists was 32.2 (SD 10.9) years. Four out of five of them had never been married. The mean IQ of the subjects was 93.0 (SD 19.7). 16 (17.8%) were mentally retarded (IQ under or 70), in 18 (20%) the principal diagnosis was psychosis, in 47 (52%) personality disorder (immature [n = 11], narsistic [n = 1], explosive [n = 2], antisocial [n = 20], borderline [n = 11], dependent [n = 1], NUD [n = 1]) and in 9 (10%) other diagnoses. A total of 55 subjects (61%) suffered from comorbid alcohol abuse/dependency and 61 persons (68%) committed the index arson under acute alcohol intoxication (mental retardation: 5 /16 persons, psychosis: 11 /18, personality disorder: 38/47, organic brain disorders: 0/2, mood disorders 3/3, alcoholism 4/4 persons). Pyromania After excluding persons with psychosis, mental retardation, organic brain syndrome and antisocial personality disorder (n = 56) 12 of the remaining 34 arson recidivists fulfilled the DSM-IV-TR inclusion criteria for pyromania. However, 9 of these 12 persons were under acute alcohol intoxication during the index arson. Pure vs. nonpure arsonists In all, 43 (48%) of the arson recidivists had only arsons in their previous criminal histories at the time of the evaluation; 15 out of 16 mentally retarded (χ2 = 16.483, df 1, p = 0.000) and 15 out of 18 psychotic persons (χ2 = 11.400, df 1, p = 0.001) belonged to this group of pure arsonists. In contrast, only 12 out of 47 subjects with personality disorder (immature [n = 5], borderline [n = 5], dependent [n = 1], NUD [n = 1]) (χ2 = 19.511, df 1, p = 0.000) and 18 subjects out of 55 (χ2 = 12.840, df1, p = 0.000) with comorbid alcoholism belonged to this group. The median IQ of the pure arsonists was lower (84.5 [Q1-3 67.25–105.5] vs. 101 [Q1-3 90–110], Z = -3.65, p = 0.000) than of nonpure arsonists. There were no differences in the median ages or marital status between these two subgroups. Discussion In this consecutive sample of arsonists, 22.4% were recidivists, which is similar to the results of some fire-setting studies undertaken among the forensic population [7,11]. The most important diagnostic categories of arson recidivists were personality disorders, psychosis and mental retardation, often with comorbid alcoholism. Nearly 70% of the arsonists were under acute alcohol intoxication during the index crime; these observations were similar to those reported in previous fire-setting studies [7-9]. The mentally retarded arson recidivists 18% of recidivists were mentally retarded persons and they were almost all pure arsonists. Originally, pathological fire-setting was considered as a phenomenon perpetrated by physically disabled or mentally retarded pubescent girls who had abnormal psychosexual development [18]. Later, due to methodological difficulties, there have been conflicting opinions of whether people with mental retardation are over- or underrepresented in the arsonist population [19]. In a recent prospective follow-up of 61 offenders with intellectual disability, both sex offence and arson were overrepresented offence types and as many as 21.3% of the total population had a history of fire-setting [20]. In contrast to our results, most of the above-mentioned 61 persons with intellectual disability had committed also other offences. A possible explanation for this discrepancy might be that the above mentioned study group included single-setters with an IQ below 80, whereas in our study all the subjects were repeaters and the IQ cutoff point was 70 or below, which generally indicates a significantly impaired intellectual functioning. As far as we know, no previous studies have focused particularly on arson recidivists in this population. Murphy and Clare [21] reported that the three most frequently identified antecedent emotions/events before fire-setting in mentally retarded patients were, in rank order, angry feelings, not being listened to/ attended to, and feeling sad/ depressed. Stereotypical behaviour is frequently found in deeply mentally retarded patients [22], which most of the persons in the present study were not. However, overrepresentation of mentally retarded persons as pure arsonists ma be explained by the idea that they represent persons who express their feelings of anger, frustration and sadness through specific, repetitious action models (in these cases arson), but who do not otherwise express interest in criminal activities. The psychotic arson recidivists 20 % of recidivists were psychotic persons and, like mentally disordered patients, they were mostly pure arsonists. A higher than expected prevalence of psychotic disorders among fire-setters was already reported over 60 years ago [23]. Motives of patients with schizophrenia have been reported to be often similar to those observed in non-psychotic persons, with hatred and revenge as the common underlying factors in arson [24]. However, delusions have also been regarded as motives or reasons [25]. In a very large sample of fire-setters as many as 40% of paranoid schizophrenics repeated fire-setting [25]. A large proportion of recidivists set fires to property that is unrelated to them in any way and they also seem to set fires on a very sporadic basis, usually within the context of situational crises [10]. In our sample, most psychotic persons were pure arsonists with no other criminal activities, suggesting, that among psychotic fire-setters there may be different subgroups with various underlying psychological mechanisms leading to fire-setting behaviour. In single-setters the motives may be more instrumental, whereas in recidivists fire-setting is probably a function of flagrant thought disturbance, precipitated by command hallucinations and delusions. The arson recidivists with personality disorders Antisocial personality disorder was the most common personality disorder in the present sample (22 % of recidivists) and, in fact, some of the best predictors of recidivist fire setting are impulsive characteristics [26]. Deliberate fire-setting is one of the diagnostic criteria in conduct disorder, the childhood and adolescent predecessor of adult antisocial personality disorder [17], and self-reported fire- setting is strongly associated with extreme antisocial behaviour in young community adolescents [27]. Recently, adolescent fire-setters were described as aggressive but on the other hand shy and rejected by their peers [28]. All recidivists with antisocial personality disorder were intoxicated during the index crime. In a prospective follow-up study, low central nervous system 5-hydroxyindole acetic acid (CNS 5-HIAA) and 3-methoxy-4-hydrophenylglycol (MHPG) concentrations, typically associated with impulsive antisocial personality and Cloninger type II alcoholism, were associated with both violent recidivist criminal offences and fire-setting [29]. In the present sample all the antisocial fire-setters were nonpure arsonists, and the arson were often aimed at revenge. The underlying motives for these impulsive acts were typically hatred and rage, which increased in severity along with acute substance intoxication. Offenders with severe personality disorder typically begin their criminal careers early and their criminal records are extended. As criminals, their acts appear to be heterogeneous and fire-setting reflects their criminal and unempathic natures [12]. In a study by Repo et al. [8] among arsonists, overall lifetime criminal recidivism was primarily associated with antisocial personality and alcohol dependence. The arson recidivists with pyromania Pyromania is regarded as an impulse-control disorder, but controversy continues over whether the condition should be classified under the impulse-control disorders or, indeed, whether it comprises a separate entity at all. Pyromania was already mentioned in DSM-I, omitted from the DSM-II, but then acknowledged again in the DSM-III [12]. In the DSM-IV-TR a wide range of exclusion criteria such as psychotic states, dementia, mental retardation, monetary gain and political ideologies are presented. The exclusion criteria cover also other criminal acts, rage, revenge and acute intoxication. The phenomenon should also not be observed as pyromania if it is better explained as conduct disorder, antisocial personality disorder or manic state. In our sample, after exclusion of persons with psychosis, mental retardation, organic brain syndrome and antisocial personality disorder, 12 of the arson recidivists fulfilled the inclusion criteria of pyromania. They expressed tension or affective arousal before the act, attraction to and interest in fire, as well as pleasure and release afterwards. However, nearly all these persons were intoxicated during the fire-setting, and they typically mentioned that the tension as well as the affective arousal increased when they were consuming alcohol. Since substance intoxication is one of the exclusion criteria, only three persons with true pyromania were found in the present sample. They all were pure arsonists and, interestingly, all these three men worked as volunteer firefighters and also expressed special interest in fire in other ways as arsonists. In conclusion, using the present strict exclusion criteria for pyromania, the disorder must be regarded as an extremely rare phenomenon as also shown in previous studies [10,11]. The question of substance intoxication as an exclusion criterion should especially be reconsidered. Methodological issues In Finland, an average of about 500–600 arson attempts are made and 100 offenders convicted of arson each year. The proportion of individuals who undergo a forensic psychiatric evaluation compared with all arsonists suspected by the police was estimated to be only 10% [3]. So, the present sample, as well as other previous pretrial samples reported in the literature, is not representative of arsonists in general. The diagnoses were based on several clinical interviews, psychological tests and observation during the two month examination periods. Finnish forensic psychiatric examination statement traditionally includes a paragraph involving a review of the person's previous official criminal history. This information instead of authentic criminal records was used in dividing arsonists into different subpopulations. The strength of the study is that we were able to collect a consecutive sample of male fire-setting recidivists over a period of more than two decades. However, further research is still needed to clarify the important question of fire-setting recidivism. Conclusion The most important diagnostic categories of arson recidivists were personality disorders, psychosis and mental retardation, often with comorbid alcoholism. Psychotic as well as mentally retarded persons with repeated fire-setting behaviour were mostly pure arsonists- persons guilty only of arsons during their criminal careers. Repeating arsonists with personality disorder, in contrast, often exhibited various types of criminal behaviour and arson appeared to be only one expression of a wide range of criminal activity. Using the criteria of the DSM-IV-TR, pyromania is an extremely rare phenomenon. Substance intoxication as an exclusion criterion for pyromania should be reconsidered. Competing interests The author(s) declare that they have no competing interest. Authors' contributions This manuscript was prepared by a multidisciplinary team consisting of: NL, generated the idea for this study, studied the forensic psychiatric evaluation statements, prepared the manuscript together with the team MMH, participated in designing the study and in data collection, carried out statistical analyses and prepared the manuscript together with the team PT, as a neurologists had a substantial contribution in theoretical background concerning mental retardation and participated actively in preparation of this manuscript MV, supervised and participated with great impact in all stages of preparation of this manuscript Pre-publication history The pre-publication history for this paper can be accessed here: ==== Refs Prins H Fire raising: its motivation and management 1994 London, New York: Routledge Soothill K Ackerley E Francis B The criminal careers of arsonists Med Sci Law 2004 44 27 40 14984212 Räsänen P Hakko H Väisänen E Arson trend increasing – a real challenge to psychiatry J Forens Sci 1995 40 976 979 Brett A "Kindling theory" in arson: how dangerous are firesetters? Aust N Z J Psychiatry 2004 38 419 425 15209833 10.1111/j.1440-1614.2004.01378.x Barnett W Recidivism and concomitant criminality in pathological fire-setters J Forensic Sci 1997 42 879 883 9304835 Barnett W Richter P Renneberg B Repeated arson: Data from criminal records Forensic Sci Int 1999 101 49 54 10376337 10.1016/S0379-0738(99)00012-2 Rix KJB A psychiatric study of adult arsonists Med Sci Law 1994 34 21 34 8159068 Repo E Virkkunen M Rawlings R Linnoila M Criminal and psychiatric histories of Finnish arsonists Acta Psychiatr Scand 1997 95 318 323 9150826 Ritchie EL Huff TG Psychiatric aspects of arsonists J Forensic Sci 1999 44 733 740 10432607 Koson DF Dvoskin J Arson: a diagnostic study Bull Am Acad Psychiatry Law 1982 10 39 49 7139133 Leong GB A psychiatric study of persons charged with arson J Forensic Sci 1992 37 1319 1326 1402754 Greenberg HR Impulse-Control Disorders Not Elsewhere Classified Kaplan & Sadock's Comprehensive Textbook of Psychiatry 2005 8 Lippincott Williams & Wilkins: Philadelphia USA 2035 2054 Report of the International Conference for the Eighth Revision of The International Classification of Diseases 1965 Geneva, World Health Organization WHO Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death, Ninth Revision 1977 1 Geneva: World Health Organization Matarazzo JD Wechsler's Measurement and Appreised of Adult Intelligence, 5th and Enlarged Edition 1972 Baltimore: The Williams & Wilkins Company Wechsler D Wechsler Adult Intelligence Scale-Revised Manual 1981 Cleveland: Psychological Corporation American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders, 4th edition, Text Revision Washington DC: American Psychiatric Press Inciardi JA The adult firesetter- a typology Criminology 1970 8 145 155 Taylor JL Thorne I Robertson A Avery G Evaluation of a group intervention for convicted arsonists with mild and borderline intellectual disabilities Crim Behav Ment Health 2002 12 282 293 12897901 Barron P Hassiotis A Banes J Offenders with intellectual disability: a prospective comparative study J Intellect Disabil Res 2004 48 69 76 14675234 10.1111/j.1365-2788.2004.00581.x Murphy GH Clare ICH Analysis of motivation in people with mild learning disabilities (mental handicap) who set fires Psychology Crime and Law 1996 2 153 164 Sturmey P Bouras N Classification: concepts, progress and future Psychiatric and Behavioural Disorders in Developmental Disabilities and Mental Retardation 1999 Cambridge University Press 3 17 Gerle B Mordbrännare (monograph) 1943 Lund: A-B Universitetsbokhandeln Virkkunen M On arson committed by schizophrenics Acta Psychiatr Scand 1974 50 152 160 4853431 Lewis NDC Yarnell H Pathological fire-setting (pyromania) Nervous and mental disease monograph no 82 1951 New York: Coolidge Foundation Foust LL The legal significance of clinical formulations of firesetting behavior Int J Law Psychiatry 1979 2 371 387 511413 10.1016/0160-2527(79)90013-X Martin G Bergen HA Richardson AS Roeger L Allison S Correlates of firesetting in a community sample of young adolescents Aust N Z J Psychiatry 2004 38 148 154 14961933 10.1111/j.1440-1614.2004.01318.x Chen YH Arria AM Anthony JC Firesetting in adolescence and being aggressive, shy, and rejected by peers: new epidemiologic evidence from a national sample survey J Am Acad Psychiatry Law 2003 31 44 52 12817842 Virkkunen M Eggert M Rawlings R Linnoila M A prospective follow-up study of alcoholic violent offenders and fire setters Arch Gen Psychiatry 1996 53 523 529 8639035	16351734	PMC1325224	CC BY	2021-01-04 16:33:02	no	BMC Psychiatry. 2005 Dec 14; 5:47	utf-8	BMC Psychiatry	2,005	10.1186/1471-244X-5-47	oa_comm
==== Front BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-1271633225310.1186/1471-2458-5-127Research ArticleIncidence of cancer in the area around Amsterdam Airport Schiphol in 1988–2003: a population-based ecological study Visser Otto [email protected] Wijnen Joop H [email protected] Leeuwen Flora E [email protected] Comprehensive Cancer Centre Amsterdam, POBox 9236, 1006 AE Amsterdam, The Netherlands2 Municipal Health Service Amsterdam, Environmental Medicine, POBox 20244, 1000 HE Amsterdam, The Netherlands3 Netherlands Cancer Institute, Dept of Epidemiology, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands2005 6 12 2005 5 127 127 8 6 2005 6 12 2005 Copyright © 2005 Visser et al; licensee BioMed Central Ltd.2005Visser 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 Amsterdam Airport Schiphol is a major source of complaints about aircraft noise, safety risks and concerns about long term adverse health effects, including cancer. We investigated whether residents of the area around Schiphol are at higher risk of developing cancer than the general Dutch population. Methods In a population-based study using the regional cancer registry, we estimated the cancer incidence during 1988–2003 in residents of the area surrounding Schiphol. We defined a study area based on aircraft noise contours and 4-digit postal code areas, since historical data on ambient air pollution were not available and recent emission data did not differ from the background urban air quality. Results In residents of the study area 13 207 cancer cases were diagnosed, which was close to the expected number, using national incidence rates as a reference (standardized incidence ratio [SIR] 1.02). We found a statistically significantly increased incidence of hematological malignancies (SIR 1.12, 95% confidence interval [CI]: 1.05, 1.19), mainly due to high rates for non-Hodgkin lymphoma (SIR 1.22, 95% CI: 1.12, 1.33) and acute lymphoblastic leukemia (SIR 1.34, 95% CI: 0.95, 1.83). The incidence of cancer of the respiratory system was statistically significantly decreased (SIR 0.94, 95% CI: 0.90, 0.99), due to the low rate in males (SIR 0.89). In the core zone of the study area, cancer incidence was slightly higher than in the remaining ring zone (rate ratio of the core zone compared to the ring zone 1.05, 95% CI 1.01, 1.10). This was caused by the higher incidence of cancer of the respiratory system, prostate and the female genital organs in the core zone in comparison to the ring zone. Conclusion The overall cancer incidence in the Schiphol area was similar to the national incidence. The moderately increased risk of hematological malignancies could not be explained by higher levels of ambient air pollution in the Schiphol area. This observation warrants further research, for example in a study with focus on substances in urban ambient air pollution, as similar findings were observed in Greater Amsterdam. ==== Body Background Amsterdam Airport Schiphol is one of the main airports of Europe. The airport is a major source of complaints about aircraft noise, noise related adverse health effects and – especially since the crash of an airplane in a suburb of Amsterdam on October 4th 1992 – about safety risks. A longstanding subject of concern of the surrounding population is the exposure to aviation fuels and their combustion products and an alleged increase of cancer risk. Particularly in warm summers the smell of aviation fuels can be distinguished outside the airport grounds. Aircraft emissions vary with the engine type, the engine load and the kind of fuel. Combustion of aviation fuels results in CO2, CO, Ce, NOx, particles, and a great number of other organic compounds, among which a number of carcinogens [1]. Among the emitted polycyclic aromatic hydrocarbons no compound characteristic for aircraft engines has been detected so far. A committee of the Health Council of The Netherlands recently reviewed the data on the health impact of large airports [2]. It was concluded that, generally, integrated health assessments are not available. In the last 30 years, several adverse health effects in relation to exposure to aircraft noise have been the subject of study, such as the use of tranquillizers, the prevalence of bronchitis and cardiovascular disease as well as child stress responses and cognition [3-6]. However, little information is available in the international literature on cancer risk in relation to airports. In the late 1980s, mortality due to cancer in the community of Haarlemmermeer, which hosts Schiphol, was investigated by the Municipal Health Service of Amsterdam on request of the general practitioners in the area [7]. The total cancer mortality and the lung cancer mortality in Haarlemmermeer during 1981–86 did not differ statistically significantly from the cancer mortality in the two standard populations that were used. The mortality due to non-Hodgkin lymphoma (NHL) was statistically significantly increased, but conclusions as to the cause of the excess mortality were not possible. In the 1990s, we carried out a first study on the incidence of cancer in the vicinity of Schiphol, as part of the health surveillance of the resident population of the Schiphol area [8]. During 1988–1993, the incidence of cancer in the area around Schiphol was close to the national average. The differences in incidence of certain types of cancer in comparison to the national average, as well as those between two study areas characterized by different levels of increased aircraft noise, were considered to be most likely due to differences in life style, such as smoking. In order to investigate whether cancer risk of the resident population of the Schiphol area (in comparison to the national average) changed since 1988–1993, we continued monitoring cancer incidence and we report here on the second, much larger population-based study of the cancer incidence around Schiphol. Methods Definition of the study population and the study area When we designed our first study, relevant exposure data on the ambient air quality around Schiphol airport were lacking and we could not define a study population exposed to increased ambient levels of aircraft emissions. The airport itself has no permanent residents and the most heavily exposed population – the airport personnel and the travelers – cannot be defined geographically. Therefore, we defined our study population as the population most heavily exposed to increased levels of aircraft noise. Since 1994, the ambient air quality outside Schiphol has been monitored and no differences with the background urban air quality have been reported for the compounds that were measured [9]. Table 1 summarizes the results of the three monitoring locations in the Schiphol area. However, it is possible that exposure to aircraft emissions has been greater in the past when aircraft engines used to be technologically and ecologically less advanced. Also, we cannot exclude that certain carcinogenic compounds specific to aviation combustion have not been monitored. Since most cancers have a long induction period and the noise contours are thought to reflect best the historical exposure of the surrounding population to aircraft emissions, we continued to use the levels of aircraft noise to define our study area. The aircraft noise levels of 1991 were available as so-called Kosten-units (Ku) [10]. We used the 35 Ku contour and extended the area with about 2 km outside the 35 Ku contour (figure 1). This total area (surrounded by the solid black line in figure 1) was redefined as 4-digit postal code areas (postal code areas surrounded by grey lines in figure 1). The four airstrips of the airport are easily recognized by noise levels over 50 Ku. We also defined a core zone for the 4-digit postal code areas within the 45 Ku contour (the area bordered by the blue line in figure 1), although we do not have empirical data showing that this zone corresponds to a zone with increased levels of ambient air pollution. The remaining study area surrounding the core zone we designated as 'ring zone'. The location of the three air quality monitoring stations (Badhoevedorp, Hoofddorp, Oude Meer) is also indicated in figure 1. The total study area with a population of 177 000 on 31 December 2003 comprised (parts of) five municipalities (table 2). Table 2 also includes figures on per capita income as approximation for socio-economic status. Figure 1 Noise exposure (in Kosten-units) in the Schiphol area in 1991. The area surrounded by the blue line indicates the core zone, the black line includes the total study area. The location of the three air quality monitoring stations are indicated by asterisks (). Table 1 Summary of the results in μg/m3 (except benzo(a)pyrene: ng/m3) of the air quality monitoring system of the Schiphol area in 2002 Pollutant Unit Limit Location of monitoring station Badhoevedorp Oude Meer Hoofddorp NO2 year average 40a 38 38 31 maximum 200b 163 544(1x > 200) 124 CO P98 (8 hours) 9000 112 100 88 P99,9 40000 134 165 160 O3 maximum 240c 185 174 266 (2x > 240) PM10 average (year) 40d,e 26 24 28 maximum (24 hours) 50d,f 81 (13x > 50) 81 (8x > 50) 132 (22x > 50) Benzene year average 10 1.4 1.1 0.7 Black smoke P98 (24 hours) 90g 42 48 34 Benzo(a)pyrene year average 1 0.14 a as of 1-1-2010 b exceeding of the limit no more than 18 times per annum c exceeding of the limit not more than 48 hours d including factor 1.3 e as of 1-1-2005 f exceeding of the limit no more than 35 times per annum g the limit expired in July 2001 PM10 = particulate matter <10 μm Table 2 Some characteristics of the Schiphol study area Zone and municipality Postal codes Inhabitants Per capita income (1998) 1-1-1988 31-12-2003 Core zone 30 590 31 850 € 10 900† Haarlemmermeer 1161 8300 7820 € 10 700 1175, 1435-8, 2143 6880 6015 € 11 000 2132 7815 10 965 € 11 000 2153 3520 3310 € 10 500 Amstelveen 1182 4075 3740 € 11 500 Ring zone 131 210 144870 € 11 800† Haarlemmermeer 1171 10 750 11 770 € 13 200 2131 10 205 11 030 € 11 400 2151-2 11 925 22 260 € 11 100 2154-8, 2165 5585 5940 € 10 300 Amstelveen 1181, 1183 31 145 30 385 € 12 400 Amsterdam 1067, 1081-3 34 960 35 080 € 14 500 Aalsmeer 1431-3 21 740 22 870 € 11 300 Haarlemmerliede & Spaarnwoude 1165, 2064-5 4900 5535 € 10 800 Total study area 161 800 176 720 € 11 700† * the national average in 1998 was € 10 000 † weighed average, rounded to € 100 Population data Annual population data covering the period 1995–2003 according to 4-digit postal code, 5-year age groups and sex, were available for all municipalities from Statistics Netherlands. For the period 1988–1994 we used data from the municipal administrations. Cancer registry data The Amsterdam Cancer Registry (ACR) is a regional, population-based cancer registry with complete regional coverage since 1988. The ACR is part of the nation-wide Netherlands Cancer Registry (NCR) [11]. Completeness of the NCR is estimated to be over 95%. The information is extracted from the medical records by registration clerks. Apart from demographic data, data are collected on tumor site, morphological classification (according to the International Classification of Diseases for Oncology [ICD-O], versions 1 and 2), stage of the tumor and treatment of the patients. The third version of the ICD-O was introduced in the NCR for cases diagnosed as of January 2001. Cases diagnosed in a hospital outside the ACR region but with residence in the ACR region are routinely obtained from the national registry and included in our regional registry. Consequently, these cases could be included in the study. We selected from the registry all cancer cases in the period 1988–2003 with residence in the area around Schiphol airport at the date of diagnosis. We stratified the cases according to type of cancer (or group of cancers), area of residence (core zone or the ring zone), 5-year age group and sex. Statistical methods In our analysis, the incidence of cancer in the national population of the Netherlands served as the reference entity. The expected numbers of cancer (E) for the Schiphol area were calculated for three periods (1988–1993, 1994–1998 and 1999–2003), based on the population data of the Schiphol area (according to 5-year age category and sex) and the 5-year age category and sex-specific cancer incidence rates from the NCR. For the period 1988–1993 we used the average incidence rates of the NCR covering the period 1989–1993 [12], because data for 1988 were not available from the NCR. For the periods 1994–1998 and 1999–2003 we used NCR-data covering 1994–1998 and 1999–2003, respectively [13]. The expected numbers were compared with the observed numbers (O) and standardized incidence ratios (SIRs) were calculated as the ratio between the observed and expected numbers. Exact 95%-confidence intervals (CI) based on the Poisson distribution of O were calculated using STATA 7.0 (STATA Corporation. College Station, Texas, USA). Rate ratios (RR) for the core zone were calculated by dividing the standardized incidence ratio of the core zone by the rate of the ring zone. Ninety five percent CIs of RRs were calculated assuming a log-normal distribution [14]. Results In 1988–2003, a total of 13 207 cancers (6 739 in males, 6 468 in females) were diagnosed among residents of the Schiphol area (table 3), which included 2 352 cases among residents of the core zone. Table 3 Observed (O) and expected (E) number of cancers in subjects with residence in the Schiphol area according to site, gender and period of diagnosis, 1988–2003 cancer site (ICD-10 code) and gender Total period (1988–2003) 1988–1993 1994–1998 1999–2003 O E SIR 95% CI O E SIR 95% CI O E SIR 95% CI O E SIR 95% CI All malignancies (C00–C96) 13207 13007.9 1.02 1.00, 1.03 4624 4538.6 1.02 0.99, 1.05 4220 4125.1 1.02 0.99, 1.05 4363 4344.3 1.00 0.97, 1.03 adult males 6697 6713.7 1.00 0.97, 1.02 2402 2363.5 1.02 0.98, 1.06 2145 2150.3 1.00 0.96, 1.04 2150 2199.9 0.98 0.94, 1.02 adult females 6436 6235.3 1.03* 1.01, 1.06 2190 2154.3 1.02 0.97, 1.06 2057 1956.5 1.05* 1.01, 1.10 2189 2124.5 1.03 0.99, 1.07 children <15 74 58.9 1.26 0.99, 1.58 32 20.7 1.54* 1.06, 2.18 18 18.3 0.98 0.58, 1.55 24 19.9 1.21 0.77, 1.79 Head & neck (C00–C14) 282 272.9 1.03 0.92, 1.16 93 93.7 0.99 0.80, 1.22 91 85.9 1.06 0.85, 1.30 98 93.3 1.05 0.85, 1.28 males 162 176.9 0.92 0.78, 1.07 53 62.8 0.84 0.63, 1.10 56 55.6 1.01 0.76, 1.31 53 58.4 0.91 0.68, 1.19 females 120 96.0 1.25* 1.04, 1.49 40 30.9 1.29 0.92, 1.76 35 30.3 1.16 0.80, 1.61 45 34.8 1.29 0.94, 1.73 Gastrointestinal tract (C15–C26) 2889 2936.0 0.98 0.95, 1.02 1034 1049.3 0.99 0.93, 1.05 899 920.4 0.98 0.91, 1.04 956 966.3 0.99 0.93, 1.05 males 1494 1538.0 0.97 0.92, 1.02 517 541.6 0.95 0.87, 1.04 470 482.3 0.97 0.89, 1.07 507 514.1 0.99 0.90, 1.08 females 1395 1398.1 1.00 0.95, 1.05 517 507.7 1.02 0.93, 1.11 429 438.1 0.98 0.89, 1.08 449 452.3 0.99 0.90, 1.09 Respiratory system (C30–C34) 1862 1975.1 0.94* 0.90, 0.99 749 752.6 1.00 0.93, 1.07 542 627.1 0.86† 0.79, 0.94 571 595.4 0.96 0.88, 1.04 males 1378 1548.0 0.89† 0.84, 0.94 604 626.3 0.96 0.89, 1.01 386 492.7 0.78† 0.71, 0.87 388 429.0 0.90 0.82, 1.00 females 484 427.1 1.13* 1.03, 1.24 145 126.3 1.15 0.97, 1.35 156 134.5 1.16 0.98, 1.36 183 166.4 1.10 0.95, 1.27 Breast (C50) 2087 1983.6 1.05 0.99, 1.11 679 676.4 1.00 0.93, 1.08 678 620.3 1.09* 1.01, 1.18 730 686.9 1.06 0.99, 1.14 Female genital organs (C51–C58) 710 730.7 0.97 0.90, 1.05 252 274.9 0.92 0.81, 1.04 237 232.9 1.02 0.89, 1.16 221 222.9 0.99 0.87, 1.13 Prostate (C61) 1291 1230.9 1.05 0.99, 1.11 382 364.6 1.05 0.95, 1.16 470 421.2 1.12* 1.02, 1.22 439 445.2 0.99 0.90, 1.08 Bladder & other urinary tract (C65–C68) 543 517.2 1.05 0.96, 1.14 211 183.9 1.15 1.00, 1.31 172 163.7 1.05 0.90, 1.22 160 169.7 0.94 0.80, 1.10 males 425 390.0 1.09 0.99, 1.20 173 139.7 1.24† 1.06, 1.44 129 123.6 1.04 0.87, 1.24 123 126.7 0.97 0.81, 1.16 females 118 127.3 0.93 0.77, 1.11 38 44.2 0.86 0.54, 1.29 43 40.1 1.07 0.78, 1.44 37 43.0 0.86 0.61, 1.19 Hematological malignancies (C81–C96) 1044 935.2 1.12† 1.05, 1.19 367 328.3 1.12* 1.01, 1.24 334 291.2 1.15* 1.03, 1.28 343 315.8 1.09 0.97, 1.21 males 598 507.6 1.18† 1.09, 1.28 210 177.3 1.18* 1.03, 1.36 184 157.8 1.17 1.00, 1.35 204 172.6 1.18* 1.03, 1.36 females 446 427.6 1.04 0.95, 1.14 157 151.0 1.04 0.88, 1.22 150 133.4 1.12 0.95, 1.32 139 143.2 0.97 0.82, 1.15 Hodgkin lymphoma 48 61.2 0.78 0.58, 1.04 19 23.6 0.81 0.48, 1.26 12 17.7 0.68 0.35, 1.18 17 19.9 0.85 0.50, 1.37 non-Hodgkin lymphoma 516 423.5 1.22† 1.12, 1.33 176 150.0 1.17* 1.01, 1.36 181 133.5 1.36† 1.17, 1.57 159 140.0 1.14 0.97, 1.33 plasma cell tumors 169 156.1 1.08 0.93, 1.26 59 55.6 1.06 0.81, 1.37 56 51.3 1.09 0.82, 1.42 54 49.3 1.10 0.82, 1.43 acute lymphoblastic leukemia 39 29.1 1.34 0.95, 1.83 17 10.2 1.67 0.97, 2.67 8 8.8 0.91 0.39, 1.79 14 10.2 1.38 0.75, 2.30 chronic lymphocytic leukemia 92 101.2 0.91 0.73, 1.11 33 35.0 0.94 0.65, 1.32 29 35.2 0.82 0.55, 1.18 30 31.0 0.97 0.65, 1.38 acute myeloid leukemia 109 97.8 1.11 0.92, 1.34 40 33.9 1.18 0.84, 1.61 33 31.4 1.05 0.72, 1.48 36 32.4 1.11 0.78, 1.54 other 71 66.2 1.07 0.84, 1.35 23 20.1 1.15 0.73, 1.72 15 13.3 1.13 0.63, 1.86 33 32.8 1.00 0.69, 1.41 Other sites 2499 2426.3 1.03 0.99, 1.07 857 815.0 1.05 0.98, 1.12 797 762.4 1.05 0.97, 1.12 845 848.9 1.00 0.93, 1.06 males 1391 1355.5 1.03 0.97, 1.08 485 462.8 1.05 0.96, 1.15 462 427.2 1.08 0.99, 1.18 444 465.5 0.95 0.87, 1.05 females 1108 1070.8 1.03 0.97, 1.10 372 352.2 1.06 0.95, 1.17 335 335.2 1.00 0.90, 1.11 401 383.4 1.05 0.95, 1.15 * p < 0.05; † p < 0.01; CI = confidence interval; E = expected number; O = observed number; SIR = standardized incidence ratio Table 3 shows, that the total number of observed cancers was close to the expected number (SIR 1.02, 95% CI: 1.00, 1.03), in males (SIR 1.00, 95% CI: 0.97, 1.02) as well as in females (SIR 1.03, 95% CI: 1.01, 1.06). The observed number of cancers of the respiratory system (predominantly lung cancer) in females was increased (SIR 1.13, 95% CI: 1.03, 1.24), but the number in both sexes combined was statistically significantly decreased (SIR 0.94, 95% CI: 0.90, 0.99). This was caused by a relatively low incidence in males (SIR 0.89, 95% CI: 0.84, 0.94). A similar pattern was observed for cancer of head and neck (SIR females 1.25, 95% CI 1.04, 1.49; SIR males 0.92, 95% CI 0.78–1.07). The incidence was statistically significantly increased for hematological malignancies (SIR 1.12, 95% CI: 1.05, 1.19, 1044 cases). The raised risk was most prominent in males (SIR males 1.18, 95% CI: 1.09, 1.28, SIR females 1.04, 95% CI: 0.95, 1.14). A statistically significantly increased incidence was observed for NHL (SIR 1.22, 95% CI: 1.12, 1.33, 516 cases), while the confidence interval for acute lymphoblastic leukemia (ALL; SIR 1.34, 95% CI: 0.95, 1.83, 39 cases) included unity. A relatively low rate was observed for Hodgkin lymphoma (SIR 0.78, 95% CI: 0.58, 1.04). Classification of lymphoid malignancies according to the WHO-classification, revealed relatively high rates for lymphoplasmocytic lymphoma (SIR 1.5, 95% CI: 1.1, 2.9), follicular lymphoma (SIR 1.5, 95% CI: 1.2, 1.8), diffuse large B-cell lymphoma (SIR 1.6, 95% CI: 1.4, 1.9) and T-cell lymphoma (SIR 1.4, 95% CI: 1.0, 1.8). The rates for plasma cell tumors (SIR 1.1, 95% CI 0.9, 1.3), small lymphocytic lymphoma/chronic lymphocytic leukemia (SIR 0.8, 95% CI: 0.6, 1.0) and other & unspecified lymphoma/leukemia (SIR 1.0, 95% CI: 0.8, 1.2) were not increased. Cancer was diagnosed in 74 children up to 15 years of age, which was relatively high (SIR 1.26, 95% CI 0.99, 1.58), due to the higher than expected number of children with ALL (23 cases, SIR 1.59, 95% CI 1.01, 2.39). For most cancer sites, the SIRs for the periods 1988–1993, 1994–1998 and 1999–2003 were quite similar. The increased risk of hematological malignancies was consistently observed in the three time periods. An increased number of breast cancer cases was observed in the 1994–1998 period (SIR 1.09, 95% CI 1.01, 1.18). An increased number of cancer of the bladder and other urinary organs in males was only observed in 1988–1993 (SIR 1.24, 95% CI 1.06, 1.44). Cancer incidence in the core zone Table 4 shows that cancer incidence in the core zone was slightly increased in comparison to the national incidence (SIR 1.06, 95% CI 1.02, 1.10) as well as in comparison to the ring zone (RR 1.05, 95% CI 1.01, 1.10), mostly because of an increased incidence in males (SIR 1.07, 95% CI 1.01, 1.13; RR 1.09, 95% CI 1.03, 1.16). Statistically significantly increased numbers in the core zone in comparison to the ring zone were observed for cancer of the respiratory system (RR 1.27, 95% CI 1.12, 1.45) and prostate (RR 1.17, 95% CI 1.02, 1.34) in males and for cancer of the genital organs in females (RR 1.24, 95% CI 1.04, 1.50), based on increased RRs for each specific site (cervix 1.20, corpus 1.04, ovary 1.55, vulva & other 1.17). In comparison to the national incidence only cervical cancer and ovarian cancer were increased (SIR cervix 1.29, corpus 1.00, ovary 1.32, vulva & other 0.98). In the core zone, the incidence rate of bladder cancer in males (SIR 1.26, 95% CI 1.01, 1.56; RR 1.20, 95% CI 0.94, 1.52) was also relativity high. The incidence of hematological malignancies was higher in the core zone than in the to the ring zone, but the increase was not statistically significant (RR 1.06, 95% CI 0.91, 1.24). Table 4 Number of cancer cases in subjects with residence in the Schiphol area according to site, gender and area of residence, 1988–2003 cancer site (ICD-10 code) and gender area of residence ring zone core zone parameter parameter number of cases SIR# 95% CI number of cases SIR# 95% CI RR## 95% CI All malignancies (C00–C95) 10 855 1.01 0.99, 1.03 2 352 1.06* 1.02, 1.10 1.05* 1.01, 1.10 adult males 5 440 0.98 0.96, 1.01 1 257 1.07 1.01, 1.13 1.09* 1.03, 1.16 adult females 5 356 1.03 1.00, 1.06 1 080 1.04 0.98, 1.11 1.01 0.95, 1.08 children (<15) 59 1.25 0.95, 1.61 15 1.29 0.72, 2.13 1.04 0.59, 1.83 Head & neck (C00–C14) 238 1.06 0.93, 1.21 44 0.90 0.65, 1.21 0.85 0.61, 1.17 males 133 0.92 0.77, 1.09 29 0.89 0.59, 1.27 0.96 0.64, 1.44 females 105 1.32* 1.08, 1.59 15 0.93 0.52, 1.53 0.70 0.41, 1.21 Gastrointestinal tract (C15–C26) 2 422 0.99 0.95, 1.03 467 0.96 0.87, 1.05 0.97 0.88, 1.07 males 1 240 0.98 0.92, 1.03 254 0.95 0.83, 1.07 0.97 0.85, 1.11 females 1 182 1.00 0.95, 1.06 213 0.97 0.85, 1.11 0.97 0.84, 1.12 Respiratory system (C30–C34) 1 484 0.91* 0.86, 0.96 378 1.10 0.99, 1.22 1.21* 1.08, 1.35 males 1 085 0.85* 0.80, 0.90 293 1.08 0.96, 1.21 1.27* 1.12, 1.45 females 399 1.13* 1.02, 1.24 85 1.17 0.93, 1.44 1.04 0.82, 1.31 Breast (C50) 1 731 1.05* 1.01, 1.11 356 1.04 0.93, 1.15 0.99 0.88, 1.11 Female genital organs (C51–C58) 567 0.93 0.86, 1.01 143 1.16 0.98, 1.37 1.24* 1.04, 1.50 Prostate (C61) 1 044 1.02 0.96, 1.08 247 1.19* 1.05, 1.35 1.17* 1.02, 1.34 Bladder & other urinary tract (C65–C68) 468 1.09 0.99, 1.19 102 1.18 0.96, 1.43 1.16 0.93, 1.43 males 341 1.05 0.95, 1.17 84 1.26* 1.01, 1.56 1.20 0.94, 1.52 females 100 0.93 0.76, 1.13 18 0.92 0.54, 1.44 0.98 0.60, 1.63 Hematological malignancies (C81–C96) 855 1.10* 1.03, 1.18 189 1.17* 1.01, 1.35 1.06 0.91, 1.24 males 483 1.16* 1.06, 1.27 115 1.26* 1.04, 1.51 1.09 0.89, 1.33 females 372 1.04 0.94, 1.15 74 1.06 0.83, 1.33 1.02 0.79, 1.31 Other sites 2 073 1.03 0.99, 1.08 426 1.02 0.93, 1.12 0.99 0.89, 1.10 males 1 148 1.03 0.97, 1.09 243 1.01 0.89, 1.14 0.98 0.85, 1.12 females 925 1.03 0.97, 1.10 183 1.04 0.90, 1.20 1.01 0.86, 1.18 * p < 0.05 # reference population: the Netherlands 1989–2003 ## ratio of the SIR of the core zone and the SIR of the ring zone CI = confidence interval; SIR = standardized incidence ratio; RR = rate ratio Discussion The major finding of our study is that total cancer incidence in the area around Schiphol airport was almost equal to the national cancer incidence (SIR 1.02). Furthermore, the incidence of hematological malignancies was statistically significantly increased, while the incidence of cancer of the respiratory system was statistically significantly decreased. We observed an excess risk in children aged 0–14 (SIR 1.26). The cancer incidence in the core zone was slightly increased in comparison to the ring zone, due to an excess risk of cancer of the respiratory tract and prostate in males and cancer of the genital organs in females. As the overall incidence of cancer of the respiratory tract was decreased (SIR 0.94), this observation does not support a positive association between the airport and the occurrence of cancer of the respiratory tract. The incidence pattern of respiratory system cancer in the Schiphol area, i.e. low rates in males and somewhat higher rates in the core zone and among females, is well within the normal regional variation in the Netherlands. Because smoking is the most important risk factor for lung cancer [15], and there is evidence of substantial regional variation in smoking habits in the Netherlands [16], smoking is likely to be responsible for the differences in respiratory system cancer (mainly lung cancer) between the Schiphol area and the Netherlands overall. Unfortunately, no data on smoking habits according to postal code in the Schiphol area are available. Lung cancer incidence in the 1990s in males was lowest in high income areas in the Netherlands. In females, low rates were found in rural areas, while high rates were observed in urban areas [17]. The slightly increased incidence of cancer of the respiratory system in females is in accordance with the moderately urbanized status of the Schiphol area. The data on per capita income (table 2) support the assumption that the low incidence of cancer of the respiratory system in males is related to the high per capita income of the Schiphol area. However, within the Schiphol area only a weak association was observed between the incidence of lung cancer and per capita income by postal code area (data not shown). This may be due to relatively small numbers by postal code area and the long induction period of lung cancer as the regional variation in lung cancer incidence can best be explained by the smoking habits 10 to 30 years ago. In a number of studies in urban areas an increase of lung cancer incidence or mortality was observed [18,19], mostly attributed to differences in smoking habits. However, there is increasing evidence for a relation between lung cancer risk and ambient air pollution [20,21]. Although we cannot exclude the possibility that the incidence of cancer of the respiratory system in the absence of the airport would even have been lower than the observed incidence, the pattern of the observed incidence does not render this very likely. The statistically significantly increased rate for breast cancer in 1994–1998 (SIR 1.09 for the total study area) is also within the observed regional variation in the Netherlands. Part of this variation can be explained by local variation in the start of the national screening program for breast cancer. The relatively high incidence of breast cancer in 1994–1998 is probably related to the start of screening in the Schiphol area in that period. We do not have an explanation for the relatively high incidence of cancer of the female genital organs in the core zone (RR in comparison to the ring zone 1.24). Possibly, this is only a chance finding, as despite the large variation in risk factors for the specific sites the incidence of all specific sites was increased, while the SIR in comparison to the general population was not statistically significantly increased. An association with pollution has not been described for cancer of the female genital organs. Moreover, in the total study area the incidence of these cancers was not increased (SIR 0.97). The most striking observation in our study is the increased incidence of hematological malignancies, which was observed consistently over three periods, mostly in males but also in females. The increase was more pronounced in the core zone. The increased incidence was mostly due to increased numbers of cases of ALL and NHL (especially lymphoplasmocytic lymphoma, follicular lymphoma, diffuse large B-cell lymphoma and T-cell lymphoma, but not small lymphocytic lymphoma/chronic lymphocytic leukemia [SLL/CLL]), while the incidence of Hodgkin lymphoma was decreased. However, pathology could not be reviewed in this study and different classification systems of lymphoma have been used by pathologists during the study period. The moderately increased cancer incidence in children (about one supplementary case per year) was mainly caused by the increased number of cases of ALL, as ALL occurs mainly in children. Also from a national perspective, the number of cases of NHL was markedly increased. In the Netherlands in 1989–1998, the highest rate of NHL in males was found in Greater-Amsterdam, which includes the Schiphol area (source: the Netherlands Cancer Registry). In females in Greater-Amsterdam, the incidence of NHL was also relatively high. Chance is not a likely explanation for our finding of an increased incidence of NHL, since the increase was consistently observed over three time periods. The relatively high incidence of NHL in the Schiphol area is also consistent with an increased mortality due to NHL already reported in Haarlemmermeer in 1981–1986 [7]. In several studies, an increased risk of hematological malignancies was found for farmers, which is related to the use of pesticides, infectious micro-organisms or working with beef cattle [22-27]. However, although the Schiphol area includes a few areas with intensive agricultural activities, the increased risk for hematological malignancies was also found in areas with few agricultural activities. Several studies have shown that the incidence of NHL is correlated with nitrate in municipal drinking water due to nitrogen fertilizers [28,29] and is increased in urban/industrialized areas [30,31]. Hatzissabas et al found that the incidence of large cell high malignancy lymphomas is highest in industrialized regions with pollution of water supplies by more toxic and immunosuppressive substances, while CLL is more frequent in areas with rather low-dose chronic influences such as from the use of fertilizers and pesticides in farming [32]. The pattern of NHL in the Schiphol area – increase of follicular and diffuse large B-cell lymphoma, but not SLL/CLL – might indicate a relation with pollution which is also found in urban areas. However, an association between the incidence of hematological malignancies and the environment in the Schiphol area is not supported by the available data on ambient air quality. Measurements in 1989 at the airport grounds of Schiphol showed increased levels of ambient air pollutants, including polycyclic aromatic hydrocarbons which are probably or possibly carcinogenic according to the International Agency for Research on Cancer, but not in the direct vicinity outside the airport ground [33]. Morphology and composition of soot emitted by aircraft at Schiphol showed great similarities with soot emitted by the road traffic. Only different profiles of hydrocarbons in the range of C6–C12 in emissions from aircraft engines, aviation fuels and road traffic were reported. Since 1994, three locations in the vicinity of Schiphol are part of the provincial monitoring network for ambient air quality measurement [34]. During 1994–2002, the concentrations of the air pollutants NO2, CO, O3, PM10 (particulate matter <10 μm), benzo(a)pyrene, benzene and black smoke at the three locations in the Schiphol area were stable and well comparable to urban background levels in Amsterdam [9]. A more detailed investigation at 59 additional locations in the Schiphol area in 2000/2001 revealed that the average contribution of air traffic emissions and of aviation fuel storage and transfer to the total concentration of volatile hydrocarbons in the area around Schiphol were only 3%, and up to 5–7% at individual locations [35]. Road traffic contributed 28%. For CO, NO2 and PM10, no relevant influence of emissions of Schiphol on ambient pollutant levels could be determined. Although we cannot exclude the possibility that residents of the Schiphol area have been exposed to air pollutants that were not measured or that higher levels of air pollutants have existed in the past, the results of the ambient air quality monitoring and the source appointment of air pollutants render it unlikely that aircraft emissions have contributed substantially to the total levels of pollutants in the ambient air of the Schiphol area. It therefore seems unlikely that the increased incidence of hematological malignancies is specifically related to ambient air pollution caused by aircraft emissions. Our results should be interpreted considering the strengths and limitations of the study design. An advantage is the availability of high quality data from a population-based cancer registry over a period of sixteen years. However, the use of the national cancer incidence as a reference has its limitations. Preferably, the cancer incidence in a population which is comparable to the Schiphol region as far as urbanization, socio-economic status and smoking habits, should be used. Unfortunately, such a reference population is not available. Another limitation of the study is that only cancer cases that were residents of the Schiphol area at the date of diagnosis were included in the study. Part of the original residents will have left the area, while others only recently settled in the area. The effect of migration (non-differential misclassification) usually results in an underestimation of the risk at study. Conclusion The overall cancer incidence in the Schiphol area was similar to the national incidence in the Netherlands. An association was found between residence in the Schiphol area and a moderately increased incidence of hematological malignancies, especially NHL and ALL. However, the increased risk of hematological malignancies could not be explained by higher levels of ambient air pollution in the Schiphol area, while similarly increased rates were observed in Greater Amsterdam. Further studies, for example a study with focus on substances in urban ambient air pollution, are necessary in order to elucidate the causes of the observed association. Competing interests The author(s) declare that they have no competing interests. Authors' contributions JvW and FvL conceived the study and were involved in the design. OV performed the statistical analysis and wrote the first draft of the article. All authors were involved in the interpretation of the results and the revision of the draft. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements We thank the registration clerks of the Comprehensive Cancer Centre Amsterdam for the collection of the data. ==== Refs Tesseraux I Risk factors of jet fuel combustion products Toxicol Lett 2004 149 295 300 15093276 10.1016/j.toxlet.2003.12.040 Passchier W Knottnerus A Albering H Walda I Public health impact of large airports Rev Environ Health 2000 15 83 96 10939086 Knipschild PG Medische gevolgen van vliegtuiglawaai (Health effects of aircraft noise) Thesis 1976 Amsterdam: Coronel laboratorium (UvA) Rosenlund M Berglind N Pershagen G Jarup L Bluhm G Increased prevalence of hypertension in a population exposed to aircraft noise Occup Environ Med 2001 58 769 773 11706142 10.1136/oem.58.12.769 Heisterkamp SH Doornbos G Nagelkerke NJ Assessing health impact of environmental pollution sources using space-time models Stat Med 2000 19 2569 2578 10960872 10.1002/1097-0258(20000915/30)19:17/18<2569::AID-SIM588>3.0.CO;2-K Haines MM Stansfeld SA Job RF Berglund B Head J A follow-up study of effects of chronic aircraft noise exposure on child stress responses and cognition Int J Epidemiol 2001 30 839 45 11511614 10.1093/ije/30.4.839 Van Bruggen M Van Wijnen JH De kankersterfte in de gemeente Haarlemmermeer (1981–1986). Een oriënterend descriptief onderzoek. (Cancer mortality in Haarlemmermeer; a descriptive study) 1989 Amsterdam: Bureau Medische Milieukunde Visser O Van Wijnen JH Benraadt J Van Leeuwen FE Incidentie van kanker in de omgeving van Schiphol in 1988–1993.(Incidence of cancer in the vicinity of Schiphol airport, 1988–1993) Ned Tijdschr Geneeskd 1997 141 468 473 9173287 Meijer W De Jonge D Datarapport Luchtkwaliteit Haarlemmermeer. Resultaten 2002. (Air quality Haarlemmermeer; results 2002) 2003 Haarlem: Provincie Noord-Holland Staatsen BAM Franssen EAM Doornbos G Abbink F Van der Veen AA Heisterkamp SH Lebret E Gezondheidskundige evaluatie Schiphol. (Health evaluation Schiphol) 1993 Bilthoven: RIVM Van der Sanden GA Coebergh JWW Schouten LJ Visser O Van Leeuwen FE Cancer Incidence in the Netherlands in 1989 and 1990. First results of the nationwide Netherlands Cancer Registry Eur J Cancer 1995 31 1822 9 8541107 10.1016/0959-8049(95)00355-M Visser O Coebergh JWW Schouten LJ Incidence of cancer in the Netherlands 1993 Fifth report of the Netherlands Cancer Registry 1996 Utrecht: Vereniging van Integrale Kankercentra Visser O Van Dijck JAAM Siesling S Incidence of cancer in the Netherlands 1999/2000 Eleventh report of the Netherlands Cancer Registry 2003 Utrecht: Vereniging van Integrale Kankercentra Altman DG Practical statistics for medical research 1991 London: Chapman & Hall Doll R Peto R Mortality in relation to smoking: 20 years' observation on male British doctors Br Med J 1979 2 1525 1536 1009386 Mackenbach JP Kunst AE Looman CWN Van Beeck EF Regionale sterfteverschillen in Nederland. (Regional mortality patterns in the Netherlands) T Soc Gezondheidsz 1992 70 396 404 Van Dijck JAAM Coebergh JWW Siesling S Visser O Trends of cancer in the Netherlands, 1989–1998 2002 Utrecht: Vereniging van Integrale Kankercentra Doll R Atmospheric pollution and lung cancer Environ Health Perspect 1978 22 23 31 648488 Hoogendoorn D Regionale verschillen in kankersterfte. (Regional cancer mortality patterns) Ned T Geneesk 1983 127 1516 1525 6633686 Cohen AJ Pope CA III Lung cancer and air pollution Environ Health Perspect 1995 103 219 224 8741787 Vineis P Forastiere F Hoek G Lipsett M Outdoor air pollution and lung cancer: recent epidemiologic evidence Int J Cancer 2004 111 647 52 15252832 10.1002/ijc.20292 Khuder SA Mutgi AB Schaub EA Tano BD Meta-analysis of Hodgkin disease among farmers Scand J Work Environ Health 1999 25 436 441 10569464 Khuder SA Schaub EA Keller-Byrne JE Meta-analyses of non-Hodgkin lymphoma and farming Scand J Work Environ Health 1998 24 255 261 9754856 Viel JF Richardson ST Lymphoma, multiple myeloma and leukaemia among French farmers in relation tot pesticide exposure Soc Sci Med 1993 37 771 777 8211293 10.1016/0277-9536(93)90371-A Amadori D Nanni O Falcini F Chronic lymphocytic leukaemias and non-Hogdkin's lymphomas by histological type in farming-animal breeding workers: a population case-control study based on job titles Occup Environ Med 1995 56 374 379 7627313 Keller JE Howe HL Case-control studies of cancer in Illinois farmers using data from the Illinois State Cancer Registry and the U.S. Census of Agriculture Eur J Cancer 1994 30A 469 473 8018404 10.1016/0959-8049(94)90421-9 Fritschi L Johnson KC Kliewer EV Fry R Canadian Cancer Registries Epidemiology Research Group Animal-related occupations and the risk of leukemia, myeloma, and non-Hodgkin lymphoma in Canada Cancer Causes Control 2002 13 563 571 12195646 10.1023/A:1016331128897 Gulis G Czompolyova M Cerhan JR An ecologic study of nitrate in municipal drinking water and cancer incidence in Trnava District, Slovakia Environ Res 2002 88 182 187 12051796 10.1006/enrs.2002.4331 Weisenburger DD Environmental epidemiology of non-Hodgkin lymphoma in eastern Nebraska Am J Ind Med 1990 18 303 305 2220835 Schouten LJ Meijer H Huveneers JAM Kiemeney LALM Urban-rural differences in cancer incidence in the Netherlands, 1989–1991 Int J Epidemiol 1996 24 729 736 8921449 Doll R Urban and rural factors in the aetiology of cancer Int J Cancer 1991 47 803 810 2010224 Hatzissabas I Krueger GR Medina JR Bedoya VA Papadakis T Environmental pollution and malignant lymphomas: a tentative contribution to geographic pathology Anticancer Res 1993 13 411 417 8517656 Van den Anker IM Van Velze K Onderlinden D Luchtverontreiniging door de luchthaven Schiphol (Air pollution by Schiphol airport) 1989 Bilthoven: Rijksinstituut voor volksgezondheid en milieuhygiëne Nieuwenhuis JW Van der Meij M Lucas MPA De Jonge D Luchtkwaliteit Haarlemmermeer 1994 (Air quality Haarlemmermeer) 1995 Haarlem: Provincie Noord-Holland, Dienst Milieu en Water, Bureau Milieutechnisch Onderzoek Thijsse TR Van Loon M Nader onderzoek naar de luchtkwaliteit in de omgeving van Schiphol en de bijdrage van te onderscheiden bronnen. (Air quality in the Schiphol area; contribution of different sources) 2001 Apeldoorn: TNO Milieu Energie en Procesinnovatie	16332253	PMC1325225	CC BY	2021-01-04 16:28:59	no	BMC Public Health. 2005 Dec 6; 5:127	utf-8	BMC Public Health	2,005	10.1186/1471-2458-5-127	oa_comm
==== Front BMC Health Serv ResBMC Health Services Research1472-6963BioMed Central London 1472-6963-5-761632115510.1186/1472-6963-5-76Research ArticleCorrelates of health and healthcare performance: applying the Canadian health indicators framework at the provincial-territorial level Arah OA [email protected] GP [email protected] Department of Social Medicine, Academic Medical Center, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, the Netherlands2 Center for Prevention and Health Services Research, National Institute of Public Health and the Environment (RIVM), PO Box 1, 3720 BA Bilthoven, the Netherlands3 Tranzo, Faculty of Social and Behavioural Sciences, Tilburg University, PO Box 90153, 5000 LE Tilburg, the Netherlands2005 1 12 2005 5 76 76 9 5 2005 1 12 2005 Copyright © 2005 Arah and Westert; licensee BioMed Central Ltd.2005Arah and Westert; 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, at the health system level, there is little research into the possible interrelationships among the various indicators of health, healthcare performance, non-medical determinants of health, and community and health system characteristics, we conducted this study to explore such interrelationships using the Canadian Health Indicators Framework. Methods We conducted univariate correlational analyses with health and healthcare performance as outcomes using recent Canadian data and the ten Canadian provinces and three territories as units of the analyses. For health, 6 indicators were included. Sixteen healthcare performance indicators, 12 non-medical determinants of health and 16 indicators of community and health system characteristics were also included as independent variables for the analysis. A set of decision rules was applied to guide the choice of what was considered actual and preferred performance associations. Results Health (28%) correlates more frequently with non-medical determinants than healthcare does (12%), in the preferred direction. Better health is only correlated with better healthcare performance in 13% of the cases in the preferred direction. Better health (24%) is also more frequently correlated with community and health system characteristics than healthcare is (13%), in the preferred direction. Conclusion Canadian health performance is a function of multiple factors, the most frequent of which may be the non-medical determinants of health and the community characteristics as against healthcare performance. The contribution of healthcare to health may be limited only to relatively small groups which stand to benefit from effective healthcare, but its overall effect may be diluted in summary measures of population health. Interpreting multidimensional, multi-indicator performance data in their proper context may be more complex than hitherto believed. ==== Body Background In June 2000, in its ambitious comparison of 191 countries in terms of their ability to meet three main goals – improving health, increasing responsiveness to meet the legitimate demands of the population, and ensuring that financial burdens are distributed fairly – the World Health Organization (WHO) ranked Canada 30th in overall health system performance [1]. This was considered a further blow to an already shaken collective psyche in Canada [2]. Canada, which WHO also ranked 35th on health level performance, has taken these rankings to be indicative of serious performance problems, despite the methodological criticisms leveled against the rankings [3]. In its letter to the then minister of health, the Canadian Medical Association called the report "a serious wake-up call" [3]. By September 2000, Canada's Prime Minister and the First Ministers had made a commitment to produce regular public reports on the performance of their health system [4]. As such, the Canadian government has invested heavily in measuring and reporting on the performance of its health system at various levels [4-6]. In doing so and in line with its longstanding 'health determinants' approach to national health policy [7-9], Canada takes a broad health performance approach to quantifying health and healthcare progress [10]. This has entailed the development and use of a multi-dimensional 'health indicators framework' [10] (see Figure 1). This Canadian Health Indicators Framework (CHIF) has four main tiers, namely (a) health status (4 fields); (b) non-medical determinants of health (4 fields); (c) health system performance, or more appropriately referred to as healthcare performance in this paper, (8 dimensions or fields); and (d) community and health system characteristics (3 fields). Many of the fields or dimensions within the framework have so far been populated with indicators. The choice of a health performance framework built on the Lalonde health determinants model should come as no surprise since this was the country that introduced the Lalonde model to the world about three decades ago [9]. It is expected that the multi-dimensional framework will aid the gauging of health progress in a fair and balanced manner. Particularly, it is often assumed that the various fields are interlinked in complex ways that contribute to health and healthcare performance [11]. Healthcare functioning is also taken to be an important contributor to health, notably for those specific populations that stand to benefit from healthcare services. Yet, there has been no research into whether such links exist within the CHIF. This paper aims to examine such links. Using the CHIF as a linked model, this study analyzes the possible relations between, on the one hand, indicators of health (and healthcare performance) and, on the other hand, indicators of non-medical determinants of health, community and health system characteristics, and healthcare performance. Thus, this study poses the question whether the performance of the Canadian provinces/territories on a health (or healthcare) indicator is related to their performance on an indicator of non-medical determinants, healthcare or community characteristic. This study provides an illustrative interlinking of multi-dimensional performance at the provincial/territorial level, but not necessarily the causal interrelationships between and among indicators. We define 'health performance' as the ultimate health outcomes (measured as health status, morbidity or mortality) of a society given its mix of healthcare and non-medical determinants of health. We also define 'healthcare performance' as the degree of maintenance of healthcare system functioning (measured in terms of dimensions such as effectiveness, patient-centeredness and so forth) that is in keeping with the system's societal, professional and user goals and norms. Therefore, healthcare performance should, in principle, contribute to health performance. Methods Study population, data and measures We used recent (2001 to 2003) secondary data on the performance of the thirteen Canadian provinces (ten in number) and territories (a total of 3) usually reported on by the government [12-16] (see Table 1). The ten provinces are Newfoundland & Labrador, Prince Edward Island, Nova Scotia, New Brunswick, Quebec, Ontario, Manitoba, Saskatchewan, Alberta, and British Columbia. The three territories comprise Yukon Territory, Northwest Territories, and Nunavut. These are the 13 jurisdictions that have constitutional responsibility for Canadian health and healthcare. The provinces and territories are of varying sizes, demographic constitution and capacities. The data underpin the annual Health Indicator publications which accompany the national Health Care in Canada reports [12,15]. The data which cover 95% of the Canadian population are appropriately age, population and gender weighted. Their primary collection sources include the Canadian Community Health Survey (in 2000/01 and 2003), Hospitality Mortality Database (CIHI), Discharge Abstract Database (CIHI) and the databases of the Centre for Infectious Disease Prevention and Control (Health Canada). Table 1 Descriptive statistics for the Canadian provinces and territories, in 2001 [12-14] Province/territory Total population Elderly population (%) Urban population (%) Total health expenditure per capita ($) Life expectancy at birth (years) Newfoundland and Labrador (NF) 534 000 11.9 57.6 3,468 78.1 Prince Edward Island (PE) 139 000 13.4 44.5 3,324 78.9 Nova Scotia (NS) 943 000 13.5 55.6 3,208 78.9 New Brunswick (NB) 756 000 13.1 50.2 3,267 79.0 Quebec (QC) 7 418 000 13.0 80.2 3,112 79.4 Ontario (ON) 11 895 000 12.6 84.6 3,492 79.9 Manitoba (MB) 1 149 000 13.6 71.7 3,706 78.6 Saskatchewan (SK) 1 017 000 14.6 64.1 3,422 79.2 Alberta (AB) 3 059 000 10.2 80.7 3,552 79.7 British Columbia (BC) 4 102 000 13.2 84.6 3,569 80.4 Yukon Territory (YK) 30 000 5.8 58.7 4,789 77.5 Northwest Territories (NT) 41 000 4.3 58.3 6,450 77.0† Nunavut (NU) 28 000 2.5 32.4 6,306 69.4 Canada 31 111 0000 12.6 79.6 3,416 79.6 †Pooled over 1995–97 To address the question whether provinces and territories that are better off in one indicator are also better off in the preferred direction in other indicators, we chose two main outcome tiers from the CHIF. First, for the outcome 'health performance' (health) we included 6 indicators from the health status tier (see Table 2): well-being (1 indicator), health conditions (3 indicators), human function (1 indicator), and deaths (1 indicator). Second, for the outcome 'healthcare performance' (healthcare) we included sixteen indicators for healthcare performance tier to cover acceptability, accessibilty, appropriateness, effectiveness, safety, and health surveillance dimensions. There are currently little or no routinely reported indicators for the competence, continuity and efficiency dimensions. The choice of indicators from the entire set of measures within the CHIF was guided by data availability and completeness (for at least ten provincial/territorial units). The independent variables consisted of indicators of non-medical determinants of health, community and health system characteristics, and healthcare performance. For the non-medical determinants of health (see Additional file 1), twelve indicators were chosen covering health behaviors, living and working conditions, personal resources and environmental factors. In addition, sixteen indicators were selected from the community and health system characteristics tier to cover the community, health system and resource fields. The details of these explanatory variables or indicators are available on request from the first author, and can also be found in the reference [12-16]. An indicator was considered positive (negative) if higher (lower) values of it would be preferred in reality. Whenever it was unclear whether a higher or lower level of an indicator would be preferred, we labeled it as both positive and negative (+/-). Table 2 Health and healthcare performance indicators: descriptions, estimates, Canadian averages, and data sources [12-16] Indicator Description Provincial & Territorial value Range Canadian average (95% C.I.) Data source Health status 'Tier' • Well-being Self-rated health (excellent or very good) [+] Percentage of the population aged 12 and over who rate their own health status as either excellent or very good 53.2–66.2% 61.4 (61.0–61.8) Statistics Canada, CCHS 2000–01 • Health conditions Body mass index higher than 27 [-] Body weight in kilograms divided by the squared value of the height in meters with values greater than 27 (overweight) for those aged 20 to 64 29.0–42.8 kg/m2 31.9 (31.4–32.3) Statistics Canada, CCHS 2000–01 Asthma rate [-] Percentage of the population aged 12 and over who report that they have been diagnosed by a health professional as having asthma 3.6–9.2% 8.4 (8.2–8.6) Statistics Canada, CCHS 2000–01 Diabetes rate [-] Percentage of the population aged 12 and over who report that they have been diagnosed by a health professional as having diabetes 1.9–5.8% 4.1 (4.0–4.3) Statistics Canada, CCHS 2000–01 • Human function Functional health (perfect of very good) [+] Percentage of the population aged 12 and over reporting measures of overall functional health, based on eight dimensions of functioning (vision, hearing, speech, mobility, dexterity, feelings, cognition and pain) 66.3–84.9% 80.5 (80.1–80.8) Statistics Canada, CCHS 2000–01 • Deaths Life expectancy [+] Life expectancy at birth calculated in years for overall population 69.4–80.4 years 79.6 (79.6–79.7) Statistics Canada Healthcare performance 'Tier' • Acceptability Satisfied with family doctor [+] Population aged 15 and above who report being very or somewhat satisfied with the most recent family doctor or other physician care received 88.3–93.5% 91.4 (90.7–92.0) Statistics Canada, CCHS 2003 Satisfied with health care services [+] Population aged 15 and above who report being very or somewhat satisfied with health care services received in the past 12 months 74.2–88.6% 84.9 (84.3–85.6) Statistics Canada, CCHS 2003 Satisfied with community health care [+] Population aged 15 and above who report being very or somewhat satisfied with community health care received in the past 12 months 78.0–91.9% 83.0 (81.4–84.6) Statistics Canada, CCHS 2003 • Accessibility Screening mammography [+] Women aged 50 to 69 who reported having had a mammogram for routine screening within the past 2 years 36–54% 52 Statistics Canada, CCHS 2000/01 Pap smear [+] Women aged 18 to 69 who reported having had a Pap smear test for routine for routine screening within the past 3 years 65–81% 73 Statistics Canada, CCHS 2000/01 Difficulties accessing routine care [-] Population aged 15 and above who report difficulties accessing routine or on-going care, among those who required care at any time of day 12.2–20.4% 16.4 (15.3–17.5) Statistics Canada, CCHS 2003 Difficulties accessing health information [-] Population aged 15 and above who report difficulties accessing health information or advice, among those who required care at any time of day 12.3–17.8% 15.8 (14.7–16.9) Statistics Canada, CCHS 2003 • Appropriateness Vaginal birth after Caesarean section [+] Proportion of women who have previously had a caesarean section who give birth via vaginal delivery in an acute care hospital 12.5–60.7% 26.7 (26.2–27.2) CIHI, Discharge abstract database Caesarean sections [-] Proportion of women delivering babies in acute care hospitals by caesarean section (stillbirths are excluded from denominator due to database characteristic) 9.2–27.9% 22.5 (22.3–22.6) CIHI, Hospital mortality database • Effectiveness In-hospital 30-day stroke mortality [-] Risk-adjusted rate of all-cause in-hospital death occurring within 30 days of first admission to an acute care hospital with a diagnosis of stroke (aged 20 to 105 years) 15.5–24.2% 18.7 CIHI, Hospital mortality database Pneumonia readmission rate [-] Risk-adjusted rate of unplanned readmission following discharge for pneumonia (aged 15 to 84 years) within 28 days of index episode; based on three years of pooled data 2.7–6.9% 3.2 CIHI, Discharge abstract database Ambulatory care sensitive conditions [-] Age-standardized inpatient acute care hospitalization rate for conditions where appropriate ambulatory care prevents or reduces the need for hospitalization, per 100,000 population 243–1,114 346 (344–348) CIHI, Hospital mortality database Pneumonia & influenza hospitalizations [-] Age-standardized acute care hospitalization rate for pneumonia and influenza, per 100,000 population aged 65 and older 482–2566 768 (760–777) CIHI, Hospital mortality database • Safety Hip fracture hospitalizations [-] Age-standardized hospitalization rate for fracture of the hip, per 100,000 population aged 65 and older 495–660 554 (547–561) CIHI, Hospital mortality database • Other: health surveillance Chlamydia [-] Number of cases of genital chlamydia reported, per 100,000 population 89.14–2514.29 149.19 CIDPC, 2000 Hepatitis C [-] Number of cases of hepatitis C reported, per 100,000 population 9.74–156.67 60.42 CIDPC, 2000 [-] implies that lower levels of the indicator are preferred; [+] implies that higher levels of the indicator are preferred CCHS: Canadian Community Health Survey CIHI: Canadian Institute for Health Information CIDPC: Centre for Infectious Disease Prevention and Control, Health Canada Analysis Pearson's correlation coefficient r was used to estimate the univariate associations between the indicators of the 2 main outcomes (health and healthcare) and the indicators of non-medical determinants, community and health system characteristics, and healthcare performance. The unit of analysis was each of the thirteen Canadian provinces and territories. The indicators were log-transformed to avoid spurious correlations between rate-based indicators. Furthermore, given the small sample size (N = 13), we used critical (cut-off) values of r to ensure that the correlation was real and to minimize the chances of committing type I error, that is incorrectly rejecting a true statistical null hypothesis. At a significance level of 5%, the critical values of r for small sample sizes are as follows: r* ≥ 0.552, 0.576, 0.602 or 0.631 for N = 13, 12, 11, or 10 respectively. At a significance level of 1%: r* ≥ 0.683, 0.707, 0.734 or 0.764 for N = 13, 12, 11, or 10 respectively [17]. Another conservative statistical choice was that we used two-sided p-values to assess the significance of the correlations. We applied decision rules to aid the interpretation of significant correlations. A correlation between any two indicators i and j was considered a significant preferred performance association if the coefficient r was positive when both i and j were positive or when both were negative. However, if one indicator was positive while the other was negative, the correlation between them was considered a significant preferred performance association if the coefficient r was negative. In all cases, the abovementioned critical value (r) requirement must have been met. Assuming a null hypothesis H0 that r = 0, the following decision rules were applied to the univariate correlations: (a) if indicators i and j were both either positive (+) or negative (-), then the following decision rule applied to their preferred positive correlation rij: r ≥ r → reject H0 r <r* → do not reject H0 (b) if only one of the indicators i and j was positive (+), then the following decision rule applied to their preferred negative correlation r (where r* took on a negative value): r ≤ r* → reject H0 r > r* → do not reject H0 Whenever the correlation was significant but not in the preferred direction, the result was considered a suboptimal performance association (termed in this paper as not preferred performance) that could point towards possibilities for improvement. To estimate the uncertainty around each estimated r, we wrote a simple spreadsheet for calculating the 95% confidence interval (C.I.) for each sample size. (This spreadsheet is available on request from the first author.) As far as the decision rules were concerned, caution was exercised in applying them to indicators for which it was unclear whether higher or lower values would be preferred. For this reason, the decision rules were not applied to three indicators of community characteristics, namely population size, elderly population, and urban population. All analyses were carried out using SPSS version 12.0.2 (SPSS Inc., Chicago, IL) and Microsoft® Excel 2002 SP3 (Microsoft Corporation, Redmond, WA). Results Table 3 shows that significant 'preferred' correlations between health and non-medical determinants of health range from -0.853 (95% C.I.: -0.955 to -0.570; P < 0.05) for the association between body mass index and dietary practices to 0.836 (95% C.I.: 0.528 to 0.950; P < 0.01) for the association between unemployment rate and diabetes rate. Smoking status, having high proportions of high school and post-secondary graduates displayed unfavorable ('not preferred') associations with health indicators (Table 3). Table 3 Correlations between health (status) indicators and non-medical determinants of health Self-rated health (excellent or very good) [+] Body mass index higher than 27 [-] Asthma rate [-] Diabetes rate [-] Functional health (perfect or very good) [+] Life expectancy [+] Non-medical determinants of health Health behaviors Smoking status [-] -0.586† 0.337 -0.688† -0.794‡ -0.583† -0.633‡ Frequency of heavy drinking [-] -0.391 0.571† -0.252 -0.437 -0.161 -0.783‡ Leisure-time physical activity [+] 0.121 -0.600† 0.079 -0.339 0.116 0.710† Dietary practices [+] 0.420 -0.853† 0.258 0.135 0.269 0.633† Living and working conditions High school graduates [+] 0.490 -0.485 0.855‡ 0.622† 0.712‡ 0.754† Post-secondary graduates [+] 0.451 -0.456 0.711‡ 0.237 0.556† 0.585 Unemployment rate [-] 0.524 0.546 -0.497 0.836‡ 0.432 -0.727† Youth unemployment [-] 0.396 0.398 -0.436 0.791‡ 0.260 -0.284 Low income rate [-] -0.050 -0.143 -0.419 0.145 0.556† -0.196 Average personal income [+] -0.203 -0.411 0.207 -0.546† 0.063 0.821‡ Personal resources Life stress [-] -0.008 -0.468 0.652† 0.177 0.516 0.581 Environmental factors Exposure to second-hand smoke [-] -0.252 0.590† 0.083 -0.071 0.238 -0.676† [-] implies that lower levels of the indicator are preferred [+] implies that higher levels of the indicator are preferred † P < 0.05 ‡ P < 0.01 Bold: correlation is significant in the possibly preferred direction and exceeds the critical level necessary for the sample size Italicized: correlation is in the possibly preferred direction but is not significant at the critical level necessary for the sample size Table 4 shows the correlations between health and healthcare performance indicators. Here, the significant 'preferred' correlations range from -0.782 (95% C.I.: -0.932 to -0.406; P < 0.01) for the association between provincial/territorial performance on diabetes rate and vaginal birth after Caesarean section to 0.754 (95% C.I.: 0.347 to 0.922; P < 0.01) for the association between provincial/territorial performance on diabetes rate and in-hospital 30-day stroke mortality. Table 5 shows the correlations of health indicators with community and health system characteristics. Again, significant 'preferred' correlations range from -0.893 (95% C.I.: -0.968 to -0.673; P < 0.01) for the association between knee replacement and functional health status (as perfect or very good) to 0.810 (95% C.I.: 0.468 to 0.941; P < 0.01) for the association between provincial/territorial performance on diabetes rate and hysterectomy. Likewise, several 'not preferred' performance associations exist between health indicators and mostly (health system) resource indicators (Table 5). Table 4 Correlations between health (status) indicators and healthcare performance indicators Self-rated health (excellent or very good) [+] Body mass index higher than 27 [-] Asthma rate [-] Diabetes rate [-] Functional health (perfect or very good) [+] Life expectancy at birth [+] Healthcare performance Acceptability Satisfied with family doctor [+] 0.300 0.356 0.360 0.788‡ 0.440 -0.605 Satisfied with health care services [+] 0.479 -0.012 0.756‡ 0.727‡ 0.708 -0.234 Satisfied with community health care [+] -0.170 0.367 -0.445 0.055 -0.073 -0.500 Accessibility Screening mammography [+] 0.168 -0.362 -0.025 0.247 -0.266 0.668† Pap smear [+] 0.275 0.346 0.622† 0.488 0.403 0.470 Difficulties accessing routine care [-] 0.595 0.248 -0.103 0.447 0.760† -0.657† Difficulties accessing health information [-] 0.141 -0.074 0.188 0.277 0.021 -0.049 Appropriateness Vaginal birth after Caesarean section [+] -0.416 -0.197 -0.671† -0.782‡ -0.770† 0.309 Caesarean sections [-] 0.511 -0.034 0.700† 0.706‡ 0.667† -0.005 Effectiveness In-hospital 30-day stroke mortality [-] 0.273 0.687† -0.381 0.754‡ 0.079 -0.754† Pneumonia readmission rate [-] 0.190 -0.524 0.406 -0.347 0.112 0.307 Ambulatory care sensitive conditions [-] -0.112 0.467 0.304 0.016 0.308 -0.392 Pneumonia & influenza hospitalizations [-] -0.434 0.502 -0.089 -0.380 -0.153 -0.410 Safety Hip fracture hospitalizations [-] -0.305 0.070 -0.190 -0.615† -0.153 0.398 Other: health surveillance Chlamydia [-] -0.661† 0.216 -0.818‡ -0.869‡ -0.776‡ -0.061 Hepatitis C [-] -0.082 -0.669† 0.309 -0.321 0.212 0.823‡ [-] implies that lower levels of the indicator are preferred [+] implies that higher levels of the indicator are preferred † P < 0.05 ‡ P < 0.01 Bold: correlation is significant in the possibly preferred direction and exceeds the critical level necessary for the sample size Italicized: correlation is in the possibly preferred direction but is not significant at the critical level necessary for the sample size Table 5 Correlations between health (status) indicators and community & health system characteristics Self-rated health (excellent or very good) [+] Body mass index higher than 27 [-] Asthma rate [-] Diabetes rate [-] Functional health (perfect or very good) [+] Life expectancy at birth [+] Community & health system characteristics Community¥ Population [+/-] 0.308 -0.552 0.221 0.125 0.239 0.620 Elderly population [+/-] 0.461 -0.084 0.512 0.841 0.449 0.780‡ Dependency ratio [-] -0.525 0.252 -0.739‡ -0.497 -0.814‡ -0.638† Urban population [+/-] 0.350 -0.646† 0.528 0.213 0.518 0.708† Health system Hip replacement [-] -0.105 -0.092 0.243 -0.065 -0.311 0.196 Knee replacement [-] -0.652‡ 0.228 -0.628† -0.749‡ -0.893‡ -0.774‡ Hysterectomy [-] 0.396 0.252 0.459 0.810‡ 0.356 0.460 Bypass surgery [-] 0.634† 0.236 -0.029 0.584‡ 0.155 -0.701† Resources Total health expenditure per capita [+] -0.467 0.068 -0.727‡ -0.814‡ -0.694‡ -0.773‡ Public sector health expenditure per capita [+] -0.473 0.084 -0.544 -0.807‡ -0.510 -0.417 General/family physicians [+] 0.504 -0.416 0.645‡ 0.436 0.641‡ 0.575† Certified specialists [+] 0.399 -0.280 0.556† 0.626† 0.499 0.760‡ Registered nurses [+] -0.310 0.738‡ -0.175 -0.122 0.011 -0.435 Licensed practical nurses [+] 0.496 0.663‡ -0.386 0.756‡ 0.206 -0.336 Pharmacists [+] 0.546 0.010 0.798‡ 0.646‡ 0.728‡ 0.782‡ Total physicians [+] 0.576† -0.445 0.772‡ 0.696‡ 0.730‡ 0.872‡ [-] implies that lower levels of the indicator are preferred [+] implies that higher levels of the indicator are preferred † P < 0.05 ‡ P < 0.01 Bold: correlation is significant in the possibly preferred direction and exceeds the critical level necessary for the sample size Italicized: correlation is in the possibly preferred direction but is not significant at the critical level necessary for the sample size ¥ Only the dependency ratio indicator is assessed here using the decision rule since the other indicators could be preferred either way depending on the goal and audience The table in Additional file 2 gives an overview of the correlations between healthcare performance, on the one hand, and non-medical determinants of health and community and health system characteristics, on the other hand. The significant 'preferred' correlations range from -0.863 (95% C.I.: -0.958 to -0.595; P < 0.01) for the association between provincial/territorial performance on screening mammography and its frequency of heavy drinking to 0.944 (95% C.I.: 0.819 to 0.983; P < 0.01) for the performance association between smoking status and chlamydia cases per unit population. Several 'not preferred' associations also exist for the healthcare performance outcome. For instance, the share of public sector health expenditure per capita shows several strong correlations with healthcare indicators: ranging from -0.716 (95% C.I.: -0.909 to -0.273; P < 0.01) for being satisfied with family doctor to 0.851 (95% C.I.: 0.565 to 0.954; P < 0.01) for chlamydia cases per unit population. Preferred inter-correlations among healthcare performance indicators ranged from -0.867 (95% C.I.: -0.960 to -0.605; P < 0.01) for Caesarean section rate versus vaginal birth after Caesarean section to 0.716 (95% C.I.: 0.273 to 0.909; P < 0.05) for the association between being satisfied with family doctor and being satisfied with healthcare services. Tables 6 and 7 present the summaries of the correlates of health and healthcare performance. Table 6 shows that there are relatively more 'preferred' associations between health and non-medical determinants (that is, 20 out of 72 correlations, or 28%) than between health and healthcare performance indicators (that is, 12 out of 96 correlations, or almost 13%). Similarly, there are relatively fewer 'not preferred' correlations between health and non-medical determinants (that is, 6 out of 72 correlations, or 8%) than between health and healthcare performance indicators (that is, 11 out of 96 correlations, or 11%). Also, the associations between health and community/health system characteristics out-number those between health and healthcare performance. There are 19 significant 'preferred' associations out of 78 correlations between health and community/health system characteristics (that is, almost 24%), while there are 18 significant 'not preferred' associations (that is, 23%). Table 7 shows that out of 190 correlations between healthcare and non-medical determinants, the significant 'preferred' and 'not preferred' associations are respectively 23 (12%) and 18 (9%). Based on 208 correlations between healthcare and community/health system characteristics, there are 27 (13%) and 35 (17%) significant 'preferred' and 'not preferred' associations respectively. Interrelationships among healthcare indicators are few in number within and between dimensions (details not shown but summarized in Table 7; see Additional file 2). Table 6 Summary of significant 'preferred' and 'not preferred' performance correlates of health (status) indicators† Self-rated health (excellent or very good) Body mass index higher than 27 Asthma rate Diabetes rate Functional health (perfect or very good) Life expectancy Row Total¥ Non-medical determinants of health (number of indicators) Health behaviors (4) 1/0 3/0 0/1 0/1 1/0 4/0 9/2/24 Living and working conditions (6) 0/0 0/0 0/2 3/1 2/1 3/0 8/4/36 Personal resources (1) 0/0 0/0 1/0 0/0 0/0 0/0 1/0/6 Environmental factors (1) 0/0 1/0 0/0 0/0 0/0 1/0 2/0/6 Sub-Total 20/6/72 Healthcare performance (number of indicators) Acceptability (3) 0/0 0/0 0/1 0/2 0/0 0/0 0/3/18 Accessibility (4) 0/0 0/0 0/1 0/0 0/1 2/0 2/2/24 Appropriateness (2) 0/0 0/0 2/0 2/0 0/2 0/0 4/2/12 Effectiveness (4) 0/0 1/0 0/0 1/0 0/0 1/0 3/0/24 Safety (1) 0/0 0/0 0/0 0/1 0/0 0/0 0/1/6 Other: health surveillance(2) 1/0 0/1 1/0 0/1 1/0 0/1 3/3/12 Sub-Total 12/11/96 Community and health system characteristics (number of indicators) Community (1)‡ 0/0 0/0 0/1 0/0 1/0 1/0 2/1/6 Health system (4) 1/1 0/0 0/1 2/1 1/0 2/0 6/3/24 Resources (8) 1/0 0/2 1/4 2/4 3/1 4/3 11/14/48 Sub-Total 19/18/78 Column Total 4/1 5/3 5/11 10/11 9/5 18/4 - †Numbers (x/y) in cells respectively represent the number of indicators which showed significant preferred performance association (x) and the number of indicators which showed significant not preferred performance association (y) at the provincial/territorial level ¥Total number of significant preferred correlations/total number of significant not preferred correlations/total number of tested correlations for that performance dimension or indicator group ‡Only the dependency ratio indicator is assessed here using the decision rule since the other indicators could be preferred either way depending on the goal and audience Table 7 Summary of significant 'preferred' and 'not preferred' correlates of healthcare performance indicators† Acceptability (3) Accessibility (4) Appropriateness (2) Effectiveness (4) Safety (1) Other: health surveillance (2) Row Total¥ Non-medical determinants of health (number of indicators) Health behaviors(4) 1/1 3/0 0/2 4/1 0/0 2/1 10/5/64 Living and working conditions (6) 1/3 4/1 3/3 2/1 0/1 1/4 11/13/96 Personal resources (1) 0/0 1/0 0/0 0/0 0/0 0/0 1/0/16 Environmental factors (1) 0/0 0/0 0/0 1/0 0/0 0/0 1/0/16 Sub-Total 23/18/190 Community and health system characteristics (number of indicators) Community (1)‡ 0/0 1/1 0/2 0/0 0/0 1/0 2/3/16 Health system (4) 3/2 0/2 1/2 0/0 0/1 1/1 6/8/64 Resources (8) 4/4 2/3 3/6 3/7 2/1 5/3 19/24/128 Sub-Total 27/35/208 Healthcare performance (number of indicators) Acceptability (3) - 0/0 0/1 0/1 1/0 2/0 - Accessibility (4) - - 0/2 2/0 0/0 1/1 - Appropriateness (2) - - - 0/0 0/0 0/2 - Effectiveness (4) - - - - 2/0 1/1 - Safety (1) - - - - - 1/0 - Other: health surveillance (2) - - - - - - - Column Total 9/10 11/7 7/18 12/10 5/3 15/13 - †Numbers (x/y) in cells respectively represent the number of indicators which showed significant preferred performance association (x) and the number of indicators which showed significant not preferred performance association (y) at the provincial/territorial level ¥ Total number of significant preferred correlations/total number of significant not preferred correlations/total number of tested correlations for that performance dimension or indicator group ‡Only the dependency ratio indicator is assessed here using the decision rule since the other indicators could be judged either way depending on the audience -; not applied Discussion This is the first study to estimate the correlates of health and healthcare performance of Canadian provinces/territories. It suggests that relatively better performance on non-medical determinants of health is related to better health. Healthcare performance is, however, less frequently related to health. In addition, health is relatively better associated with community/health system characteristics than healthcare performance is. Provincial/territorial healthcare performance shows relatively more 'preferred' than 'not preferred' associations with non-medical determinants. Healthcare correlations with community and health system characteristics show the reverse picture, with relatively more 'not preferred' associations than 'preferred associations'. This again suggests there is still more to be desired in how provinces/territories simultaneously optimize their performance in terms of health and healthcare, given their community and health system characteristics. Interrelationships between healthcare performance indicators suggest that how the healthcare system performs in terms of one indicator is often not related to its performance in terms of another indicator. Importantly, there are at least two ways of looking at the correlations. First, the correlations can be interpreted as possible associations between the epidemiological factors that underlie the indicators (that is, epidemiological associations) in ideal circumstances; for example, body mass index is associated with dietary practices in an epidemiological sense. Second, the correlations could be viewed as no more than associations between the actual performance attainments of provinces/territories in terms of different indicators (that is, performance associations) in everyday circumstances. Although both views are related and implied in this study, we recognize that the latter can undermine the former when real "epidemiological associations" are not observed in sub-optimal or "not preferred" performance scenarios. Explanation of results Health is a function of multiple factors or determinants that work in complex, sometimes unclear ways [7-9,18]. The results of this study, although based on multiple univariate correlations, support this notion. In a similar correlational analysis used in a study of 311 local administrative units covering 70 million populations in Japan, a group of nine health determinant indices (namely, healthcare resources, preventive health activities, environmental quality, housing urban clutter, local economy, employment, income, and education) explained almost 52% of the variances in health index levels in the cities studied [19]. In our study, it could be shown that, if independently assessed, non-medical determinants could explain between 40% and 67% (calculated from r-squared) of the variance in life expectancy as a measure of health at the provincial/territorial level (see Table 3). Similarly, healthcare performance indicators could independently account for 44% to 57% of the variance in life expectancy. Unsurprisingly, health has relatively more associations with non-medical determinants than with healthcare indicators. Studies looking at healthcare inputs and resources to explain variations in the health of countries or other ecological units of analysis have mostly failed to demonstrate any or consistent associations [20-22]. However, it is also possible that the non-medical determinants correlate more frequently with health levels because they represent factors (such as dietary practices, smoking and so on) that are more or less related to disease risk profiles, prevalence and incidence in a general population. Healthcare factors reflect mostly corrective or management measures that marginally influence the prevalence or incidence of chronic ill-health or diseases, particularly in the face of co-morbidities in risk populations. This does not imply that healthcare is not life-saving for those who need it, when they need it. The point is that the contribution of healthcare to the general health of a population may be limited only to relatively small groups, in time and space, which stand to benefit from effective healthcare, but its overall effect will be diluted in summary measures of population health or well-being. Therefore, using the prevalence of diagnosed health conditions such as diabetes rates as indicators of health performance can only yield disappointing results in relation to healthcare performance. This explanation may be relevant to the results detailed in Table 4 when compared to Table 3 findings. There are also results (the significant 'not preferred' correlations) which suggest that provinces may be struggling with optimizing their health performance given their health determinants, and healthcare and community profiles. For example, the higher the percentage of high school or post-secondary graduates in a province/territory, the higher the asthma rate (r = 0.855 or 0.711, P < 0.01, in Table 3). This may be expected given that asthma is more prevalent among the younger age groups. The comparable frequency with which health displays both 'preferred' and 'not preferred' associations with community and health system characteristics also points to the possibility that health levels are shaped community needs in complex ways that this study cannot disentangle. Moreover, it is difficult to say which indicator precedes the other in this study. Healthcare indicators may just be current responses to perceived previous shortcomings in health. For instance, the negative correlation (r = -0.773, P < 0.01, in Table 5) between health expenditure and life expectancy could be due to an increase in total healthcare spending in provinces/territories with a long history of lower health levels. Nunavut, for example, is a collection of 26 communities with 28,000 inhabitants living in a vast territory about one-fifth the size of Canada. Nunavut is only accessible by air or sea, and has substantial difficulties in recruiting and retaining health professionals although it spends twice as much as the Canadian average on health per capita (Table 1). Given, the poorer-than-Canadian-average health in Nunavut, the government has been investing a lot in health and healthcare there. Therefore, health expenditure will understandably display a negative association with life expectancy. In this study, we assumed that the lower rates of knee replacement or bypass surgery and other contextual health system indicators, when seen in the context of higher health outcomes, will be preferred. It could, however, be argued that lower rates of such contextual indicators could be indicative of unmet needs (locally). Given such interpretations, we would have to reverse the interpretations and associated correlations to reflect the possibility of under-use of needed healthcare in such communities. Nonetheless, our current interpretations allow for the possibilities of over-use of appropriate healthcare in communities where health outcomes are already high. These considerations further highlight the often overlooked difficulties that are inherent in understanding published performance data, regardless of the amount of contextual information provided. In a sense, the meaning and excellence of performance may be in the eye of the beholder. Implications Recommending policies on health and healthcare in the provinces and territories must take into account the responsibility structure, organization, delivery and funding of healthcare in the concerned areas. Blanket recommendations will probably miss the point by being too generic and insensitive to local needs. The Canadian healthcare system is mainly publicly funded (Medicare) [4]. The provinces and territories have primary constitutional responsibility for health, and the management and delivery of health services, although they all adhere to a set of federal principles in view of the Canadian history of fiscal transfers from the federal to the provincial governments [23]. A number of interlocking general revenue-financed health insurance plans cover hospital in-patient and out-patient services, pharmaceutical products, physician services and public health services. Therefore, there is, in principle, a lot of improvement leverage points that provincial, territorial and federal governments can use to better the health of Canadians. Nevertheless, there are serious challenges and tensions posed by the varying needs of multiple stakeholders and the use of broad summary indicators or a parsimonious set of indicators. It is advisable that governments invest more in investigating and interpreting possible linkages among performance results in order to aid learning and to facilitate the simultaneous optimization of different performance dimensions and indicators. That said, it is also becoming increasingly clearer that investing in public health, especially in promoting healthy life style, disease prevention and health protection, may still offer new avenues for dealing with population health in western societies [24-26]. The current narrow focus on technical care may not be the best approach to improving and maintaining population health when most gains are to be made by living and working well, for example. The social choice arrangements needed to achieve better health for Yukon Territory, Northwest Territories and Nunavut will be have to be ambitious. There are signs that some provinces are already investing more in both health and healthcare performance of their communities [27-29]. Study limitations The data used in this study come from multiple sources with different data elements and quality. Although the Canadian government continues to invest in the quality and coverage of the data used in constructing the indicators, the system is still not perfect [16]. To ensure data quality and comparability in performance reporting, the Canadian First Ministers have been giving policy support to the federal government and the 13 provincial/territorial jurisdictions since September 2000 [4,16]. In February 2003, the First Ministers' Accord on Health Care Renewal directed Health Ministers to further develop indicators to supplement the work on comparable indicator reporting. So far, about 70 indicators have been standardized for comparable reporting at the provincial/territorial and federal government levels. Furthermore, it is beyond the scope of this study to ascertain which indicator is really a cause or an effect. Correlation does not imply causality, and this is troublesome in ecological observational study designs [30]. Bearing this in mind, we only hope to speak to the (sub)optimization of a pair of indicators based on the attained performance. We also realize that statistical significance does not necessarily imply substantive significance of performance. Unconfounded associations, temporality, and real-world translation of the performance associations, particularly causal ones, are more appropriate criteria for assessing importance of the association. Such assessments will also entail value judgments pertaining to how good the performance levels may be. Besides, it is quite possible there are important lag effects of health determinants and other indicators on health and healthcare performance that this study will be unable to pick up, given its contemporaneous cross-sectional ecological design. A vexing limitation is the issue of multiple correlations and significance testing, given also that the health and healthcare indicators are not independent. At a significance level of 5%, there is a 1 in 20 chance of getting a spurious significance, just by chance. Thus, given the large number of correlations conducted in this study, it is quite likely that some correlations occurred by mere chance. However, given the rather low to moderate number of significant results and the fact that we actually pre-specified our paired analyses, it is likely that the magnitude of errors introduced by the multiple independent correlations using the same sample file will be relatively minimal. Besides, our decision rules were rather conservative. We could have set a stricter significance level by, say, dividing 0.05 by the number of anticipated tests (that is, using the Bonferroni method) [31]. Although this would minimize our type I error rate, it would have depleted our statistical power, thus giving a higher probability of type II error (that is, the probability of rejecting an association that actually exists). A trade-off between committing type I error and having enough statistical power was thus necessary, particularly given our already small sample size. Conclusion The results of this exploratory study should serve as a provocative basis for future research into performance interrelationships. The prevailing assumption that publishing a comprehensive battery of indicators will automatically lead to clearer understanding and contextualization of performance is not tenable. This study forces us to take a closer look at how we actually interpret such indicators in relation to one another, and we have seen that there are no easy rules for understanding possible links between what a community attains in one indicator and what it achieves in another. Since performance is interventionist in nature, health and healthcare performance can be influenced by those who have the ability and resources to do so. This study suggests that indicators which are correlated with how well Canadian provinces and territories perform in terms of health and healthcare can act as leverage points for improving health. This study has many implications for further research on linkages within performance frameworks now being used in several industrialized countries, and for choosing a national performance framework. For instance, a framework that focuses mostly on healthcare performance (e.g. the US National Healthcare Quality Report framework or the now old UK NHS performance assessment framework) does so at the expense of understanding the links between non-healthcare determinants and population health. Further elucidation of the meanings, nature, and extent of the interrelationships among the different fields and domains of the Canadian or any other performance framework will aid actual performance improvement by pointing the responsible governments in the right direction. Competing interests The author(s) declare that they have no competing interests. Authors' contributions OAA conceived and supervised the study, collected the data, completed the analyses, and led the writing. GPW assisted with the study, interpretation of results and critical review of the manuscript for intellectual content. Figure 1 Canadian health indicators framework (adapted from public domain sources [12-14]). (Legend/Footnote: The third tier "healthcare performance" is originally titled "health system performance" in the Canadian public domain documents, refs. [12-14]) Pre-publication history The pre-publication history for this paper can be accessed here: Supplementary Material Additional file 1 Canadian indicators of non-medical determinants of health (Table in Word format; describing the non-medical determinants of health used in this study) Click here for file Additional file 2 Correlations between healthcare performance indicators§, and non-medical determinants of health and community & healthcare system characteristics (Table in Word format; displaying extensively the correlation coefficients between healthcare performance indicators, on the one hand, and non-medical determinants of health, community characteristics and health system characteristics, on the other hand) Click here for file Acknowledgements This study was supported within Project S/260806 (on performance indicators) atthe National Institute of Public Health and the Environment (RIVM), Bilthoven, the Netherlands. We are grateful to Karien Stronks, PhD and Niek Klazinga, MD, PhD for their insightful comments. ==== Refs World Health Organization The world health report 2000 Health systems: improving performance 2000 Geneva: WHO Lewis S Donaldson C Mitton C Currie G The future of health care in Canada BMJ 2001 323 926 929 11668143 10.1136/bmj.323.7318.926 Deber R Why did the World Health Organization rate Canada 30th? Some thoughts on league tables Longwoods Review 2004 2 2 7 Arah OA Klazinga NS Delnoij DMJ ten Asbroek AHA Custers T Conceptual frameworks for health systems performance: a quest for effectiveness, quality and improvement Int J Qual Health Care 2003 15 377 398 14527982 10.1093/intqhc/mzg049 Canadian Institute for Health Information, Statistics Canada Canadian health information roadmap initiative indicators framework 2000 Ottawa: Canadian Institute for Health Information Commission on the Future of Health are in Canada (Roy J. Romanow commissioner) Building on values: the future of health care in Canada Final report 2002 Ottawa: Queen's Printer Evans RG Stoddart GL Producing health, consuming health care Soc Sci Med 1990 31 1347 1363 2126895 10.1016/0277-9536(90)90074-3 Evans RG Stoddart GL Consuming research, producing policy? Am J Public Health 2003 93 371 379 12604475 Lalonde M A new perspective on the health of Canadians 1974 Ottawa: Office of the Canadian Minister of National Health and Welfare Arah OA Westert GP Perspectives on conceptualization of health and healthcare systems performance 2004 Bilthoven: RIVM International Organization for Standardization Health informatics – health indicators definitions, relationships and attributes ISO TC 215/SC N339 2003 Geneva: ISO Canadian Institute for Health Information Health indicators 2004 2004 Ottawa: CIHI Canadian Institute for Health Information Health indicators 2003 2003 Ottawa: CIHI Canadian Institute for Health Information Health indicators 2002 2002 Ottawa: CIHI Canadian Institute for Health Information Health indicators Health Canada Healthy Canadians A federal report on comparable health indicators 2004 2004 Ottawa: Her Majesty the Queen in Right of Canada Altman DG Practical statistics for medical research 1991 London: Chapman and Hall Kindig D Stoddart G What is population health? Am J Public Health 2003 93 380 383 12604476 Takano T Nakamura K An analysis of health levels and various indicators of urban environments for Healthy Cities projects J Epidemiol Community Health 2001 55 263 270 11238582 10.1136/jech.55.4.263 Nolte E McKee M Does healthcare save lives? Avoidable mortality revisited 2004 London: The Nuffield Trust Mackenbach JP Health care expenditure and mortality from amenable conditions in the European community Health Policy 1991 19 245 255 10115995 10.1016/0168-8510(91)90011-L Carr-Hill RA Hardman GF Russell IT Variations in avoidable mortality and variations in heaalth care resources Lancet 1987 1 789 792 2882193 10.1016/S0140-6736(87)92810-8 Wolfson M Alvarez R Smith P Towards integrated and coherent health information systems for performance monitoring: the Canadian experience Measuring up: Improving health system performance in OECD countries 2002 Paris: Organisation for Economic Cooperation and Development 133 155 World Health Organization The world health report 2002 Reducing risks, promoting healthy life 2002 Geneva: World Health Organization Arah OA Performance reexamined. Concepts, content and practice of measuring health system performance PhD thesis 2005 Amsterdam: Academic Medical Center, University of Amsterdam Arah OA Westert GP Delnoij DM Klazinga NS Health system outcomes and determinants amenable to Public Health in industrialized countries: a pooled, cross-sectional time series analysis BMC Public Health 2005 5 81 16076396 10.1186/1471-2458-5-81 Ontario Ministry of Health and Long-Term Care Ontario's health system performance report 2004 2004 Ontario: Ontario Ministry of Health and Long-Term Care Statistics Canada Federal, provincial and territorial reports on comparable health indicators Greenberg A Angus H Sullivan T Brown AD Development of a set of strategy-based system-level cancer care performance indicators in Ontario, Canada Int J Qual Health Care 2005 17 107 114 15665069 10.1093/intqhc/mzi007 Morgenstern H Rothman KJ, Greenland S Ecologic studies Modern epidemiology 1998 Philadelphia: Lippincott 459 480 Bland JM Altman DG Statistics notes: Multiple significance tests: the Bonferroni method BMJ 1995 310 170 7833759	16321155	PMC1325226	CC BY	2021-01-04 16:31:49	no	BMC Health Serv Res. 2005 Dec 1; 5:76	utf-8	BMC Health Serv Res	2,005	10.1186/1472-6963-5-76	oa_comm
==== Front BMC Cell BiolBMC Cell Biology1471-2121BioMed Central 1471-2121-6-401631367210.1186/1471-2121-6-40Research articleThe peroxisomal multifunctional protein interacts with cortical microtubules in plant cells Chuong Simon DX [email protected] Nam-Il [email protected] Michelle C [email protected] Robert T [email protected] Douglas G [email protected] Department of Biological Sciences, University of Calgary, 2500 University Dr. NW, Calgary, AB, Canada T2N 1N4, Canada2 Department of Molecular and Cellular Biology, University of Guelph, Guelph, ON, Canada N1G 2W1, Canada2005 28 11 2005 6 40 40 29 7 2005 28 11 2005 Copyright ©2005 Chuong et al; licensee BioMed Central Ltd.2005Chuong 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 The plant peroxisomal multifunctional protein (MFP) possesses up to four enzymatic activities that are involved in catalyzing different reactions of fatty acid β-oxidation in the peroxisome matrix. In addition to these peroxisomal activities, in vitro assays revealed that rice MFP possesses microtubule- and RNA-binding activities suggesting that this protein also has important functions in the cytosol. Results We demonstrate that MFP is an authentic microtubule-binding protein, as it localized to the cortical microtubule array in vivo, in addition to its expected targeting to the peroxisome matrix. MFP does not, however, interact with the three mitotic microtubule arrays. Microtubule co-sedimentation assays of truncated versions of MFP revealed that multiple microtubule-binding domains are present on the MFP polypeptide. This indicates that these regions function together to achieve high-affinity binding of the full-length protein. Real-time imaging of a transiently expressed green fluorescent protein-MFP chimera in living plant cells illustrated that a dynamic, spatial interaction exits between peroxisomes and cortical microtubules as peroxisomes move along actin filaments or oscillate at fixed locations. Conclusion Plant MFP is associated with the cortical microtubule array, in addition to its expected localization in the peroxisome. This observation, coupled with apparent interactions that frequently occur between microtubules and peroxisomes in the cell cortex, supports the hypothesis that MFP is concentrated on microtubules in order to facilitate the regulated import of MFP into peroxisomes. ==== Body Background Peroxisomes are single-membrane-bound organelles that lack a genome and, therefore, must import their entire complement of constituent proteins. All proteins that are targeted to the peroxisome are synthesized on free polyribosomes in the cytosol and are imported post-translationally. Several distinct import pathways exist for membrane and matrix proteins (reviewed in [1]). For example, peroxisomal membrane proteins can be targeted either directly to the peroxisome from their sites of synthesis in the cytosol, or indirectly to peroxisomes via the endoplasmic reticulum ([2,3], and references therein). On the other hand, peroxisomal matrix proteins can be imported from the cytosol in their fully-folded conformation and as oligomeric protein complexes. Two types of peroxisomal matrix protein import pathways have been identified and well characterized [4,5]. Most matrix-destined proteins possess a type 1 peroxisomal targeting sequence (PTS1) that consists of an uncleaved carboxyl-terminal tripeptide sequence (small, basic and hydrophobic amino acids or a variant thereof). The cognate receptor for PTS1-bearing proteins, peroxin 5 (Pex5p), is proposed to carry its protein cargo into the peroxisome matrix as it cycles between the cytosol and the matrix. Alternatively, it may release its cargo after docking with the import machinery located on the peroxisomal surface [6]. In contrast to the PTS1, the type 2 PTS (PTS2) is located near the amino terminus of a smaller set of peroxisomal matrix-destined proteins. In plants and mammals the PTS2 is proteolytically cleaved following import. The receptor protein for PTS2 targeted proteins, Pex7p, is probably best characterized in yeast cells where it is proposed to cycle in and out of peroxisomes, similar to its Pex5p counterpart [7]. Pex7p also relies on Pex5p for the import of PTS2-containing proteins, indicating that the two matrix protein pathways are coupled [8,9]. In all organisms examined to date, peroxisomes are remarkably dynamic in terms of their shapes and intracellular movements [10-15]. Peroxisome morphology ranges from spherical and dumbbell-shaped, to extensively elongated and reticulated. Peroxisome movement and distribution is also highly variable and is mediated by either MTs or actin filaments, depending on the species. For instance, mammalian peroxisomes utilize a MT-based system for movement that is directed by dynein/dynactin and possibly kinesin motors [13]. In plant and yeast cells, actin filaments serve as the tracks for peroxisome movement, and myosin motors have been reported to be responsible for this movement [11,14,16-18]. Examples of actin-based plant peroxisome movements include rapid oscillations at fixed locations, stop-and-go movements in forward and reverse directions, and rapid longer distance movements that achieve velocities of up to 10 μm per second (for review, see [10]). While MT involvement has not been reported for the directed motility of peroxisomes in plant cells, recent evidence indicates that MTs play a role in peroxisome protein import. Specifically, the peroxisomal multifunctional protein (MFP) has been shown to possess MT-binding activity in vitro [19]. MFP is a PTS1-containing peroxisomal matrix protein that possesses up to four enzymatic activities involved in catalyzing different reactions of the fatty acid β-oxidation pathway [20,21]. We proposed that the MT-binding activity of MFP serves to concentrate MFP in order to improve the efficiency of its import into peroxisomes [10]. This would necessitate that peroxisomes frequently be in close proximity to MTs in order for MT-bound MFP to interact with the import machinery located on the peroxisomal surface. Indirect evidence supporting a general role for MTs in peroxisomal protein import has come from a recent large scale proteomic study demonstrating the in vitro tubulin binding activity of five additional plant peroxisomal matrix proteins [22]. Although there is increasing biochemical evidence for MT-binding activities associated with plant peroxisomal matrix proteins, there have been no reports for this interaction in vivo. Here we demonstrate that the peroxisomal MFP is an authentic MT-binding protein that associates specifically with interphase cortical MTs, but not with the MT arrays that form during cell division. We also present deletion analysis data indicating that multiple MT-binding domains are present on the MFP polypeptide. Additionally, live-cell imaging of a GFP-MFP chimera that localized to both peroxisomes and MTs revealed that peroxisomes are frequently in close association with MTs in the cell. This suggests that peroxisomes interact with cortical MTs as they move along actin filaments or oscillate at fixed locations. Overall, the localization of MFP to cortical MTs in situ and the apparent cortical MT-peroxisome interactions support the hypothesis that MTs have an important role in the targeting of MFP to peroxisomes. Results MFP localizes to both peroxisomes and cortical MTs in onion epidermal cells To begin to examine the putative interaction between MFP and MTs in plant cells, an expression construct encoding GFP fused to the amino terminus of MFP (GFP-MFP) was introduced into onion epidermal cells by particle bombardment. Cells expressing the fusion protein displayed a fluorescence pattern that included numerous, small punctate structures (Figure 1A) that were presumed to be peroxisomes. These were similar in appearance to the punctate structures observed in cells expressing a peroxisomal marker protein consisting of GFP appended to a carboxyl-terminal PTS1 ([23], GFP-SRM, Figure 1B). In addition, GFP-MFP transformed cells often displayed less intense, uniformly labeled filamentous structures that were concentrated in the cell cortex and were presumed to correspond to MTs (Figure 1A, inset). In contrast, cells expressing GFP alone displayed only diffuse nuclear and cytosolic fluorescence, as expected (Figure 1C). Figure 1 Expression of a GFP-MFP fusion protein in onion epidermal cells labels peroxisome-like structures and MT-like filaments. Onion epidermal layers were bombarded with DNA constructs encoding either GFP-MFP (A), GFP containing a carboxyl-terminal PTS1 (GFP-SRM) (B), or GFP alone (C), and visualized by epifluorescence microscopy. The inset in (A) shows a representative confocal image of peroxisome-like structures and MT-like filaments in the same optical Z-section in the cortical region of a GFP-MFP-expressing cell. Bars, 20 μm and 3 μm (A, inset). Indirect immunofluorescence and cytoskeleton-depolymerizing drug treatment experiments were conducted to determine if the punctate and filamentous structures identified in cells expressing GFP-MFP (Figure 1A) did indeed correspond to peroxisomes and MTs, respectively. Antibodies raised against the peroxisome matrix marker enzyme catalase were used to label peroxisomes in chemically fixed onion epidermal peels bombarded with the GFP-MFP fusion construct. Figure 2 shows that the punctate structures in GFP-MFP-expressing cells clearly co-localized with endogenous anti-catalase antibody labeling in the same cells, confirming that at least a portion of GFP-MFP was localized to peroxisomes. To confirm also that the filamentous structures labeled by GFP-MFP were MTs, GFP-MFP-transformed cells that displayed strongly fluorescing filaments were treated with reagents that cause cytoskeleton depolymerization. Figure 3A and 3B show that a 45 min treatment with the MT depolymerizing drug oryzalin specifically disrupted the filamentous structures in GFP-MFP expressing cells. As expected, MTs in cells expressing the MT-binding protein GFP-MAP4 [24] were also disrupted by oryzalin (compare Figures 3D and 3E). Conversely, treatments with latrunculin B, a potent inhibitor of actin filament assembly, had no effect on the stability of the filamentous structures in GFP-MFP (or GFP-MAP4) expressing cells (compare Figures 3A and 3C; 3D and 3F). However, latrunculin B treatment did effectively disrupt actin filaments in cells expressing a GFP-talin fusion protein [25] (compare Figures 3G and 3I). These results indicated that the labeled filaments observed in GFP-MFP transformed onion epidermal cells were MTs. Figure 2 GFP-MFP partially co-localizes with the endogenous peroxisomal matrix protein catalase. Onion epidermal cells expressing GFP-MFP were fixed and then probed with monoclonal antibodies raised against catalase. GFP-MFP fluorescence (A), mouse anti-catalase antibody staining (B), and an overlay of the two images (C) are shown in a region of a single transformed cell by epifluorescence microscopy. The acquisition of these images was set at low exposure to capture the brightly fluorescing peroxisomes. The less intensely fluorescing MTs in this GFP-MFP expressing cell were not visible at this exposure. Bar, 4 μm. Figure 3 The filamentous labeling in GFP-MFP-expressing cells is abolished when cells are treated with the MT-disrupting agent oryzalin. Onion epidermal cells expressing GFP-MFP (A, B, C), GFP-MAP4 (D, E, F) or GFP-talin (G, H, I) were treated with DMSO (A, D, G), oryzalin (B, E, H) or latrunculin B (C, F, I) for 45 minutes and visualized by epifluorescence microscopy. Bar, 3 μm. To determine if endogenous onion MFP also interacts with MTs and that the interaction of GFP-MFP to MTs described above was not a consequence of its transient over expression, we performed co-immunolocalization experiments using rabbit anti-MFP and mouse anti-tubulin antibodies. In addition to labeling peroxisomes, anti-MFP antibodies also labeled faint filamentous structures that corresponded to cortical MTs (Figure 4). We also performed immunolocalization experiments using Arabidopsis suspension culture cells in order to determine if MFP binds to the MT arrays that form during plant mitosis. These mitotic MT arrays include the pre-prophase band of MTs, the mitotic spindle, and the cytokinetic phragmoplast. Suspension culture cells are useful for observing these arrays since they divide rapidly during their logarithmic growth phase. Anti-MFP antibody labeling of cortical MTs was evident in non-dividing cells, similar to that observed in onion cells (not shown). However, we did not observe any fluorescence attributable to MFP binding to the mitotic MT arrays, as only peroxisome-specific fluorescence was visible in dividing cells (Figure 4D). Figure 4 Immunolocalization of MFP labels faint filaments that co-localize with cortical MTs, but does not label mitotic MT structures. Indirect immunofluorescence analysis of a fixed onion epidermal cell probed with affinity purified rabbit anti-MFP antibodies (A) and mouse anti-tubulin antibodies (B) and visualized by epifluorescence microscopy. An overlay (C) shows the co-localization of MT labeling by the two antibodies (arrowheads). (D) Fluorescence immunostaining of MT structures with a tubulin antibody in four stages of mitosis in Arabidopsis suspension cells labels the pre-prophase band, the mitotic spindle in metaphase and anaphase, and the phragmoplast in telophase. The MFP antibody labels peroxisomes but not these MT structures in dividing cells. DNA was stained with DAPI (blue). Bars, 6 μm. Multiple MT-binding regions are present on the MFP polypeptide In an effort to identify the region(s) of MFP that is responsible for binding to MTs we generated four truncated versions of recombinant MFP (Figure 5A) and determined their affinity for MTs using MT co-sedimentation assays. Two of the truncated polypeptides, C2 and N1, contained deletions that terminated within an internal region of MFP that is similar in sequence to a MT-binding domain found in several well-characterized MT-associated proteins including MAP2, MAP4, tau and TOGp [26]. The consensus sequence of this domain consists of the amino acid residues, -LX5KI/VGSE/DNK-, and is defined by a characteristic serine phosphorylation tetrapeptide motif (-KXGS-) known to regulate the MT-binding activity of this domain [26]. The amino acid sequence of the potential MT-binding site in MFP (-LEGLVKRGSLTKDK-) begins at amino acid position 356. We were unable to obtain a more detailed deletion analysis of the carboxyl-terminus of the protein (Figure 5A), as repeated efforts to express truncated versions of the protein in this region were not successful. Figure 5 Binding affinity of various truncated MFP polypeptides to MTs. (A) Polypeptide map of wild type MFP and four deletion polypeptides of MFP (C1, C2, N1, N2) used for the expression of recombinant proteins in E. coli and GFP chimeras in onion epidermal cells. Arrow indicates the location of the putative KXGS MT-binding domain. This putative domain is disrupted between the leucine and valine residues (see Results) in the truncated polypeptides C2 and N1. (B) The recombinant proteins were used in MT co-sedimentation assays to generate MT-binding curves. Increasing concentrations of each truncated protein were used (0.25, 0.5, 0.75, 1.0, 2.0, 3.0, 4.0, 7.5, 10.0, 15.0, and 25.0 μM) with a constant amount of MTs (5 μM tubulin). Dissociation constant values (Kd, μM) generated from the data in each of the curves are shown. (C) The subcellular localization pattern of the full-length MFP (GFP-MFP), truncated versions of MFP (C1, C2, N1, N2), and MFP lacking the PTS1 tripeptide (ΔSRM) fused to the carboxyl-terminus of GFP in onion epidermal cells. Faint MTs were occasionally evident in cells expressing the GFP-C1 fusion protein (arrow). Each image is a selected region from an individual cell. Bar, 10 μm. Figure 5B shows that full-length MFP had a high affinity for MTs as indicated by its dissociation constant (Kd) value of 0.7 μM, a value similar to what we determined previously for the same protein (Kd = 0.8 μM, [19]). In contrast, each of the amino- and carboxyl-terminal deletions of the MFP polypeptide (C1, C2, N1 and N2) displayed a significantly reduced affinity for MTs when compared to the full-length polypeptide. None of these deletions, however, resulted the complete elimination of MT-binding activity. The amino-terminal deletion polypeptide, C1, had the highest affinity for MTs (Kd = 6.8 μM, Figure 5B) when compared to the other truncated proteins although its affinity for MTs was 10-fold less than the full-length polypeptide. Polypeptide N2 was composed of the region of MFP that was deleted in polypeptide C1 yet its affinity for MTs was the lowest of all the truncated polypeptides (Kd = 266.5 μM). This indicates that this region alone does not possess significant binding activity, but has an important contribution to the overall binding activity of the full-length MFP. Polypeptides C2 (Kd = 22.2 μM) and N1 (Kd = 46.3 μM), on the other hand, were intermediate in their MT-binding activities despite being disrupted in the putative KXGS MT-binding domain. Overall, this deletion analysis indicates that there are multiple MT-binding regions located on the MFP polypeptide, and that they work synergistically to achieve the high MT-binding affinity observed for full-length MFP. We performed next a series of transient expression experiments with the same truncated versions of MFP by expressing them in onion epidermal cells as fusions to the carboxyl-terminus of GFP (e.g., GFP-C1, Figure 5C). The objective of these experiments was to determine whether their relative MT-binding affinities as determined by MT co-sedimentation assays (Figure 5B) corresponded to their in vivo MT-binding characteristics. For those fusion constructs that lacked portions of the carboxyl-terminus of MFP (i.e., N1 and N2, Figure 5A), the endogenous MFP PTS1 sequence (-SRM) was restored so that these deleted polypeptides would retain their targeting to peroxisomes. All fusion constructs, along with wild-type GFP-MFP, were bombarded individually into onion epidermal cells and observed throughout a 3 to 24 hour period. While cells expressing full-length GFP-MFP often displayed MT labeling, the only truncated MFP fusion protein to label MTs was the amino-terminal deletion construct C1 (GFP-C1; Figure 5C). The number of GFP-C1-transformed cells displaying MT labeling, however, was low, as was the fluorescence intensity of the MTs in these cells. Nevertheless, that GFP-C1 localized to MTs was consistent with the observation that this polypeptide possessed the highest MT-binding affinity of the truncated recombinant proteins (Figure 5B). While cells expressing the N2, N1 and C2 fusion proteins did not display MT fluorescence, it is possible that these proteins bound to MTs in vivo at levels too low to be detected using our imaging system. We hypothesized earlier [10,19] that the MT-binding activity of MFP may serve to concentrate MFP on MTs prior to its import into peroxisomes. Since the carboxyl-terminal PTS1 tripeptide (SRM) of MFP is both necessary and sufficient for import into peroxisomes (R.T. Mullen and D.G. Muench, unpublished observations), we anticipated that deletion of this sequence from MFP in the context of the GFP-MFP chimera would cause an accumulation of the protein in the cytosol and a resulting increase in binding to MTs. Surprisingly, expression of this modified chimera (i.e., GFP-MFPΔSRM) in onion epidermal cells resulted in only a diffuse cytosolic fluorescence with no detectable filamentous structures (Figure 5C). This apparent necessity of the MFP PTS1 for high affinity MT-binding in vivo further demonstrates the complex nature of the MT-binding activity of MFP and the requirement for multiple regions of the protein to achieve efficient binding. Confocal microscopy and real-time imaging indicate that peroxisomes and MTs interact frequently Previous live-cell microscopy studies revealed that plant peroxisomes exhibit high velocity unidirectional movements, bi-directional and stop-and-go movements, as well as rapid oscillations at fixed locations within the cell, and that these movements occurred on actin filaments with no apparent role for MTs (reviewed in [10]). Using confocal microscopy of GFP-MFP-expressing onion epidermal cells we observed that peroxisomes and cortical MTs were regularly located in the same optical section (Figure 1A, inset). This indicated that peroxisomes are frequently in close association with MTs as they move through the cell cortex on actin filaments. In addition, we noted in real-time movies that the characteristic stop-and-go (or "pausing") of peroxisomes frequently occurred at sites occupied by MTs (Figure 6, see Additional data file 1: Movie 1, and Additional data file 2: Movie2), also suggesting that these two structures interact. We determined the frequency of these peroxisome pausing events on MTs in cells that had a relatively sparse cortical MT network such as those observed in Additional data files Movie 1 and Movie 2. The stop-and-go movements of peroxisomes (n = 246) from 14 GFP-MFP-expressing cells were observed in movies of 15 to 20 seconds duration. We determined that 67 ± 13% of the pausing events occurred at sites that were coincident with MTs. There were also numerous peroxisomes that exhibited oscillatory movements at fixed locations. These were observed to be in frequent contact with cortical MTs during their movements 73 ± 19% of the time. Notably, while treatment of GFP-MFP-expressing cells with latrunculin B for one hour caused peroxisomes to oscillate at fixed locations within the cell [16,17,27,28], treatment with oryzalin for one hour did not have an observable effect on the frequency of peroxisome pausing events associated with motile peroxisomes (see Additional data file 3: Movie3). This suggests that MTs are not directly involved in a majority of the peroxisomal pausing events, but that another mechanism is responsible for these pausing events. Figure 6 Interactions are apparent between peroxisomes and MTs in onion epidermal cells. Time-lapse images of a region from a GFP-MFP-expressing onion epidermal cell showing apparent transient and long-term interactions of peroxisomes and MTs. The movements of three peroxisomes were followed over 18 image frames. One of these peroxisomes (star) showed, throughout the entire series, oscillatory movements at a fixed location that were co-incident with a MT. Another peroxisome (arrow) remained fixed at a site coincident with a MT for approximately two seconds at the beginning of the image sequence, and then moved out of the field of view during the final two seconds. The initial movement is marked by the arrow in image frame 1.82. A dumbbell-shaped peroxisome (arrowhead) moved through the cytosol and appeared to transiently tether to a MT (second arrowhead, image frame 1.30). This peroxisome then released and tethered to another MT (third arrowhead, image frame 3.12) before releasing again and continuing its movement. The cell observed in this image sequence was from an unpeeled epidermal layer that remained associated with the leaf segment. The movements of the peroxisomes shown in this figure are also shown in a real-time image sequence in Additional data file Movie 1. Numbers indicate elapsed time in seconds. Bar, 2 μm. Interestingly, we observed also that when epidermal peels were treated with oryzalin or latrunculin B immediately following bombardment of the GFP-MFP fusion construct and throughout the incubation period (see Materials and Methods for details about long-term drug treatments), there was a subtle effect on the rate of GFP-MFP labeling of peroxisomes compared to cells from untreated peels. The average frequency of cells showing peroxisome fluorescence 3 to 4 hours after bombardment in untreated peel segments (normalized value of 0.81 ± 0.08) was higher than in segments treated with oryzalin (0.65 ± 0.11) or latrunculin B (0.74 ± 0.08). These data indicate that the there is a reduction in the rate of synthesis and/or import of GFP-MFP as a result of cytoskeleton disruption. Discussion The processes involved in the import of proteins into the peroxisome matrix are distinct from those described for protein import into other membrane-bound organelles. One major difference is that peroxisomal matrix proteins can be imported fully folded and/or in an oligomerized conformation [6,29,30]. This characteristic implies that these proteins have the potential to possess functional activities in the cytosol prior to their import. We reported previously that the plant peroxisomal MFP binds to MTs and RNA in vitro, in addition to possessing enzymatic activities related to fatty acid β-oxidation. We suggested that the MT-binding activity may function to facilitate the import of MFP into peroxisomes [10,19]. In this paper, we used indirect immunofluorescence microscopy and the expression of a GFP-MFP chimera to show that MFP interacts uniformly with interphase cortical MTs in vivo (Figures 1 and 4), thereby demonstrating that MFP is a bona fide MT-binding protein. The specific binding of MFP to the cortical MTs implies that the MFP-MT interaction is limited to interphase cells. Array-specific binding by other MT-binding proteins has been observed previously [31-33]. These specific interactions may be regulated by post-translational modifications such as phosphorylation/dephosphorylation of the MT-binding protein or modifications to the acidic carboxyl-terminal tail of tubulin [33,34]. Although the techniques used here showed that MFP binds to MTs in vivo, the level of detection of this interaction varied. Indirect immunofluorescence microscopy using an affinity-purified anti-MFP antibody labeled peroxisomes intensely, whereas MTs possessed only weak immunofluorescence and were often not visible at all. Similarly, many transformed cells expressing GFP-MFP lacked observable MT labeling, or MTs in these cells were visible only after computer software enhancement of digitized images. However, in numerous other bombardment experiments MTs were readily observable in up to one-third of the transformed cells. We were unable to determine whether this variation in MT labeling by GFP-MFP was due to differences in the level of expression of the fusion protein, the metabolic state of the cell, or the quality of the leaf material used in individual bombardment experiments. MT co-sedimentation experiments of truncated versions of MFP indicated that there are multiple MT-binding domains within the MFP polypeptide (Figure 5). Each of the truncations resulted in a significant decrease in MT-binding affinity of the remaining polypeptide, indicating that maximal binding of the full-length MFP likely depends on several points of contact to MTs, as has been suggested for other MT-binding proteins [32]. MT-binding domains have been identified for numerous proteins, some of which contain well characterized domains such as those found in the animal MAPs tau, MAP4, MAP2, and TOGp [26], whereas others possess less well known MT-binding domains that are not as well studied [35,36]. We identified by sequence analysis a putative MT-binding domain in MFP that showed sequence similarity to a domain centered around a KXGS motif present singly or in repeats in some animal MAPs [26,33]. However, this region does not contribute significantly to the MT-binding activity of MFP in vitro (Figure 5). We have demonstrated that the Arabidopsis homolog of the rice MFP also possesses MT-binding activity (SDX Chuong and DG Muench, unpublished observations). Despite sharing 82% amino acid similarity to the rice sequence, the Arabidopsis protein does not possess the putative KXGS motif region that is present in the rice MFP. This is consistent with the observed minimal influence of this motif on MT binding affinity in vitro (Figure 5). Interestingly, transient expression of a modified version of GFP-MFP that lacked the MFP carboxyl-terminal PTS1 resulted in this fusion protein (GFP-MFPΔSRM) being localized exclusively to the cytosol, with no apparent labeling of MTs (Figure 5C). This indicates that the carboxyl terminus of MFP is also required for its efficient binding to MTs in vivo. The lack of MT binding by this fusion protein appears to be an effect of expression in vivo since recombinant MFP△SRM binds to MTs in vitro with an affinity that is similar to the full-length MFP protein (SDX Chuong and DG Muench, unpublished results). The PTS1 normally interacts with the peroxisomal import receptor, Pex5p, to facilitate the targeting of PTS1-bearing proteins from the cytosol to the peroxisome surface and possibly across the peroxisome membrane [6,37]. Pex5p is then recycled and is available to interact with other PTS1-containing proteins. The necessity of the PTS1 for efficient binding of MFP to MTs in vivo may, therefore, highlight a novel role for Pex5p at an early stage of peroxisome protein sorting. The immunofluorescence and live-cell imaging data presented here support our previous hypothesis that was based on in vitro MT-binding assays [19] and stated that the interaction between MFP and MTs assists in the efficient and regulated sorting of MFP to peroxisomes [10]. That is, MFP-MT interactions serve to concentrate MFP in the cell cortex, a region of the cell where peroxisomes are abundant and are frequently moving along actin filaments. Using confocal microscopy, we observed here that peroxisomes and cortical MTs regularly occupied the same three-dimensional space in the cell cortex (Figure 1A, inset). This arrangement would provide frequent opportunities for motile peroxisomes to interact with MTs as peroxisomes move in the cell cortex, a premise that is supported by previous evidence for a direct association between cortical MTs and actin filaments [38]. In addition, real-time fluorescence imaging revealed apparent MT interactions with peroxisomes in the form of pausing events by motile peroxisomes as well as peroxisomes that oscillated at fixed locations (Figure 6, Movie Supplement 1). These types of peroxisome-MT interactions would readily allow for the transfer of MT-bound MFP to the import machinery located on the peroxisomal surface. Additional support for a MT role in peroxisomal protein import comes from the discovery that six other plant peroxisomal matrix proteins also possess tubulin-binding activities [22,39]. Most of these proteins have activities involved in catalyzing the reactions of fatty acid β-oxidation in the peroxisome, however, one of them (malate dehydrogenase) is involved in the glyoxylate cycle. β-oxidation enzymes presumably do not have a metabolic role in the cytosol, since their associated metabolic pathways have not been shown to exist in that compartment. It is more likely, therefore, that the binding of these peroxisomal matrix proteins to MTs functions in a general sorting pathway to the peroxisome, as discussed above for MFP. If the cytoskeleton is required for the efficient delivery of MFP to the peroxisome surface as described above, then treatment of cells with cytoskeleton destabilizing agents should affect the import of MFP into peroxisomes. We determined that long-term treatments of epidermal peels with oryzalin and latrunculin B immediately after bombardment of GFP-MFP resulted in a small reduction in the rate of fluorescence accumulation by peroxisomes, indicating that at least subtle roles exist for the cytoskeleton in MFP import. However, since our transient expression system involved an over-expression of GFP-MFP, this may have overridden any requirement for MTs without significantly affecting protein import. We are currently conducting detailed experiments that are beyond the scope of the present study which utilize alternative approaches to determine a potential role of the cytoskeleton on MFP sorting to the peroxisome. Conclusion These data provide conclusive evidence for an interaction between peroxisomal MFP and MTs in plant cells, and revealed that this association is specific to the cortical MT array. Confocal and live cell imaging indicated also that interactions between peroxisomes and MTs occur frequently in the cell cortex. These observations reinforce our hypothesis for a MT role in the sorting of MFP to the peroxisome. Methods GFP expression constructs and biolistic transformation of onion epidermal cells Full-length and truncated versions of a rice peroxisomal MFP cDNA (Genbank accession C26129, [19]) were synthesized by PCR amplification for the production of GFP fusion constructs (Figure 5A) using the following oligonucleotide primers: gMFP1 – CACTCTAGAATGGCGGGGGCGATCCGCGTCACC; gMFP10 – CGTCTAGATCACATGCGTGACCTCGA; gMFP11 – CGTCTAGATCACATGCGTGAAAGACCACCTTCTTTC; gMFP12 – GCGTCTAGATCACATGCGTGAGACTAGACCCTCAAGATTAG; gMFP14 – CACTCTAGAGTTGATGCCCTGTGCTCC; gMFP15 – CACTCTAGAGGCTCATTGACAAAGCAC; gMFP16 – CCATCAGCTGAAATAGCATCTAGAATGCGTTACCTCGACGAAGCCTGCTGG. Primer pairs were used to amplify either full-length MFP (primers gMFP14 and gMFP10), the amino-terminal truncations C1 (primers gMFPl4 and gMFP10) and C2 (primers gMFP15 and gMFP10), the carboxyl-terminal truncations N1 (primers gMFP1 and gMFP11), N2 (primers gMFP1 and gMFP12), or GFP-MFP minus the PTS1 sequence (GFP-MFPΔSRM, primers gMFP1 and gMFP16). Each of the PCR primers contained an XbaI sequence in their 5' flanking region. Digestion of the PCR products with XbaI allowed for cloning into the XbaI site of pRTL2ΔNS/GFP-XbaI [40], resulting in an in-frame fusion of the full-length MFP or truncated MFP open reading frame to the carboxyl-terminus of GFP. pRTL2ΔNS/GFP-XbaI is a plant expression vector containing the cauliflower mosaic virus 35S promoter and nopaline synthase terminator, and a modified GFP with an in-frame XbaI site instead of a stop codon. GFP-talin [25] and GFP-MAP4 [24] expression constructs were used as controls in some transient expression experiments to label actin filaments and MTs, respectively. Transient expression of the various GFP fusion constructs was achieved by biolistic bombardment into onion (Allium cepa) epidermal cells using protocols that were described previously [41]. Plasmid DNA was purified using a DNA Maxiprep kit (Qiagen, Mississauga, ON, Canada) and 5 to 10 μg of the DNA was coated onto gold particles (1 μm diameter) as described by the manufacturer (BioRad, Hercules, CA). DNA-coated gold particles were bombarded into the adaxial surface of onion bulb leaf segments (1 cm2) from a distance of 10 cm using a Biolistic PDS-1000/He particle delivery system (Bio-Rad) at a pressure of 1200 psi. After bombardment, the leaf sections were placed on moist filter paper in petri dishes, adaxial side down, and incubated in the dark at room temperature for 3 to 24 hours. Alternatively, the epidermal layer was peeled immediately after bombardment and floated on liquid MS media (pH 5.9) containing 3% sucrose, and peels were then placed in the dark at room temperature for 3 to 24 hours. Expression and purification of recombinant proteins and MT co-sedimentation assays PCR was used to amplify full-length or truncated versions of the rice MFP coding region for the production of recombinant proteins (Figure 5A) using the following oligonucleotide primers: rMFP1 – CACGGATCCGCGGGGGCGATCCGCGTCACCATG; rMFP10 – CACCTGCAGCATGCGTGACCTCGACGAAGCCTG; rMFP11 – GCGCTGCAGAAGACCACCTTCTTTCCCTTCCTT; rMFP12 – GCGCTGCAGGACTAGACCCTCAAGATTAGCTGC; rMFP14 – CACGGATCCGTTGATGCCCTGTGCTCCCCTGAT; rMFP15 – CACGGATCCAAGAGGGGCTCATTGACAAAGGAC. Primer pairs were used to amplify either full-length MFP (primers rMFP1 and rMFP10), the amino-terminal truncations C1 (primers rMFP14 and rMFP10) and C2 (primers rMFP15 and rMFP10), and the carboxyl-terminal truncations N1 (primers rMFP1 and rMFP11) or N2 (primers rMFP1 and rMFP12). Each primer contained either BamHI or PstI restriction enzyme sites that were compatible for cloning into the expression vector pQE30 (Qiagen). Recombinant proteins were expressed in E. coli and purified as described previously [19]. The purified recombinant proteins were used in bovine MT co-sedimentation assays to generate binding curves for the determination of MT/MFP dissociation constants (Kd) [19]. The co-sedimentation assays were repeated twice for each of the recombinant MFPs, and their mean values and standard deviations were presented. BSA was used as an internal negative control in selected MT co-sedimentation assays to ensure that soluble protein was not trapped by MTs in the pellet fraction. The Kd value for each recombinant protein was determined using the Prism kinetics program (GraphPad Prism, Version 3.02, GraphPad Software Inc., San Diego, CA). Affinity purification of antibodies and immunolocalizations Two mg of the full-length recombinant MFP was covalently coupled to 1 mL of a matrix consisting of an equal mixture of Affigel-15 (Bio-Rad) and Sepharose CL-6B (Sigma-Aldrich Ltd., Oakville, Canada), and the matrix was then poured into a disposable 2 mL column (Poly Prep, BioRad). One mL of MFP antiserum [19] was added to the column matrix, incubated overnight at 4°C, and the matrix was then washed with 5 column volumes of phosphate buffered saline (PBS). IgGs were eluted from the column with 2 mL of 100 mM glycine, pH 2.5, and the eluent was immediately neutralized by the addition of one-tenth volume 1.5 M Tris-HCl pH 8.8. Purified IgGs were concentrated using a Centricon-10 microfiltration apparatus (Amicon, Millipore, Billerica. MA) and stored at 4°C in the presence of 0.05% (w/v) sodium azide. Onion epidermal peels were fixed for 5 min in freshly prepared paraformaldehyde (7.4% [w/v], Sigma-Aldrich Ltd.) and glutaraldehyde (0.015% [w/v], Sigma-Aldrich Ltd.) in a cytoskeleton-stabilizing buffer (50 mM HEPES pH 6.9, 2 mM MgSO4, 5 mM EGTA, and 0.1% Triton X-100)[42]. After fixing, the peels were washed three times in PBS for 5 min each, followed by treatment with 0.5% cellulase (Onazuka R-10) and 0.05% pectolyase Y23 (Seishin Pharmaceutical Co., Tokyo) in MS media (pH 5.9) containing 0.3 M mannitol for 3 minutes at 37°C. The peels were washed three times in PBS for 5 min each and then treated with 1% Triton-X 100 for 5 min at room temperature. Arabidopsis suspension cells were used for immunolocalization studies three days after subculturing. One mL of cell culture was fixed for 15 min at room temperature in freshly prepared 4% paraformaldehyde dissolved in cytoskeleton-stabilizing buffer. After fixing, the cells were washed three times with PBS for 5 min each, and then digested with 0.1% pectolyase Y-23 in PBS for 15 min at 30°C. The cells were washed three times in PBS for 5 min each and then treated with 1% Triton-X 100 for 5 min at room temperature. The fixed and permeabilized epidermal peels and suspension cells were incubated for 15 min at room temperature in blocking solution (PBS containing 3% BSA), followed by incubation for 60 min in blocking buffer containing the appropriate primary antibodies. The following primary antibodies were used: affinity purified rabbit anti-MFP IgGs ([19], final concentration, 0.5 μg/mL); mouse anti-tobacco catalase (undiluted hybridoma medium purchased from the Princeton Monoclonal Antibody Facility); and mouse anti-α-tubulin monoclonal antibody (Sigma-Aldrich Ltd.), 1:500 (v/v). After three, 5 min washes in PBS, the peels/cells were incubated in secondary antibodies diluted 1:500 for 30 min at room temperature, and then washed again in PBS. The secondary antibodies used were goat anti-rabbit IgG-Alexa-594 conjugate and goat anti-mouse IgG-Alexa-488 conjugate (Molecular Probes, Eugene, OR). Pre-incubation of anti-MFP antibodies with recombinant MFP protein prior to using the antibodies in immunolocalization experiments was used to confirm the specific labeling of the anti-MFP antibodies. Microscopic analysis Unpeeled onion leaf segments were secured to a glass slide with 1% agarose, and the epidermal cells were observed through a bead of water or MS media using a 63× long working distance water immersion lens (HCX Apo L, Leica, Germany) mounted on an epifluorescence microscope (Leica DMR). Alternatively, fresh or fixed epidermal peels were mounted on a glass slide in MS media or PBS, covered with a coverglass, and observed using either a Plan Fluotar 40× or Plan Apo 63× oil immersion objective lens (Leica). Images were captured using a cooled CCD camera (Retica 1350 EX, QImaging, Burnaby, BC, Canada) through the FITC or rhodamine filter sets. Image enhancement and deconvolution confocal algorithm manipulations were performed using Openlab software (Version 3.0, Improvision, Lexington, MA). Quicktime movies were generated from sequential images using the Openlab software, and single images were modified using Adobe Photoshop image processing software (Version 8.0, Adobe Systems Inc., San Jose, CA). The frequency of peroxisome pausing events on MTs was determined from 14 GFP-MFP expressing cells, and variation between cells was reported as standard deviation from the mean. Cytoskeleton drug treatments Short-term cytoskeleton drug treatments of 45 min to one hour were conducted on transiently transformed onion epidermal layers that were bombarded with the GFP-MFP construct 3 to 24 hours in advance, and that possessed numerous cells showing high levels of fluorescence. Final drug concentrations were 0.2 μM latrunculin B (Sigma-Aldrich Ltd.) or 2 μM oryzalin (Crescent Chemical Company, Islandia, NY). Stock solutions of these drugs were prepared in DMSO. Epidermal peels were treated with the drugs by floating the peels on liquid MS media (pH 5.9) containing 3% sucrose and the appropriate amount of DMSO or drug. To demonstrate that these drug concentrations were effective in MT and actin filament disruption, epidermal cells expressing the GFP-MAP4 or GFP-talin constructs were treated with the corresponding drugs. For experiments performed to determine the long-term effect of cytoskeleton disrupting drugs on GFP-MFP import into peroxisomes, each bombarded epidermal peel was divided equally into thirds, and the resulting sections were then treated individually with DMSO, oryzalin or latrunculin B immediately after bombardment by floating the peels on liquid MS media containing the appropriate amount of DMSO or drug. The bombarded epidermal peels were observed by epifluorescence microscopy, and the number of cells showing peroxisome labeling were determined 3 to 4 hours post-bombardment, since control sections typically showed the first signs of fluorescence during this period. Data from sections derived from individual peels were pooled for each experiment, and the total dataset presented included results from at least three individual experiments. Normalization of the raw data was performed, since the total number of transformed cells varied between individual bombardment experiments. Variation between experiments was reported as standard deviation from the mean. Authors' contributions SDXC contributed to the experimental design, conducted a majority of the experiments, and assisted in manuscript writing, data analysis, and figure preparation. N-IP and MCF assisted with the immunofluorescence and transient expression experiments. RTM provided expression vectors, assisted in the transient expression experiments, and contributed to the experimental design and writing of the manuscript. DGM wrote the original draft of the manuscript, contributed to the experimental design, and assisted with the immunofluorescence and transient expression experiments, data analysis, and figure preparation. All authors read and approved the final manuscript. Supplementary Material Additional data file 1 Movie1.mov – Real-time image sequence of an onion epidermal cell expressing GFP-MFP shows apparent interactions between peroxisomes and cortical MTs. Epifluorescence images (n = 200) were captured every 0.13 seconds. A mobile, dumbbell-shaped peroxisome (tracked with the green marker) appears to interact with MTs at several locations during the initial part of the image sequence before moving out of focus. Upon returning into the focal plane near the end of the image sequence, this peroxisome again appears to interact with MTs. The movement of this peroxisome is also shown in the fixed image sequence in Figure 6. Another peroxisome demonstrated a pausing event at a location where MTs were absent (tracked with the red marker), although it also paused several times at locations that coincided with MTs. The peroxisome tracked with the blue marker demonstrated on two occasions a subtle, lateral gliding movement alongside MTs. The cell observed in this image sequence was from an unpeeled epidermal layer that remained associated with the leaf segment. Click here for file Additional data file 2 Movie2.mov – Real-time image sequence of an onion epidermal cell expressing GFP-MFP that demonstrated a lower frequency of peroxisome pausing events at sites that were coincident with MT locations. Epifluorescence images (n = 270) were captured every 0.2 seconds. Click here for file Additional data file 3 Movie3.mov – Real-time image sequence of an oryzalin treated onion epidermal cell expressing GFP-MFP. This movie demonstrates that MT depolymerization does not result in an obvious reduction of peroxisome pausing events in cell cortex. Epifluorescence images (n = 150) were collected every 0.2 seconds. Click here for file Acknowledgements We would like to thank Eric Davies (North Carolina State University) for his helpful discussions, Richard Cyr (Penn State University) for the GFP-MAP4 fusion construct, and Nam-Hai Chua (Rockefeller Foundation) for the GFP-talin construct. We are grateful to the Japanese Ministry of Agriculture, Forestry and Fisheries (MAFF) Genome Program for providing us with the rice MFP cDNA clone. This work is supported by Natural Sciences and Engineering Research Council of Canada (NSERC) grants to DGM and RTM. ==== Refs Heiland I Erdmann R Biogenesis of peroxisomes. Topogenesis of the peroxisomal membrane and matrix proteins Febs J 2005 272 10 2362 2372 10.1111/j.1742-4658.2005.04690.x 15885087 Hoepfner D Schildknegt D Braakman I Philippsen P Tabak HF Contribution of the endoplasmic reticulum to peroxisome formation Cell 2005 122 1 85 95 10.1016/j.cell.2005.04.025 16009135 Trelease RN Lingard MJ Robinson D Participation of the plant ER in peroxisomal biogenesis. DG Robinson, ed. The Plant Endoplasmic Reticulum The Plant Endoplasmic Reticulum Heidelberg: Springer-Verlag in press Brown LA Baker A Peroxisome biogenesis and the role of protein import J Cell Mol Med 2003 7 4 388 400 14754507 Eckert JH Erdmann R Peroxisome biogenesis Rev Physiol Biochem Pharmacol 2003 147 75 121 12687401 Subramani S Hitchhiking fads en route to peroxisomes J Cell Biol 2002 156 3 415 417 10.1083/jcb.200112122 11827979 Nair DM Purdue PE Lazarow PB Pex7p translocates in and out of peroxisomes in Saccharomyces cerevisiae J Cell Biol 2004 167 4 599 604 10.1083/jcb.200407119 15545321 Matsumura T Otera H Fujiki Y Disruption of the interaction of the longer isoform of Pex5p, Pex5pL, with Pex7p abolishes peroxisome targeting signal type 2 protein import in mammals. Study with a novel Pex5-impaired Chinese hamster ovary cell mutant J Biol Chem 2000 275 28 21715 21721 10.1074/jbc.M000721200 10767287 Woodward AW Bartel B The Arabidopsis peroxisomal targeting signal type 2 receptor PEX7 is necessary for peroxisome function and dependent on PEX5 Mol Biol Cell 2005 16 2 573 583 15548601 10.1091/mbc.E04-05-0422 Muench DG Mullen RT Peroxisome dynamics in plant cells: a role for the cytoskeleton Plant Science 2003 164 307 315 10.1016/S0168-9452(02)00426-0 Fagarasanu M Fagarasanu A Tam YY Aitchison JD Rachubinski RA Inplp is a peroxisomal membrane protein required for peroxisome inheritance in Saccharomyces cerevisiae J Cell Biol 2005 169 5 765 775 10.1083/jcb.200503083 15928207 Yan M Rayapuram N Subramani S The control of peroxisome number and size during division and proliferation Curr Opin Cell Biol 2005 17 4 376 383 10.1016/j.ceb.2005.06.003 15978793 Schrader M Thiemann M Fahimi HD Peroxisomal motility and interaction with microtubules Microsc Res Tech 2003 61 2 171 178 10.1002/jemt.10326 12740823 Hoepfner D van den Berg M Philippsen P Tabak HF Hettema EH A role for Vpslp, actin, and the Myo2p motor in peroxisome abundance and inheritance in Saccharomyces cerevisiae J Cell Biol 2001 155 979 990 10.1083/jcb.200107028 11733545 Wiemer EA Wenzel T Deerinck TJ Ellisman MH Subramani S Visualization of the peroxisomal compartment in living mammalian cells: dynamic behavior and association with microtubules J Cell Biol 1997 13 71 80 10.1083/jcb.136.1.71 Jedd G Chua NH Visualization of peroxisomes in living plant cells reveals actomyosin-dependent cytoplasmic streaming and peroxisome budding Plant Cell Physiol 2002 43 384 392 10.1093/pcp/pcf045 11978866 Mathur J Mathur N Hulskamp M Simultaneous visualization of peroxisomes and cytoskeletal elements reveals actin and not microtubule-based peroxisome motility in plants Plant Physiol 2002 128 1031 1045 11891258 10.1104/pp.011018 Hashimoto K Igarashi H Mano S Nishimura M Shimmen T Yokota E Peroxisomal localization of a myosin XI isoform in Arabidopsis thaliana Plant Cell Physiol 2005 46 5 782 789 10.1093/pcp/pci085 15792961 Chuong SDX Mullen R Muench DG Identification of a rice RNA- and MT-binding protein as the multifunctional protein (MFP), a peroxisomal enzyme involved in the β-oxidation of fatty acids J Biol Chem 2002 277 2419 2429 10.1074/jbc.M109510200 11706039 Preisig-Müller R Gühnemann-Schäfer K Kindl H Domains of the tetrafunctional protein acting in glyoxysomal fatty acid β-oxidation: demonstration of epimerase and isomerase activities on a peptide lacking hydratase activity J Biol Chem 1994 269 20475 20481 8051146 Mullen RT Baker A, Graham IA Targeting and import of matrix proteins into peroxisomes Plant Peroxisomes: Biochemistry, Cell Biology and Biotechnological Applications 2002 Kluwer Academic Publishers 339 383 Chuong SDX Good AG Taylor GJ Freeman MC Moorhead GB Muench DG Large-scale Identification of Tubulin-binding Proteins Provides Insight on Subcellular Trafficking, Metabolic Channeling, and Signaling in Plant Cells Mol Cell Proteomics 2004 3 10 970 983 10.1074/mcp.M400053-MCP200 15249590 Lee M Mullen R Trelease R Oilseed isocitrate lyases lacking their essential type 1 peroxisomal targeting signal are piggybacked to glyoxysomes Plant Cell 1997 9 185 197 9061950 10.1105/tpc.9.2.185 Marc J Granger C Brincat J Fisher D Kao T-h McCubbin A Cyr RJ A GFP-MAP4 reporter gene for visualizing cortical microtubule rearrangements in living epidermal cells Plant Cell 1998 10 1927 1939 9811799 10.1105/tpc.10.11.1927 Kost B Spielhofer P Chua N-H A GFP-mouse talin fusion protein labels plant actin filaments in vivo and visualizes the actin cytoskeleton in growing pollen tubes Plant J 1998 16 393 401 10.1046/j.1365-313x.1998.00304.x 9881160 Charrasse S Schroeder M Gauthier-Rouviere C Ango F Cassimeris L Gard DL Larroque C The TOGp protein is a new human microtubule-associated protein homologous to the Xenopus XMAP215 J Cell Sci 1998 111 1371 1383 9570755 Collings DA Harper DI Marc J Overall RL Mullen RT Life in the fast lane: actin-based motility of plant peroxisomes Can J Bot 2002 80 1 11 10.1139/b02-036 Mano S Nakamori C Hayashi M Kato A Kondo M Nishimura M Distribution and characterization of peroxisomes in Arabidopsis by visualization with GFP: dynamic morphology and actin-dependent movement Plant Cell Physiol 2002 43 331 341 10.1093/pcp/pcf037 11917088 McNew JA Goodman JM An oligomeric protein is imported into peroxisomes in vivo J Cell Biol 1994 127 5 1245 1257 10.1083/jcb.127.5.1245 7962087 Walton PA Hill PE Subramani S Import of stably folded proteins into peroxisomes Mol Biol Cell 1995 6 6 675 683 7579687 Muller S Smertenko A Wagner V Heinrich M Hussey PJ Hauser MT The plant microtubule-associated protein AtMAP65-3/PLE is essential for cytokinetic phragmoplast function Curr Biol 2004 14 5 412 417 10.1016/j.cub.2004.02.032 15028217 Smertenko AP Chang HY Wagner V Kaloriti D Fenyk S Sonobe S Lloyd C Hauser MT Hussey PJ The Arabidopsis microtubule-associated protein AtMAP65-l: molecular analysis of its microtubule bundling activity Plant Cell 2004 16 8 2035 2047 15273298 10.1105/tpc.104.023937 Cassimeris L Spittle C Regulation of microtubule-associated proteins Int Rev Cytol 2001 210 163 226 11580206 Westermann S Weber K Post-translational modifications regulate microtubule function Nat Rev Mol Cell Biol 2003 4 12 938 947 10.1038/nrm1260 14685172 Moore RC Cyr RJ Association between elongation factor-lalpha and microtubules in vivo is domain dependent and conditional Cell Motil Cytoskeleton 2000 45 4 279 292 10.1002/(SICI)1097-0169(200004)45:4<279::AID-CM4>3.0.CO;2-4 10744861 Wang PJ Huffaker TC Stu2p: A microtubule-binding protein that is an essential component of the yeast spindle pole body J Cell Biol 1997 139 5 1271 1280 10.1083/jcb.139.5.1271 9382872 Purdue PE Lazarow PB Peroxisome biogenesis Annu Rev Cell Dev Biol 2001 17 701 752 10.1146/annurev.cellbio.17.1.701 11687502 Collings DA Asada T Allen NS Shibaoka H Plasma membrane-associated actin in bright yellow 2 tobacco cells. Evidence for interaction with microtubules Plant Physiol 1998 118 3 917 928 9808736 10.1104/pp.118.3.917 Harper JDI Weerakoon ND Gardiner JC Blackman LM Marc J A 75-kDa plant protein isolated by tubulin-affinity chromatography is a peroxisomal matrix enzyme Can J Bot 2002 80 1018 1027 10.1139/b02-088 Hwang YT Pelitire SM Henderson MP Andrews DW Dyer JM Mullen RT Novel Targeting Signals Mediate the Sorting of Different Isoforms of the Tail-Anchored Membrane Protein Cytochrome b5 to Either Endoplasmic Reticulum or Mitochondria Plant Cell 2004 16 11 3002 3019 15486098 10.1105/tpc.104.026039 Scott A Wyatt S Tsou PL Robertson D Allen NS Model system for plant cell biology: GFP imaging in living onion epidermal cells Biotechniques 1999 26 6 1128 1132 Abe S Davies E Isolation of F-actin from pea stems. Evidence from fluorescence microscopy Protoplasma 1991 162 51 61 10.1007/BF01323406	16313672	PMC1325227	CC BY	2021-01-04 18:03:09	no	BMC Cell Biol. 2005 Nov 28; 6:40	utf-8	BMC Cell Biol	2,005	10.1186/1471-2121-6-40	oa_comm
==== Front BMC Plant BiolBMC Plant Biology1471-2229BioMed Central London 1471-2229-5-241628866310.1186/1471-2229-5-24DatabasePscroph, a parasitic plant EST database enriched for parasite associated transcripts Torres Manuel J [email protected] Alexey A [email protected] Natalya [email protected] Russell L [email protected] John I [email protected] Department of Plant Sciences, University of California, Davis, Davis, CA 95616, USA2 Plant Genome Mapping Laboratory, Center for Applied Genetic Technologies, 111 Riverbend Road #224, University of Georgia, Athens, Georgia 30602, USA3 Vavilov Institute of General Genetics, Russian Academy of Sciences, Gubkina Str., 3, Moscow, Russia2005 16 11 2005 5 24 24 28 7 2005 16 11 2005 Copyright © 2005 Torres 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 Parasitic plants in the Orobanchaceae develop invasive root haustoria upon contact with host roots or root factors. The development of haustoria can be visually monitored and is rapid, highly synchronous, and strongly dependent on host factor exposure; therefore it provides a tractable system for studying chemical communications between roots of different plants. Description Triphysaria is a facultative parasitic plant that initiates haustorium development within minutes after contact with host plant roots, root exudates, or purified haustorium-inducing phenolics. In order to identify genes associated with host root identification and early haustorium development, we sequenced suppression subtractive libraries (SSH) enriched for transcripts regulated in Triphysaria roots within five hours of exposure to Arabidopsis roots or the purified haustorium-inducing factor 2,6 dimethoxybenzoquinone. The sequences of over nine thousand ESTs from three SSH libraries and their subsequent assemblies are available at the Pscroph database . The web site also provides BLAST functions and allows keyword searches of functional annotations. Conclusion Libraries prepared from Triphysaria roots treated with host roots or haustorium inducing factors were enriched for transcripts predicted to function in stress responses, electron transport or protein metabolism. In addition to parasitic plant investigations, the Pscroph database provides a useful resource for investigations in rhizosphere interactions, chemical signaling between organisms, and plant development and evolution. ==== Body Background Parasitic plants directly invade and rob nutrients from host plants [1,2]. The consequences can be devastating to the host plant and some of the world's most pernicious agricultural pests are parasitic weeds [3]. The number of parasitic angiosperms is surprisingly large with over four thousand parasitic species identified in nineteen different plant families [4]. Parasitic plants have a wide diversity of growth habits ranging from the tiny flowered mistletoes that live in the tops of trees to the enormously flowered and rootless Rafflesia whose entire vegetative body is endophytic [4]. The degree to which parasites rely on host resources also varies. Some obligate parasites, like Rafflesia, have lost photosynthetic capabilities and are fully heterotrophic. Others, like Triphysaria, are facultative parasites that can mature without a host plant but will parasitize neighboring plants when available. The single feature shared by all parasitic plants is the ability to invade host tissues via a haustorium [1]. Haustoria of parasitic plants fulfill multiple functions including host attachment, penetration, and translocation of resources from host to parasite [5]. Interestingly, the competence to develop haustoria has originated in autotrophic ancestors multiple times during the evolution of angiosperms [6]. There are two general hypotheses for the evolutionary origins of haustoria. One hypothesis suggests that the genes encoding haustorium development are derived from non-plant organisms, such as bacteria or fungi, that are endophytic or which have transferred a set of genes required for haustorium formation into the parasite genome [7]. The second is that genes encoding haustorium development are derived from those present in autotrophic angiosperms where they fulfill functions unrelated to parasitism. The identification of genes associated with haustorium development will provide insights into the evolutionary origins of plant parasitism. These genes will also elucidate the degree to which haustoria in different parasitic families are encoded by convergent or homologous genetic pathways. Parasitic plants in the Orobanchaceae develop haustoria on their roots in response to contact with host roots. Several molecules, typically products of the phenylpropanoid pathway, have been identified that induce haustorium development when applied to Orobanchaceae roots in vitro [5,8-10]. Early haustorium development in response to exogenous signal molecules is characterized by three visible phenotypes: temporary cessation in root elongation, isodiametric cortical swelling, and haustorial hair proliferation [11,12]. Molecular phylogeny places the Orobanchaceae on a single phylogenetic clade of parasites distinct from the nearest non-parasitic relative [13]. This suggests that the genetic mechanisms controlling haustorium development in the Orobanchaceae are likely similar. Triphysaria, formerly Orthocarpus, is an Orobanchaceae that grows as a common, springtime annual throughout the Pacific coast from Canada to Baja [14]. Triphysaria is a small genus of five intercrossing diploid species that are amenable to classical genetic analyses [15]. Triphysaria is closely related to the devastating agricultural weeds Striga and Orobanche; however, Triphysaria itself has no agricultural significance. Triphysaria are facultative parasites that can grow to maturity without host plants but will readily parasitize many host species when available, including Arabidopsis and maize. Triphysaria form haustoria within twelve hours of being exposed to Arabidopsis roots or root factors in vitro [16]. The speed, synchrony, and dependence on exogenous inducer makes haustorium development in Orobanchaceae an excellent system for identifying transcripts associated with subterranean plant-plant communications. Towards the goal of identifying genes associated with plant parasitism, we sequenced cDNA libraries enriched by suppression subtractive hybridization (SSH) [17] for transcripts regulated in Triphysaria roots during haustorium development. To date we have sequenced approximately nine thousand ESTs from three SSH libraries generated after treating Triphysaria roots with either intact Arabidopsis roots or the chemical haustorium inducer 2,6-dimethoxybenzoquinone (DMBQ). DMBQ, first purified as a haustorium inducer from sorghum [9], induces high rates of haustorium development in Triphysaria; however its role in mediating haustorium formation in Triphysaria – Arabidopsis interactions is not known. The Pscroph database provides on-line access to these EST and assembly sequences and provides BLAST and keyword search functions [18]. Comparative analysis with other transcriptomes will highlight genes and pathways associated with the origins of haustorium development and the evolution of heterotrophy in plants. These studies may provide insights into genetic strategies for developing crops resistant to parasitic weeds and into strategies for exploiting allelopathic interactions in agriculture generally. Construction and content Parasite treatments Triphysaria versicolor seeds were collected from thousands of cross-pollinating plants growing in a grassland stand near Napa CA. They were surface sterilized in a solution of 2% sodium hypochlorite (50% household bleach) and 0.01% Triton X-1000, rinsed thoroughly with water and germinated at 16°C in 0.25X Hoagland's solution and 1% agar [19]. After two to three weeks the seedlings were transferred along one edge of a square Petri dish containing the appropriate media and incubated at 25°C at a near vertical angle so that the roots grew down along the surface of the agar. Arabidopsis Columbia seeds were obtained from Lehle seeds (Round Rock, TX, USA), surface sterilized, and germinated in agar. Arabidopsis seedlings were then cultivated hydroponically for 40 days in liquid media (30 seedlings in 30 ml of 0.25X Hoagland's media in 250 ml flasks shaking at 50 rpm at 25°C under a 16 hours light/8 hours dark cycle). Haustorium development was induced five to seven days after the transfer of Triphysaria seedlings to vertical square plates by exposing their roots to Arabidopsis. This was done by placing roots of forty day old hydroponically grown Arabidopsis across the roots of Triphysaria seedlings as shown in figure 1a. Two ml of 0.25X Hoagland's media were applied to the roots to ensure good contact. Contact was maintained for up to five hours during which time Triphysaria roots were harvested for RNA. Haustoria were also induced by applying two ml of 10 μM DMBQ directly to Triphysaria roots for construction of the EDIT library. Early haustorium development could be detected within 24 hr of host parasite contact as swollen, hairy knobs emerging just proximal to the Triphysaria root tips (Figure 1b,c,d). Triphysaria exposed to media without Arabidopsis or DMBQ did not develop haustoria. Library construction Triphysaria roots were dissected and frozen in liquid nitrogen at various time intervals ranging from immediately after treatment to up to five hours later. Triphysaria RNA was prepared as described [20]. Suppressive subtractive hybridization was used to generate cDNA libraries enriched for up or down regulated transcripts using a commercial kit (BD Sciences, Clontech, Mountain View, CA). The Host Forward (HF) library was enriched for transcripts upregulated in Triphysaria roots after contact with Arabidopsis and was made using mock-treated Triphysaria as the hybridization driver. The Host Reverse (HR) library was enriched for transcripts down regulated after host contact and was made using Arabidopsis exposed Triphysaria as driver. The EDIT library was enriched for transcripts upregulated in root tips within five hours of exposure to DMBQ as previously described [20]. EST sequencing and assembly Subtracted cDNAs were ligated into pCR2.1-Topo (Invitrogen, Carlsbad, CA) and cloned in E. coli. Colonies were picked into 384 well microtiter plates and frozen at -80°C. Sequencing templates were prepared from the library using a rolling circle amplification technology with TempliPhi kit (Amersham Biosciences, Piscataway, NJ). Sequencing reactions were run using the BigDye terminator v3.1 (Applied Biosystems, Foster City, CA.) and the products were separated and detected using the ABI3730xl DNA Fragment Analyzer (Applied Biosystems, Foster City, Foster City, CA) at the UC Davis College of Agriculture and Environmental Sciences Genomics Facility. DNA trace files were base-called using Phred (version 0.990722.g) and low quality sequences were removed based on a Phred p value ≤ 0.05 [21]. The sequences were masked for the pCR 2.1-TOPO cloning vector, linker sequences, and repetitive sequences (excluding poly A and poly T) based on alignments generated by the BLASTN program as used by the PyMood Sequence Processor (Allometra, Davis, CA) (Alexander Kozik, pers. comm.). Sequences less than 100 nts were discarded from further analyses. Approximately five percent of the clones had linker sequences internal to an ORF sequence. These were determined to be chimeras generated by the ligation of multiple SSH fragments into a single plasmid. The chimeric sequences were computationally digested into independent ESTs. The finished ESTs were submitted to GenBank's dbEST repository [22]. FASTA files of the finished ESTs were assembled into contigs using the cap3 program. Because we assembled trimmed and masked FASTA sequences, quality files were not included and the cap3 clipping function was unnecessary. The assembly was performed at the default parameters (overlap length cutoff = 30; overlap percent identify cutoff = 75; and overlap similarity score cutoff = 500) [23]. Fifty percent of the assemblies were comprised of a single EST; an additional forty percent were comprised of two, three, or four ESTs. The assembly process identified about 1100 transcript assemblies in the HF library, 1300 in the HR library and 1400 in the EDIT library (Table 1). Database of early haustorial transcripts The EST sequences and assembly alignments are available at the Pscroph database [18]. Data are stored in a MySQL database and made available on the web using a phpMyAdmin interface. The database is housed at the University of California-Davis Genome Center. Proteins predicted to be encoded by the assemblies were annotated from the BLASTX reports comparing Triphysaria sequences to either all proteins in GenBank (rel145.fsa_aa release Dec 15, 2004) or to all predicted proteins in Arabidopsis (ATH1.pep_cm_20040228). These BLAST reports can be accessed at the web site as full text files or by keyword searches of protein annotations. The keyword search function reports the best three hits obtained from GenBank or TAIR databases with e values ≤ 10-8. Each best hit is hyperlinked to the corresponding report page at NCBI or TAIR. The web site also provides a BLAST function that allows homology searches against DNA or protein sequences in each or all libraries. Utility and discussion SSH libraries We previously published a sequence characterization of 246 cDNAs from the EDIT library [20]. At that time we sequenced clones that had been selected by colony hybridizations for transcripts most differentially abundant in the forward hybridization reaction compared to the reverse. This is suggested by the manufacturer to reduce the number of false positives. The colonies not analyzed fell into two, roughly equal sized groups; those that hybridized to both forward and reverse probes and those that hybridized to neither. While this step reduces the number of false positives, it also may eliminate interesting transcripts. In particular, colonies that hybridized with neither probe likely represented weakly expressed, low abundance transcripts [24]. Furthermore, colonies that hybridized to probes from multiple libraries may contain conserved domains in otherwise distinct proteins. Therefore we sequenced additional, unselected clones from the EDIT library and eliminated the hybridization in constructing the HF and HR libraries. The SSH procedure included an Rsa1 digestion step prior to cloning that resulted in bidirectional cloning and in single transcripts being represented by multiple, non-overlapping SSH products. In order to determine the distribution of SSH products relative to the 3' and 5' ends of the encoding gene, we mapped the virtual translations of the SSH ESTs onto the most homologous protein in the plant protein database. The tcl_blast_parser_123_V017 was used to convert BLASTX output data to a table format suitable for manipulation in a spreadsheet [25]. Using the length of the target ORF and the amino acid locations corresponding to the start and stop of the aligned region between the SSH and plant homologs, we estimated the number and length of Triphysaria sequences predicted to be either 3' or 5' non-coding (Table 2, Figure 2). These regions provide good candidate sites for identifying gene specific primers. Depending on the library, from 34 % to 62% of the Triphysaria sequences were predicted to include non-coding sequences; one to ten percent of the cDNAs included both 5' and 3' non-coding sequences (Table 2). There were more 3' than 5' non-coding sequences in all libraries; there were eight times more 3' sequences in the HR library. The 3' non coding regions recovered in the SSH libraries were also longer than those predicted for the 5' (Figure 2). The 3' bias likely results from the initial cDNA synthesis reaction that is primed with poly-T. Depending on the library, between ten and twenty percent of the SSH products had poly-A tracts. The predominance of ORF encoded sequences in the libraries demonstrates that these libraries were less biased towards 3' sequences than would be expected without the Rsa1 digestion. Interlibrary comparisons We used BLASTN to identify nucleotide sequences in common between the different libraries. This is a bioinformatics alternative to colony hybridization to identify interlibrary sequence homologies. Figure 3 is a color representation of the BLASTN results generated by the PyMood software package (Allometra, Davis, CA). The squares represent 3820 assembly sequences arrayed in the order HF, HR and EDIT. BLASTN was performed using the concatenated sequences from the virtual array as target and sequences from each library as query. PyMood parsed the BLAST output and assigned mixes of red, blue and yellow colors to each sequence based on the degree to which the target sequence had homologies in other libraries. The intensity of color was a function of the BLASTN e value and colors were mixed when sequences were present in more than one library. As shown in table 3, about seventy percent of the sequences were specific to a single library. About seven percent of the assemblies were found in both HF and EDIT libraries but not HR; these represent likely candidates for early haustorium development. However similar numbers of sequences were in common between the HF and HR libraries, indicating a basal level of interlibrary redundancy. The number of sequences in common between forward and subtracted libraries is higher than expected if there was no selection for particular cross-hybridizing sequences. If the one thousand sequenced assemblies in each library represent 2% of the approximately 20,000 root transcripts [26], about 0.4% of assemblies would be expected in both libraries by chance alone. In a previously published wet lab characterization of the EDIT library, we reported that about 20% of the clones cross-hybridized to transcripts in both forward and reverse subtracted probes. Other experiments employing SSH procedures report false positive rates of cross-hybridizing clones of 30–50% [27,28]. The Clonetech Selelect PCR users guide states that recovery rates of false positives will vary between tissue types and RNA preparations [29]. The unpredictably high rate of cross-library hybridizing transcripts was not a function of the assembly because a BLASTN analysis of EST sequences before assembly gave similar results (data not shown). Approximately half of the cross-hybridizing sequences had multiple sequence polymorphisms, suggesting these are alleles of co-expressed genes or domains. The levels of expression of cross-hybridizing sequences were estimated from the Arabidopsis MPSS database to determine if they are particularly highly expressed in roots [26]. The Arabidopsis homologs to cross-hybridizing Triphysaria sequences ranged in their root expression between six and over three thousand transcripts per trillion. An ANOVA analysis indicated no significant differences between the predicted expression levels of library specific sequences from false positives present in both forward and subtracted libraries (data not shown). One possible explanation for the unpredictably large number of false positive clones following SSH procedures is miss-priming at the first or second PCR reactions. cDNAs that were not selected during hybridization would be similarly amplified from both libraries if they have sufficient homology to the 22-mers used in the final amplifications prior to cloning. Functional classifications BLASTX was used to assign putative functions to virtual translations of each library specific assembly. Roughly 75–80% of the library specific sequences had homologies in the Arabidopsis protein database at an e value ≤ 10-8. Using the AT number of the best Arabidopsis hits, the putative Triphysaria proteins were placed into functional categories using the Gene Ontology at TAIR [30]. The GO terms obtained for each library are summarized in supplemental table 8. The TAIR output included multiple GO terms for most assemblies so there are more GO descriptors than transcripts. We used chi squared analyses to determine whether different libraries were enriched for certain GO functional categories (supplemental table 8). The frequency of a particular GO term was determined from the total number of GO terms obtained for that library. The relative frequencies of specific GO terms were then compared between libraries. Table 4 summarizes pair wise comparisons between libraries in the proportion of transcripts in each of the GO categories. The most significant functional enrichment was the overrepresentation of transcripts associated with electron transport in the HF library relative to HR (table 4). Electron transport functions were also enriched in the EDIT library relative to HR although at a lower significance (p ≤ 0.05). Correspondingly, transcripts associated with mitochondria were also over-represented in the HF and EDIT libraries relative to the HR library. The over representation of transcripts associated with electron transport is consistent with the model that haustorial inducing factors trigger development through redox mediated mechanisms [31,32]. The HF library was also enriched for transcripts associated with stress responses. This was previously recognized in the EDIT library and is consistent with the long standing hypothesis that parasitic plants recruit defense related genes for host recognition [7,20]. Transcripts associated with the metabolism of nucleic acids and proteins were significantly less abundant in the HF and EDIT compared to the HR libraries. The down regulation of DNA metabolism genes is consistent with the earlier observations that cell division and DNA synthesis is rapidly terminated in Striga upon contact with DMBQ [33]. There were also fewer transcripts predicted to encode protein metabolism functions in the HF and EDIT libraries. While changes in protein profiles have been observed in Striga following DMBQ treatment, the overall reduction in the proportion of transcripts encoding protein metabolism genes was not expected [34,35]. Conclusion Parasitic plants provide an excellent system for studying genetic mechanisms of chemical signaling between plants. In addition, parasitic weeds are among the world's most destructive agricultural pests against which few genetic resistances are available. Genetic suppression of parasite development at early stages in parasitism is a promising approach for engineering resistance against parasitic weeds but requires knowledge of the genetic factors regulating parasite development. The Pscroph database contains parasitic plant transcripts regulated by host encoded factors; these provide potential points for engineering parasite resistance. More generally, the identification of regulatory elements induced by the presence of other plants provides the potential for genetic weed control strategies. Availability and requirements The Pscroph database can be accessed at . Authors' contributions MJT developed the EST and assembly pipeline and the Pscroph database, AAT and NT made the SSH libraries and prepared them for sequencing, RLR mapped the SSH products to ORFs and manages the Pscroph databases, JIY designed and coordinated the project and wrote the manuscript with input and editing from the other authors. All authors read and approved the final manuscript. Supplementary Material Supplementary files All additional files are in Excel format. Footnote to additional files 1-7: Sequences from three libraries were compared by BLASTN to determine which sequences were present in more than one library. The BLASTN cutoff e value was 10-80. Supplemental tables 1-3 contain the library specific sequences and their annotations, tables 4-6 contain sequences present in HF and HR libraries, or HF and EDIT libraries, or HR and EDIT libraries, and table 7 contains sequences present in all three libraries. The annotations shown are parsed from a BLASTX search of the Arabidopsis protein database ATH1.pep_cm_20040228. Footnote to additional file 8: The HF, HR and EDIT specific sequences were annotated with Gene Ontogeny descriptors by cross referencing to the best BLASTX hit in ATH1.pep via TAIR [30]. The first three columns of numbers show the number of times a particular GO descriptor was assigned. Most of the assemblies were annotated with multiple GO descriptors so there are more descriptors than assemblies. A total of 5499 GO descriptors were obtained for the HF library, 7222 GO descriptors from the HR and 8381 descriptors from the EDIT library. Pair-wise comparisons were made between each library for the relative proportion of annotations in a particular GO class. For example, 5499 GO annotations were obtained from the HF assemblies; of these, 666 (12.1%) were annotated as "other physiological processes". Similarly, 931 of the 7222 (12.9%) of the GO annotations obtained from the HR library were "other physiological processes". The resultant Chi-squared value is 1.73, suggesting there are not significant differences in the proportion of assemblies annotated as "other physiological processes" between the HF and HR libraries. In contrast, the relative proportion of "other physiological processes" annotations in the EDIT library is 11.8 % (992 out of 7222 total); this is significantly different from the proportions obtained for the HR library (12.9%) but not the HF (12.1%) at p ≤ 0.05. Numbers in the cells are the Chi-squared values calculated. The yellow shaded cells indicate significant p values ≤ 0.05. Additional file 1 - HF specific assemblies Additional file 2 - HR specific assemblies Additional file 3 - EDIT specific assemblies Additional file 4 - Assemblies in both HF and HR Additional file 5 - Assemblies in both HF and EDIT Additional file 6 - Assemblies in both HR and EDIT Additional file 7 - Assemblies in HF, HR and EDIT Additional file 8 - GO category analysis Acknowledgements The authors recognize and sincerely appreciate the contributions of Marta Matvienko, Alexander Kozik, and Steve Edberg to the development of this database. The database is housed at the University of California-Davis Genome Center. This work was supported by NSF grant # 0236545. M. J. Torres was a recipient of an NIH Biotechnology Training Grant Award. Figures and Tables Figure 1 Photos of Triphysaria haustoria with Arabidopsis host roots. A. Physical contact between parasite and host roots was made by laying Arabidopsis seedlings across the roots of T. versicolor seedlings in vitro. RNA for the host-induced library was isolated from the Triphysaria roots up to five hours after contact with Arabidopsis. B and C. Haustorium development on Triphysaria roots after 36 hours contact with Arabidopsis. C shows a single Triphysaria root forming haustoria on two different Arabidopsis roots. D. Haustorium development after 24 hrs exposure to 30 μM DMBQ. Figure 2 Length of 5' and 3' non-coding sequences. Protein translations of the SSH sequences in each library were mapped by BLASTX to the most homologous plant protein in GenBank. The histogram depicts the number and size of SSH sequences predicted to extend into either the 5' or 3' non-coding regions. Figure 3 Virtual cDNA arrays and clone redundancy in different libraries. A FASTA file containing all HF, HR and EDIT assembly sequences was used as the target in BLASTN comparisons with sequences from each library as query. Each target cDNA (3820 total) was assigned a color based on homology to sequences in different libraries; sequences hybridizing to HF probes were assigned red, those hybridizing to HR probes green, and those with EDIT probes blue. The color intensity reflected the BLASTN score with higher values assigned to greater homology. Colors were mixed when sequences were present in more than one library: those present in both the HF and HR libraries were yellow, in both the HF and EDIT libraries pink, and in the HR and EDIT libraries teal. Assembly sequences with homologies in all three libraries are represented as white. Table 1 EST and contig statistics of HF, HR and EDIT libraries Total ESTs Assembled Contigs Assembled Singlets Total assemblies % with AT annotation1 HF 3386 673 401 1074 82% HR 3428 781 563 1344 80% EDIT 2216 403 999 1402 85% 1 AT annotations were determined from BLASTX comparisons with ATH1.pep_cm_20040228 with a cutoff of e ≤ 10-8 Table 2 Localization of SSH products to gene regions 1 Library Only coding Only 5' non-coding Only 3' non-coding Both 5' and 3' non-coding Total SSH products analyzed 2 HF 1302 (66%) 80 (4%) 566 (29%) 19 (1%) 1967 HR 1212 (54%) 101 (4%) 868 (39%) 51 (2%) 2232 EDIT 721 (38%) 428 (22%) 579 (30%) 171 (9%) 1899 1SSH sequences were mapped to the plant gene identified as the best hit in BLASTX searches as described in the text. The number of produces predicted to contain non-coding sequences (3', 5', or both), or only coding sequences, are shown for each library. 2 Redundant ESTs within each library were identified from the BLASTX reports and removed from the analysis. ESTs lacking BLASTX hits with E values ≤ 10-8 were also eliminated from this analysis. Table 3 Occurrence of transcripts in different libraries 1 Category Total # assemblies # assemblies in category % category specific HF specific 1,074 702 65% HR specific 1,344 910 68% EDIT specific 1,402 1,018 73% HF+HR 2,418 213 9% HF+EDIT 2,476 180 7% HR+EDIT 2,746 218 8% HF+HR+EDIT 3,820 92 2% 1BLASTN (e ≤ 10-20) was used to identify assemblies found in one or more libraries. Table 4 Functional classification of library specific sequences 1 Biological function HF vs HR HF vs EDIT HR vs EDIT cell organization and biogenesis 0.87 0.13 0.43 developmental processes 0.82 0.18 0.31 DNA or RNA metabolism 4.26 * 0.01 4.96 * electron transport or energy pathways 11.24 *** 2.25 4.33 * protein metabolism 8.42 ** 4.35 * 0.99 response to abiotic or biotic stimulus 2.20 1.31 0.17 response to stress 4.87 * 2.19 0.73 signal transduction 0.27 0.03 0.60 transcription 0.21 0.02 0.12 transport 3.15 0.12 2.51 other biological processes 0.37 0.84 2.80 other cellular processes 1.99 0.05 3.31 other metabolic processes 1.05 0.05 1.93 other physiological processes 1.73 0.24 3.99 * biological process unknown 0.07 1.58 2.72 1The proportion of clones in each library assigned a particular GO function were compared between different libraries and chi square used to indicate significance differences in functional representation between libraries (* p ≤ 0.001, p ≤ 0.01,* p ≤ 0.05). ==== Refs Kuijt J The Biology of Parasitic Flowering Plants 1969 Berkeley, University of California Press 246 Press MC Graves JD Parasitic Plants 1995 London, Chapman and Hall 292 Parker C Riches CR Parasitic Weeds of the World: Biology and Control 1993 Wallingford, CAB International 332 Nickrent D Parasitic Plant Connection Riopel JL Timko MP Press MC and Graves JD Haustorial initiation and differentiation Parasitic Plants 1995 London, Chapman and Hall 39 79 Nickrent DL Duff RJ Colwell AE Wolfe AD Young ND Steiner KE dePamphilis CW Soltis DE, Soltis PS and Doyle JJ Molecular phylogenetic and evolutionary studies of parasitic plants Molecular Systematics of Plants II DNA Sequencing 1998 Boston, Kluwer Academic Publishers 211 241 Atsatt PR Parasitic flowering plants: How did they evolve? Amer Nat 1973 107 502 510 10.1086/282853 Steffens JC Lynn DG Kamat VS Riopel JL Molecular specificity of haustorial induction in Agalinis purpurea (L.) Raf. (Scrophulariaceae). Annals of Botany 1982 50 1 7 Chang M Lynn DG The haustorium and the chemistry of host recognition in parasitic angiosperms Journal of Chemical Ecology 1986 12 561 579 10.1007/BF01020572 Albrecht H Yoder JI Phillips DA Flavonoids promote haustoria formation in the root parasite Triphysaria Plant Physiology 1999 119 585 591 9952454 10.1104/pp.119.2.585 Okonkwo SNC Nwoke FIO Initiation, development and stucture of the primary haustorium in Striga gesnerioides (Scrophulariaceae) Annals of Botany 1978 42 455 463 Baird WV Riopel JL Experimental studies of haustorium initiation and early development in Agalinis purpurea (L.) Raf. (Scrophulariaceae). American Journal of Botany 1984 71 803 814 dePamphilis CW Young NN Wolfe AD Evolution of plastid gene rps2 in a lineage of hemiparasitic and holoparasitic plants: Many losses of photosynthesis and complex patterns of rate variation. Proc Natl Acad Sci USA 1997 93 7367 7372 10.1073/pnas.94.14.7367 Hickman JC The Jepson Manual; Higher Plants of California 1993 Berkeley, CA, University of California Press 1400 Yoder JI Self and cross-compatibility in three species of the hemiparasite Triphysaria (Scrophulariaceae) Environ Exp Bot 1998 39 77 83 10.1016/S0098-8472(97)00038-5 Goldwasser Y Westwood JH Yoder JI Somerville C and Meyerowitz E The use of Arabidopsis to study interactions between parasitic angiosperms and their plant hosts The Arabidopsis Book 2002 Rockville, MD, American Society Plant Biologists 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 Pscroph A parasitic plant database Jamison DS Yoder JI Heritable variation in quinone-induced haustorium development in the parasitic plant Triphysaria Plant Physiol 2001 125 1870 1879 11299366 10.1104/pp.125.4.1870 Matvienko M Torres MJ Yoder JI Transcriptional responses in the hemiparasitic plant Triphysaria versicolor to host plant signals Plant Physiol 2001 127 272 282 11553755 10.1104/pp.127.1.272 Ewing B Hillier LD Wendl MC Green P Base-calling of automated sequencer traces using phred. I. Accuracy assessment Genome Research 1998 8 175 185 9521921 Boguski MS Lowe TM Tolstoshev CM dbEST--database for expressed sequence tags Nature Genetics 1993 4 332 333 8401577 10.1038/ng0893-332 Huang XQ Madan A CAP3: A DNA sequence assembly program Genome Res 1999 9 868 877 10508846 10.1101/gr.9.9.868 Florian F Adam K Martin. C Identification of novel putative regulatory genes induced during alfalfa nodule development with a cold-plaque screening procedure. Molecular Plant-Microbe Interactions 1998 11 358 366 9574504 Kozik A Chan B Michelmore R How to parse results of BLAST search (ESTs vs proteins) and set up tables for mySQL database Meyers BC Lee DK Vu TH Tej SS Edberg SB Matvienko M Tindell LD Arabidopsis MPSS. An online resource for quantitative expression analysis Plant Physiol 2004 135 801 813 15173564 10.1104/pp.104.039495 Ji SJ Lu YC Feng JX Wei G Li J Shi YH Fu Q Liu D Luo JC Zhu YX Isolation and analyses of genes preferentially expressed during early cotton fiber development by subtractive PCR and cDNA array Nucl Acids Res 2003 31 2534 2543 12736302 10.1093/nar/gkg358 Zheng J Zhao J Tao Y Wang J Liu Y Fu J Jin Y Gao P Zhang J Bai Y Wang G Isolation and analysis of water stress induced genes in maize seedlings by subtractive PCR and cDNA macroarray Plant Molecular Biology 2004 55 807 823 15604718 Clontech PCR-Select Differential Screening Kit, Users manual PT 3138-1 TAIR Gene Ontology at TAIR: The Arabidopsis Information Resource Smith CE Ruttledge T Zeng Z O'Malley RC Lynn DG A mechanism for inducing plant development- the genesis of a specific inhibitor Proceedings National Academy of Sciences USA 1996 93 6986 6991 10.1073/pnas.93.14.6986 Matvienko M Wojtowicz A Wrobel R Jamison D Goldwasser Y Yoder JI Quinone oxidoreductase message levels are differentially regulated in parasitic and non-parasitic plants exposed to allelopathic quinones Plant Journal 2001 25 375 387 11260494 10.1046/j.1365-313x.2001.00971.x Keyes WJ O'Malley RC Kim D Lynn DG Signaling organogenesis in parasitic angiosperms: Xenognosin generation, perception, and response J Plant Growth Regul 2000 V19 217 231 Wolf SJ Timko MP Analysis of in vivo protein sythesis and histological studies of haustorial formation in root cultures of witchweed (Striga asiatica L. Kuntze). J Exp Bot 1992 43 1339 1348 Stranger A Murphy A Corbett JM Dunn MJ Bolwell GP Stewart GR Changes in patterns of protein synthesis during haustorial development of Striga hermonthica (Del.) Benth. seedlings Journal of Experimental Botany 1995 46 277 283	16288663	PMC1325228	CC BY	2021-01-04 16:35:47	no	BMC Plant Biol. 2005 Nov 16; 5:24	utf-8	BMC Plant Biol	2,005	10.1186/1471-2229-5-24	oa_comm
==== Front Immunome ResImmunome Research1745-7580BioMed Central London 1745-7580-1-61635171210.1186/1745-7580-1-6ResearchPrediction of MHC class II binding peptides based on an iterative learning model Murugan Naveen [email protected] Yang [email protected] Department of Bioengineering (MC063), University of Illinois at Chicago, 851 South Morgan Street, Chicago, IL 60607, USA2005 13 12 2005 1 6 6 25 10 2005 13 12 2005 Copyright © 2005 Murugan and Dai; licensee BioMed Central Ltd.2005Murugan and Dai; 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 Prediction of the binding ability of antigen peptides to major histocompatibility complex (MHC) class II molecules is important in vaccine development. The variable length of each binding peptide complicates this prediction. Motivated by a text mining model designed for building a classifier from labeled and unlabeled examples, we have developed an iterative supervised learning model for the prediction of MHC class II binding peptides. Results A linear programming (LP) model was employed for the learning task at each iteration, since it is fast and can re-optimize the previous classifier when the training sets are altered. The performance of the new model has been evaluated with benchmark datasets. The outcome demonstrates that the model achieves an accuracy of prediction that is competitive compared to the advanced predictors (the Gibbs sampler and TEPITOPE). The average areas under the ROC curve obtained from one variant of our model are 0.753 and 0.715 for the original and homology reduced benchmark sets, respectively. The corresponding values are respectively 0.744 and 0.673 for the Gibbs sampler and 0.702 and 0.667 for TEPITOPE. Conclusion The iterative learning procedure appears to be effective in prediction of MHC class II binders. It offers an alternative approach to this important predictionproblem. ==== Body Background Immune responses are regulated and initiated by major histocompatibility complex (MHC) molecules, which bind to short peptides from antigens and display them on the cell surface for the recognition by T cell receptors. The specificity of this binding can be predicted from the amino acid sequence of a peptide. Such predictions can be used to select epitopes for use in rational vaccine design and to increase the understanding of roles of the immune system in infectious diseases, autoimmune diseases, and cancers. There are two types of MHC molecules, class I and class II, and both are highly polymorphic. The core binding subsequence of both MHC I and II is approximately 9 amino acids long. However, the MHC I molecules rarely bind peptides much longer than 9 amino acids, while MHC II molecules can accommodate longer peptides of 10–30 residues [1-3]. The presence of the binding core with a uniform length for MHC I molecules makes the prediction of peptide-MHC binding relatively easier. Many different methods have been developed for the prediction of peptide-MHC binding, including simple binding motifs, quantitative matrices, hidden Markov models, and artificial neural networks [4-8]. These methods can be readily applied to MHC I molecules, since the binding motif is well characterized and most of the natural peptides that bind MHC I molecules are of close to equal length. The prediction of MHC class II binding peptides is a difficult classification problem. MHC class II molecules bind peptides that are 10–30 amino acids long with a core region of 13 amino acids containing a primary and secondary anchor residues [2,9,6,11]. Analysis of binding motifs has suggested that a core of only 9 amino acids within a peptide is essential for peptide-MHC binding. Reported binding peptides usually have variable lengths and an undetermined core region for each peptide. Therefore, a search for the binding core region can circumvent the problem of variable lengths. Efforts have been focused on how to align the peptides such that a block of the peptides can be identified as the binding cores. The alignment of peptides is searched based on evolutionary algorithms [12], the Gibbs sampling method [13], and a recent method motivated by the ant colony search strategy [14]. The former looks for a position scoring matrix with the highest fitness score (predictive power) through the genetic operator of mutation. The latter two methods attempt to find an optimal local alignment by means of Monte Carlo Metropolis sampling in the alignment space or by the collective search strategy of ant colony systems, respectively. The binding cores with same length are identified from the alignment, and a scoring matrix used for prediction is established from these binding cores. In the work of Brusic et al. [12], the alignment of peptides is treated as a pre-processing procedure. Upon the determination of the binding cores, a binary classifier is then learned with artificial neural networks using amino acid sequences presented in the binding core as a positive training set and other non-binding peptides as a negative training set. In Nielsen et al. [13] and Karpenko et al. [14], a position scoring matrix is obtained from the best alignment and used for scoring peptides. Most of these alignment-based predictors have achieved reasonably good performances. However, a common complication involved in these methods is the correct choice of associated parameters. The tuning of the parameters could be complicate. A similar work is by Bhasin et al. who used a pre-processing procedure called MOTs to filter the putative binding core for binding peptide sequences and subsequently trained the classifier based on the support vector machine (SVM) [15] with those binding core sequences and random sequences [16]. Another method using an iterative approach has been developed based on a stepwise discriminant analysis model [17,18]. More recently a model based on Bayesian neural networks has been developed [19]. This work is motivated by a machine learning model designed for a training task with only positive and unlabeled examples in text mining. This type of training set is in evidence in various applications in which the identification of a positive training example is labor intensive and time consuming. The basic idea developed for this learning task is the use of a binary classifier to filter out positive examples from the unlabeled set and include them into the positive set through an iterative procedure [20,21]. A classifier is trained at each iteration by simply assigning positive examples the label 1 and unlabeled examples the label -1 to form normal binary training sets. A classifier can be learned by the use of different binary classification methods such as the Naïve Bayesian or support vector machines. The unlabeled and labeled examples in the prediction of peptide-MHC binding can be introduced naturally through the encoding mechanism. A sliding window scheme with a window length of 9 is applied to binding peptides. This procedure breaks a peptide into a set of nonamers of equal length. The binding core, which is unknown, is one of the nonamers. The nonamers from all the binding peptides serve as unlabeled examples in which the positive examples, i.e., nonamers of binding cores, are included. Similarly, all nonamers obtained from the non-binding peptides serve as negative examples. It is noted that the situation in this application is opposite to that of text mining. Here a negative set and an unlabeled set containing potential positive examples are presented. However, the same strategy described previously for text mining can be applied. The approach here is to filter out non-binding nonamers in the unlabeled set iteratively. This iterative learning model enables the use of the non-binder information for the identification of the binding cores and to generate the predictor simultaneously. This is different from the three alignment based methods mentioned earlier in which the identification of binding cores relied only on binding peptides. The linear programming (LP) model proposed by Bennett and Mangasarian [22] is used as the learning model for binary classification at each iteration. This model has several advantages over other learning methods such as support vector machines, Naïve Bayesian, and artificial neural networks. First, there are only a few parameters and they are very easy to tune. Second, a linear program can be solved very fast and it embodies favorable properties which allow sensitivity analysis. Therefore, if the subsequent linear program is only different for a small number of constraints, then the corresponding optimal solution can be found through a re-optimization procedure that uses the information of the current optimal solution. This is particularly important for the iterative learning procedure as only a small number of nonamers is removed from the positive training set at each iteration. This model was evaluated with benchmark datasets from MHCBench against other major existing methods. The computational study demonstrates overall that this method can achieve comparable or superior performance in comparison with the competing predictors, such as the Gibbs sampler [13] and TEPITOPE [10]. The average areas under the ROC (Receiver Operating Characteristic) curve [23] obtained from one variant of our model are 0.753 and 0.715 for the original and homology reduced benchmark sets, respectively. The corresponding values are 0.744 and 0.673 for the Gibbs sampler and 0.702 and 0.667 for TEPITOPE. Methods LP model for classification Consider a set of positive examples xi, i = 1,...,m+ and a set of negative examples xi, i = 1,...,m-, each of which is a point in an n-dimensional space. The LP model for a binary classification problem in (Bennett et al., 1992) is as follows. where yi = 1 or -1 is the label assigned to each positive or negative example, respectively. This model generates a separating hyperplane with the smallest amount of misclassification error. It has been proved that this linear program always returns a non-trivial solution of w, which permits a linear classification function, even in a non-linear separable case [22]. The decision function, denoted by f(x) = wTx+b, assigns a label to an example x by the sign of f(x). LP model for MHC class II problem A set of nonamers can be obtained by sliding a window of length 9 along each MHC class II binding peptide as described previously. A peptide of length s will have s - 8 nonamers (see top panel, Figure 1). These nonamers are further reduced to a set of putative nonamers based on the knowledge that the residue in the first position of the nonamer has to be hydrophobic in order for it to bind to an HLA-DR MHC II molecule. This set of putative nonamers is considered as an unlabeled set. Each nonamer in the unlabeled set is assigned the label 1 temporarily. Figure 1 Top: A peptide has been reduced to a set of nonamers. Bottom: A nonamer is encoded as a 180-dimensional vector. The negative set of nonamers can be obtained similarly from the non-binding peptides. Each nonamer in this set is assigned the label -1. All redundant nonamers in both sets are removed. The remaining nonamers are subject to further preprocessing steps, which will be described later. An amino acid at each position of the nonamer can be encoded by a 20-dimensional vector. Each coordinate of the vector is either 1 or 0, representing the presence or the absence of a particular amino acid. Accordingly, each nonamer can then be represented as a 180-dimensional vector (see bottom panel, Figure 1). Assume that there are m+ binding and m- non-binding peptides. Each encoding vector of a nonamer for a peptide i is denoted by . Assume that each binder i permits ik putative nonamers. By using the LP model given before, our problem can be formulated as the following linear program: where C1 and C2 are coefficients that will be determined through cross-validation on the training set. Note that we have extended the LP model (1) by allowing the change of coefficients C1 and C2 associated with the error terms in the objective function in LP (1). This extension aims at the control of the weights on error terms so that some non-core nonamers in the positive set are deliberately misclassified. This is a chief characteristic of our learning model. Iterative procedure The iterative training process consists of the following major steps. First, a weight vector w and value b are obtained by solving the LP (2) for fixed C1 and C2. This solution is used to score each nonamer in the positive training set based on the function f(x) = wTx+b. Nonamers with negative scores from the positive set are moved to the negative set. Subsequently, the LP is solved for the altered training sets. This process is repeated for a number of iterations, which will be determined through cross-validation (CV). The function f(x)defined with the final LP solution w, and b is used for the prediction of peptides in the testing set. A peptide that has at least one positively scored nonamer is considered as a binder; and otherwise it is considered a non-binder. If several nonamers from one peptide have a positive score, then the nonamer with the highest score is considered as the binding core for that peptide. Note that there may be no binding core identified for certain binding peptides in the final positive training set. In addition to the learning model described above, two other variants were considered. In the first variation, the nonamers in the positive set evaluated with a negative score were discarded instead of being appended to the negative set at each iteration, as these nonamers may not necessarily be true non-binders. In the second variation, at most two nonamers with the highest positive scores from each peptide were allowed to remain in the positive set and the remaining was discarded. The approach in this variant of the LP is motivated by the observation that the binding core is likely to occur among the high scoring nonamers for each peptide. (From our preliminary study on peptides from the training set with known binding core regions, it was observed that there was no significant improvement in performance from using the top three or four nonamers over the top two nonamers.) These variants of the LP method are referred to as LP_append, LP_discard, and LP_top2 in the discussions below. For LP_append, LP_discard, the number of iterations for which the LP process is repeated and the coefficients C1 and C2 were determined by a 5-fold CV on the training set. For LP_top2, the CV procedure only determines the coefficients C1 and C2, since LP_top2 terminates after the second iteration. The area under the ROC curve was the criterion for the evaluation of predictors. The final predictor for each method was obtained by training the whole training set with the optimal parameters determined from the 5-fold CV. The linear programming package GLPK [24] was used to solve the LP given (2). Data sets Training data sets for HLA-DR4 (B10401) allele The sequences of peptides binding to the MHC class II molecule HLA-DR4 (B10401) from the SYFPEITHI [6] and MHCPEP [12] databases were extracted. Since the SYFPEITHI database has more peptides now than in 1999, peptides sequences added to the database after year 1999 were eliminated to make it comparable to the dataset used in Nielsen et al. [13]. This set consists of 462 unique binding peptide sequences. Non-binders for the MHC class II molecule HLA-DR4 (B10401) were extracted from the MHCBN database [25]. This set consists of 177 unique non-binding peptides sequences. The binding peptides that do not possess a hydrophobic residue (I, L, M, F, W, Y, V) at the first position in putative binding cores were removed [12]. That is, a peptide was removed if no hydrophobic residues are present at the first n-s+1 positions, where n is the peptide length and s is the length of the sliding window. The hydrophobic filter removed 27 peptides. Furthermore, the set was reduced by removing unnatural peptide sequences with an extreme amino acid content of more than 75% alanine. Thus, the pre-processing procedure gives 462 unique binding peptides and 177 unique non-binding peptides. The length distribution in the training set ranges from 9 to 30 residues, with the majority of peptides having a length of 13 amino acids. These peptide sequences were then used to obtain nonamers with the sliding window scheme described earlier. All redundant nonamers and nonamers that do not have a hydrophobic residue at position 1 were removed. The final numbers of nonamers obtained from the binding and non-binding peptides are 796 and 903, respectively. Testing data sets for HLA-DR4 (B10401) allele Ten benchmark datasets used in Nielsen [13] were considered in our study. These 10 datasets consist of the 8 datasets described in MHCBench [26] and 2 datasets described by Southwood [27] and Geluk [28]. The same procedure presented in Nielsen et al for the determination of binders and non-binders was followed in our study. More specifically, for the 8 MHCBench datasets, peptides with an associated binding value of zero were considered be non-binders, and all other peptides were binders. For the datasets of Southwood and Geluk, an affinity of 1000 nM was taken as the threshold for peptide binding [27]. In order to reduce the chance of over-prediction, the benchmarking was also performed on the homology-reduced datasets. The homology reduction was carried out so that no peptide in the evaluation sets had a match in the training set with sequence identity >90% over an alignment length of at least nine amino acids. Table 1 shows a summary of the original and the homology-reduced benchmark datasets, respectively. Note that there is small discrepancy in the numbers in some of the reduced sets compared with the ones reported in Nielsen [13] (From the email communication with Dr. Nielsen, there was an error in reporting the numbers in the table in their paper; however, the results on prediction presented there were based on the numbers shown in Table 1). Table 1 Description of HLA-DR4 (B10401) benchmark datasets. Set Original Dataset Homology Reduced Dataset Total Binders Non Binders Total Binders Non Binders Set 1 1017 694 323 531 248 283 Set 2 673 381 292 416 161 255 Set 3a 590 373 217 355 151 204 Set 3b 495 279 216 325 128 197 Set 4a 646 323 323 403 120 283 Set 4b 584 292 292 375 120 255 Set 5a 117 70 47 110 65 45 Set 5b 85 48 37 84 47 37 Southwood 105 22 83 99 19 80 Geluk 22 16 6 21 15 6 Data sets of HLA-DRB10101 and HLA-DRB10301 for cross-validation test Two other datasets for the MHC class II molecules HLA-DRB10101 and HLA-DRB10301 were obtained from the MHCBN database [25]. The dataset for HLA-DRB10101 consists of 475 binder and 105 non-binder peptides. The dataset for HLA-DRB10301 consists of 219 binder and 150 non-binder peptides. The same pre-processing procedure described earlier was applied to these two sets. Results Testing of the benchmark data for HLA-DR4(B10401) The results of the three methods on the benchmark datasets are compared with those obtained from the Gibbs sampling technique [13] and TEPITOPE [10]. The results of the Gibbs sampler were calculated with the scoring matrix provided by Dr. Nielsen; and the results of TEPITOPE were obtained with the use of the scoring matrix from ProPred [29], which is based on the one from TEPITOPE. The performance, evaluated by the area under the ROC curve (Aroc), of each method on the 10 benchmark datasets is presented in Figure 2 and Figure 3. Table 2 gives the performance of the methods averaged over the 10 benchmark datasets. It is observed that among the three proposed methods, LP_top2 has a slightly higher average Aroc value than those obtained from the other two variants. It is also observed that all the three LP variants have higher Aroc values compared to the Gibbs sampler and TEPITOPE. Figure 2 Prediction accuracy of the various methods on the original benchmark datasets. Figure 3 Prediction accuracy of the various methods on the homology reduced datasets. Table 2 The average Aroc values for different methods. Method Average Aroc values for the benchmark datasets Original Homology Reduced LP_append 0.749 0.698 LP_discard 0.748 0.699 LP_top2 0.753 0.715 Gibbs method 0.744 0.673 TEPITOPE (Propred) 0.702 0.667 A notable observation is that the performance of the Gibbs sampler appears to deteriorate for set 5A (0.588) and set 5B (0.600), whereas the LP methods maintain the performance for those two datasets, e.g., LP_top2 has Aroc values of 0.725 and 0.760 for the original benchmark sets 5a and 5b respectively. These two datasets have higher cysteine content as compared to the training set. However, since the LP methods use both binders and non-binders to train the classifier (unlike most of the other methods in which only binders are used for training), the LP methods are more robust in performance. In addition, upon testing the method by substituting all occurrences of cysteine in all the sets by alanine [13], it was observed that the LP_top2 method obtained Aroc values of 0.815 and 0.859 for the original benchmark sets 5a and 5b, respectively; while the Gibbs sampler obtained Aroc values of 0.621 and 0.661, respectively. The details of the results are provided in Tables S1, S2, S3 and S4 in the supplementary document (see Additional files add1.doc – add4.doc). Also TEPITOPE had a poor performance for the Southwood dataset (Aroc values 0.703 and 0.630 for the original and homology datasets) due to a biased amino acid composition at position P1 and that if a modified TEPITOPE matrix at the P1 position was used, TEPITOPE could increase Aroc values to 0.786 and 0.794 for the original and homology reduced Southwood datasets, respectively [13]. For the other benchmark datasets, the performance of the modified TEPITOPE is similar to the original TEPITOPE matrix. In order to investigate the statistical significance of the results, 1000 datasets were generated by random sampling N data points with replacement for each dataset. Here, N is the number of data points in the original dataset. The performance of the different methods was evaluated for each of the original and homology reduced datasets. It was observed that among the LP variants, LP_top2 had a slightly improved performance when compared to LP_append and LP_discard methods. However there was no significant difference observed in their performance. The overall average performance of the methods for the sampled datasets was also not very different from that for the original and homology reduced datasets. The details are provided in Tables S1 and S2 in the supplementary document (see Additional files add5.doc – add6.doc) For comparison with the Gibbs sampler, the p-value for the hypothesis that the Gibbs method performs better than the LP method is estimated as the fraction of experiments where the Gibbs sampler has a better performance. LP_top2 was selected in this comparison. It was observed that for the original benchmark datasets, for 7 of the 10 datasets (sets 1, 2, 3a, 3b, 4a, 5a and 5b), LP_top2 performed better than the Gibbs sampling method (p < 0.05). For the remaining 3 datasets, there was no significant difference in performance (0.05 <p < 0.95). In case of the homology reduced datasets, for 8 of the 10 datasets (sets 1, 2, 3a, 3b, 4a, 4b, 5a and 5b), LP_top2 performed better than the Gibbs sampling method (p < 0.05). For the remaining 2 datasets there was no significant difference in performance (0.05 <p < 0.95). The same comparison was made between LP_top2 and TEPITOPE. It was observed that for the original benchmark datasets, for 2 of the 10 datasets (sets 5b and Southwood), LP_top2 performed better than the TEPITOPE sampling method (p < 0.05). For the remaining 8 datasets, there was no significant difference in performance (0.05 <p < 0.95). In case of the homology reduced datasets, for 7 of the 10 datasets (sets 1, 2, 3a, 3b, 4a, 5a and Southwood), LP_top2 performed better than TEPITOPE (p < 0.05). For the remaining 3 datasets there was no significant difference in performance (0.05 <p < 0.95). Details are given in Table S3 in the supplementary document (see Additional files add7.doc). Results for cross-validation The LP method (LP_top2) was also evaluated using a 5-fold cross-validation for the datasets of HLA-DRB10101 and HLA-DRB10301. The results were compared against those obtained from TEPITOPE (see Table 3). The TEPITOPE matrix was downloaded from ProPred [29], and was used on the testing folds. The LP method produced Aroc values 0.779 for HLA-DRB10101 data set and 0.721 for HLA-DRB10301 dataset. The corresponding values generated from TEPITOPE are 0.842 and 0.585, respectively. The LP method appears to be more consistent in performance over different alleles. Table 3 The average Aroc values from 5-fold cross validations. Method HLA-DRB10101 HLA-DRB10301 LP_top2 0.779 0.721 TEPITOPE (Propred) 0.842 0.585 Prediction of binding core The predictive ability of the LP method (LP_append) for the identification of binding cores in binding peptides was assessed for the HLA-DR4 (B10401) allele. The 68 peptide sequences which have information on experimentally determined binding cores, contained in the SYFPEITHI database were used for the verification. Nonamers in the initial set of putative binding cores for the HLA-DR4(B10401) allele that are identical to any binding cores in the 68 binding peptides were removed. It resulted in a new training set of 755 binding nonamers. The same negative nonamer set for the HLA-DR4(B10401) allele was used. The classifier was trained with the use of the previously described procedure. Among the 68 binding peptides, Fifty one binders which produce distinct binding cores were selected from the 68 binders. However, 6 of those had cores with a length less than 9 amino acids. After the removal of these exceptions, 45 peptides were left for the testing. The predicted binding core is considered as the nonamer with the highest score. The numbers of identified binding cores that were within two positions from the exact binding core by the LP method, TEPITOPE, and the Gibbs sampler are respectively 41, 43, and 42. That is, each identified binding core shares at least 7 consecutive residues with the reported cores. The reason for verifying the predicted core with a shift of a few positions of the reported binding core is because that the binding affinity is not completely determined by the binding core and the flanking amino acids on both sides of the real core may contribute to the binding affinity and stability [19,30,31]. It should be noted that the Nielsen matrix used was obtained from the original training set, which includes those 68 binders. It appears that the three methods performed almost the same. The core alignment of 11 peptides out of the 45 testing peptides obtained from the LP method and the original core alignment from the SYFPEITHI database are presented in Figure 4. Figure 4 Top: The alignment of actual binding cores (shadowed) from SYFPEITHI database. Bottom: The alignment of the predicted binding cores by the LP method. Discussion It is important to note that the Gibbs sampler involves a set of parameters that need to be optimized using a complicated procedure before the training, whereas the LP method is very simple and the only parameters that need to be determined are the coefficients for the misclassification errors and the number of iterations. Both these parameters are easily and very quickly determined through cross-validation. This process involves no modification when applied to peptide sequences respect to MHC alleles. A similar iterative approach for predicting HLA DR1 alleles using a stepwise discriminant analysis (SDA) has been reported [17,18]. This approach trains a linear discriminant function at each iteration and uses it to evaluate nonamers obtained from the original binding peptide sequences. Those nonamers passed the prediction threshold forms the positive training set in the next iteration. Therefore, the positive training set is dynamically changing over iteration to iteration. The negative training set remains unchanged. In this sense, Mallios' method is similar to our LP_discard or LP_top2. The discriminative features are selected based on the F-statistic from a one-way analysis of variance. The Mallios model is essentially a multiple linear regression which minimizes the sum of squared errors, while our model minimizes a weighted sum of errors. In a recent work, a Bayesian neural network [19] was used for the prediction of MHC class II peptide binding. They concluded that their method outperforms the neural network model [12] and the SVM model [16]. Since their datasets were not available, a direct comparison could not be performed. Conclusion An iterative supervised learning model has been developed for the prediction of peptide binding to MHC class II molecules. This approach was motivated by a model for building a classifier with the positive and unlabeled training sets in text mining. The major feature of this method is its iterative extracting of binding core nonamers. The iterative training process functions like an 'adaptive loop', feeding back useful information by validating against the training data. The results indicate that the performance of the new method for HLA-DR4 (B10401) allele is competitive to other methods. Furthermore, the method can incorporate new peptides into the training data easily. This feature makes the method far more adaptive. It is expected that the predictive accuracy will be improved, if the information on other key anchor positions is incorporated [13] and a support vector machine learning model is adapted. Supplementary Material Additional File 1 This file includes Table S1 – The average of Aroc values and standard deviation for the 1000 random sampling datasets on the original benchmark datasets. Click here for file Additional File 2 This file includes Table S2 – The average of Aroc value and standard deviations for the 1000 random sampling datasets on the homology reduced benchmark datasets. Click here for file Additional File 3 This file includes Table S3 – P values for the statistical tests. Click here for file Additional File 4 This file includes Table S4 – The Aroc values for the original benchmark datasets. Click here for file Additional File 5 This file includes Table S5 – The Aroc values for the reduced benchmark datasets. Click here for file Additional File 6 This file includes Table S6 – The Aroc values for the original benchmark datasets (Cysteine substituted). Click here for file Additional File 7 This file includes Table S7 – The Aroc values for the reduced benchmark datasets (Cysteine substituted). Click here for file Acknowledgements The authors are grateful to Lei Huang for helpful discussion and Deepa Vijayaraghavan for the assistance with the computing environment. The authors also thank Dr. Morten Nielsen for sharing of the homology reduced dataset and a scoring matrix. This research is supported in part by National Science Foundation (EIA-022-0301) and Naval Research Laboratory (N00173-03-1-G016). ==== Refs Sette A Buus S Appella E Smith JA Chesnut R Miles C Colon SM Grey HM Prediction of major histocompatibility complex binding regions of protein antigens by sequence pattern analysis Proc Natl Acad Sci USA 1989 86 3296 3300 2717617 Max H Halder T Kropshofer H Kalbus M Muller CA Kalbacher H Characterization of peptides bound to extracellular and intracellular HLA-DR1 molecules Hum Immunol 1993 38 193 200 8106277 Castellino F Zhong G Germain RN Antigen presentation by MHC class II molecules: invariant chain function protein trafficking and the molecular basis of diverse determinant capture Hum Immunol 1997 54 159 169 9297534 Parker KC Bednarek MA Coligan JE Scheme for ranking potential HLA-A2 binding peptides based on independent binding of individual peptide side-chains J Immunol 1994 152 163 175 8254189 Brusic V Rudy G Harrison LC Stonier RJ, Xu XH Prediction of MHC binding peptides using artificial neural networks Complex Systems: Mechanism of Adaptation 1994 Amsterdam, Holland: IOS Press 253 260 Rammensee HG Bachmann J Emmerich NPN Bachor OA Stevanoviæ S SYFPEITHI: database for MHC ligands and peptide motifs Immunogenetics 1999 50 213 219 10602881 Buus S Lauemoller SL Worning P Kesmir C Frimurer T Corbet S Fomsgaard A Hilden J Holm A Brunak S Sensitive quantitative predictions of peptide-MHC binding by a 'Query by Committee' artificial neural network approach Tissue Antigens 2003 62 378 384 14617044 Nielsen M Lundegaard C Worning P Lauemoller SL Lamberth K Buus S Brunak S Lund O Reliable prediction of T-cell epitopes using neural networks with novel sequence representations Protein Sci 2003 12 1007 1017 12717023 Marshall KW Wilson KJ Liang J Woods A Zaller D Rothbard JB Prediction of peptide affinity to HLA DRB10401 J Immunol 1995 154 5927 5933 7751636 Sturniolo T Bono E Ding J Raddrizzani L Tuereci O Sahin U Braxenthaler M Gallazzi F Protti MP Sinigaglia F Hammer J Generation of tissue-specific and promiscuous HLA ligand databases using DNA microarrays and virtual HLA class II matrices Nat Biotechnol 1999 17 555 561 10385319 Borras-Cuesta F Golvano J Garcia-Granero M Sarobe P Riezu-Boj J Huarte E Lasarte J Specific and general HLA-DR binding motifs: comparison of algorithms Hum Immunol 2000 61 266 278 10689116 Brusic V Rudy G Honeyman G Hammer J Harrison L Prediction of MHC class II-binding peptides using an evolutionary algorithm and artificial neural network Bioinformatics 1998 141 121 130 9545443 Nielsen M Lundegaard C Worning P Hvid CS Lamberth K Buus S Brunak S Lund O Improved prediction of MHC class I and class II epitopes using a novel Gibbs sampling approach Bioinformatics 2004 20 1388 1397 14962912 Karpenko O Shi J Dai Y Prediction of MHC class II binders using the ant colony search strategy Artificial Intelligence in Medicine 2005 35 147 156 16061368 Vapnik VN The Nature of Statistical Learning Theory 1995 Springer Bhasin M Raghava GPS SVM based method for predicting HLA-DRB10401 binding peptides in an antigen sequence Bioinformatics 2004 20 421 423 14960470 Mallios RR Class II MHC quantitative binding motifs derived from a large molecular database with a versatile iterative stepwise discriminant analysis meta-algorithm Bioinformatics 1999 15 432 439 10383468 Mallios RR Predicting class II MHC/peptide multi-level binding with an iterative stepwise discriminant analysis meta-algorithm Bioinformatics 2001 17 942 948 11673239 Burden FR Winkler DA Predictive Bayesian neural network models of MHC class II peptide binding J Mol Graph Model 2005 23 481 489 15878832 Lee WS Liu B Learning with positive and unlabeled examples using weighted logistic regression Proceedings of International Conference on Machine Learning (ICML) 2003 448 455 Liu B Dai Y Li X Lee WS Yu P Building text classifiers using positive and unlabeled examples Proceedings of the Third IEEE International Conference on Data Mining (ICDM'03) 2003 Melbourne Florida 179 188 Bennett KP Mangasarian OL Robust Linear Programming Discrimination of Two Linearly Inseparable Sets Optimization Methods and Software 1992 1 23 34 Swets JA Measuring the accuracy of diagnostic systems Science 1988 240 1285 1293 3287615 GLPK Bhasin M Singh H Raghava GPS MHCBN: a comprehensive database of MHC binding and non-binding peptides Bioinformatics 2003 19 665 666 12651731 MHCBench Southwood S Sidney J Kondo A del Guercio MF Appella E Hoffman S Kubo RT Chesnut RW Grey HM Sette A Several common HLA-DR types share largely overlapping peptide binding repertoires J Immunol 1998 160 3363 3373 9531296 Geluk A van Meijgaarden KE Schloot NC Drijfhout JW Ottenhoff TH Roep BO HLA-DR binding analysis of peptides from islet antigens in IDDM Diabetes 1998 47 1594 1601 9753297 Singh H Raghava GPS ProPred: prediction of HLA-DR binding sites Bioinformatics 2001 17 1236 1237 11751237 Gulukota K Sidney J Sette A DeLisi C Two complementary methods for predicting peptides binding major histocompatibility complex molecules Journal of Molecular Biology 1997 267 1258 1267 9150410 Godkin AJ Smith KJ Willis A Tejada-Simon MV Zhang J Elliott T Hill AVS Naturally processed HLA Class II peptides reveal highly conserved immunogenic flanking region sequence preferences that reflect antigen processing rather than peptide-MHC interactions J Immunol 2001 166 6720 6727 11359828	16351712	PMC1325229	CC BY	2021-01-04 16:24:10	no	Immunome Res. 2005 Dec 13; 1:6	utf-8	Immunome Res	2,005	10.1186/1745-7580-1-6	oa_comm
==== Front BMC Med EducBMC Medical Education1472-6920BioMed Central London 1472-6920-5-381635171810.1186/1472-6920-5-38Research ArticleUnhappiness and dissatisfaction in doctors cannot be predicted by selectors from medical school application forms: A prospective, longitudinal study McManus IC [email protected] Sheeraz [email protected] Amuthan [email protected] E [email protected] Joanna [email protected] Department of Psychology, University College London, Gower Street, London WC1E 6BT, UK2 Department of Psychology, University College London, Gower Street, London WC1E 6BT, UK3 Department of Psychology, University College London, Gower Street, London WC1E 6BT, UK4 School of Psychology, University of Nottingham, Nottingham NG7 2RD, UK5 School of Psychology, University of Nottingham, Nottingham NG7 2RD, UK2005 13 12 2005 5 38 38 14 8 2005 13 12 2005 Copyright © 2005 McManus et al; licensee BioMed Central Ltd.2005McManus 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 Personal statements and referees' reports are widely used on medical school application forms, particularly in the UK, to assess the suitability of candidates for a career in medicine. However there are few studies which assess the validity of such information for predicting unhappiness or dissatisfaction with a career in medicine. Here we combine data from a long-term prospective study of medical student selection and training, with an experimental approach in which a large number of assessors used a paired comparison technique to predict outcome. Methods Data from a large-scale prospective study of students applying to UK medical schools in 1990 were used to identify 40 pairs of doctors, matched by sex, for whom personal statements and referees' reports were available, and who in a 2002/3 follow-up study, one pair member was very satisfied and the other very dissatisfied with medicine as a career. In 2005, 96 assessors, who were experienced medical school selectors, doctors, medical students or psychology students, used information from the doctors' original applications to judge which member of each pair of doctors was the happier, more satisfied doctor. Results None of the groups of assessors were significantly different from chance expectations in using applicants' personal statements and the referees' reports to predict actual future satisfaction or dissatisfaction, the distribution being similar to binomial expectations. However judgements of pairs of application forms from pairs of doctors showed a non-binomial distribution, indicating consensus among assessors as to which doctor would be the happy doctor (although the consensus was wrong in half the cases). Assessors taking longer to do the task concurred more. Consensus judgements seem mainly to be based on referees' predictions of academic achievement (even though academic achievement is not actually a valid predictor of happiness or satisfaction). Conclusion Although widely used in medical student selection to assess motivation, interest and commitment to a medical career, the personal statement and the referee's report cannot validly be used by assessors, including experienced medical school selectors, to identify doctors who will subsequently be dissatisfied with a medical career. ==== Body Background Many doctors in the UK, and probably elesewhere, are unhappy with their careers in medicine, with a fifth or more of junior doctors considering leaving medicine [1]. Medical school selectors do not wish only to select academically able students who can cope academically with the medical course, but also want to assess motivation for studying medicine, becoming a doctor, and for a lifetime of practising medicine. An important question, therefore, is whether the information available during student selection can be used to predict which doctors will be happy or unhappy with a career in medicine. Applicants to UK medical schools apply by means of the UCAS (previously UCCA) form, which contains standard demographic data, information on educational achievement and qualifications, a personal statement from the applicant, and a report from a referee [2]. Empirically, educational achievement and intellectual ability have been found to be unrelated to stress, burnout and dissatisfaction with medicine [3], and therefore the most likely source for information which might be able to predict happiness with a medical career is the applicants' personal statements and their referees' reports. Experimental research manipulating the informational content of personal statements, resumes and cover letters shows that applicants are judged as more competent and to have greater potential when the resumes contained relevant educational references [4], whereas trying to present oneself in a positive light was not always beneficial [4,5]. Although informative, such studies do not address the actual predictive utility of personal statements, and whether selectors can identify those applicants who will eventually be successful or happy in their career, particularly when UCAS/UCCA personal statements are not structured to help selectors make their judgments. In this study we used a simple experimental design to assess the ability of assessors to distinguish pairs of doctors who had been qualified for five or six years, one of whom who was very happy and satisfied with their medical career and the other was very unhappy and dissatisfied, using only the personal statement and referee's report submitted as a part of the doctors' UCCA application forms to medical school, twelve years previously. Methods This study was approved under the normal procedures of the Ethics Committee of the UCL Department of Psychology. The study was exempt from the need for specific permission by the UCL Research Ethics Committee (see ; ). Design The study used a paired comparison (two-alternative, forced-choice) design in which a group of assessors in 2005 decided which of two doctors, one very happy and one very unhappy, was the happier doctor (assessed in 2002/3), based on the personal statement and referee's report from the application forms the doctors had submitted to UCCA in 1990. The doctors and their application forms The doctors were part of the 1991 Cohort Study and had applied to medical school during the autumn of 1990 for admission in October 1991 [6], when a wide range of measures was collected, including O-level/GCSE results, predicted A-level grades and actual A-level grades. In 2002/3 these doctors had taken part in a follow-up study which amongst other things assessed happiness and satisfaction with a medical career, as well as stress and burnout [7]; the doctors in the 2002/3 study also completed a brief Big Five personality measure [7]. The present study considered only those doctors who had returned questionnaires in the follow-up study carried out in December 2002, and for whom transcribed UCCA forms were also available. Satisfaction with medicine was assessed using a composite satisfaction score derived from three separate, temporally-anchored questions (see table 1), which correlated strongly with measures of stress and burnout (see Additional File 1). Figure 1 shows the distribution of satisfaction scores; happy doctors were classified as those scoring 16 or above, whereas unhappy doctors were those scoring 9 or below. Table 1 Responses of happy and unhappy doctors. How often do the following statements describe the way you feel about working as a doctor? Group Every day A few times a week Once a week A few times a month Once a month or less A few times a year Never a) I think of giving up medicine for another career Happy 6 34 Unhappy 17 8 5 9 1 b) I reflect on the satisfaction that I get from being a doctor Happy 12 25 3 Unhappy 1 11 12 12 4 c) I regret my decision to have become a doctor Happy 5 35 Unhappy 5 11 5 11 7 1 The three questions used for assessing happiness and satisfaction in the doctors. For the first and third questions (a and c) a response of 'every day' is scored as 0, 'A few times a year' as 1, through to 'Never' being scored as 6, whereas the second question (b) is scored in the reverse order. The responses of the 40 happy doctors are shown in bold, whereas the responses of the 40 unhappy doctors are shown in italics. Figure 1 Distribution of Satisfaction scores in the doctors for whom transcribed UCCA forms were available. Doctors with a score of 16 or above were in the high satisfaction group, and those with a score of 9 or below were in the low satisfaction group. Questionnaire booklets The assessors made their judgements in questionnaire booklets which contained twenty pairs of UCCA forms, with each side of a double-page spread showing the personal statement and referee's report from a satisfied and a dissatisfied doctor, randomised to left (labelled A) or right (labelled B). Pairs were matched for sex, and were from applicants who were UK nationals aged under 21 when they applied to medical school. All forms were transcribed, checked against the originals, fully anonymised and printed in a standard format. For each pair of forms, there was a question at the bottom of the page: "Which is the satisfied, happy doctor? Definitely A/Probably A/Probably B/Definitely B". For the main statistical analysis, judgements of 'definitely' and 'probably' were combined; for correlations with the use of 'definite' judgements, see Additional File 1. Two separate booklets were assembled, Book 1 and Book 2, each containing 20 different pairs of doctors. Overall therefore, forms from 80 doctors were used, 40 being very satisfied and 40 very dissatisfied, but, in order not to overload assessors, each individual assessor saw only one booklet containing 20 pairs of forms from a total of 40 doctors. It should be emphasised that we made no attempt to differentiate 'happiness' and 'satisfaction', and for the purposes of this paper these words should be regarded as synonyms. Assessors The pairs of forms were assessed by four different groups of assessors: 35 experienced medical school selectors (recruited with the assistance of the CHMS Admissions to Medical Schools Group), a convenience sample of 19 doctors known to SI, 22 medical students from UCL, and 20 psychology students from UCL (for details see Additional File 1). The student groups were included because of studies suggesting that students can sometimes be better assessors than experienced judges [8]. After judging the pairs of forms, assessors provided brief data on age, sex, experience of medicine and student selection, and how long the task had taken, and they completed brief assessments of their own personality (Big Five, empathy, and communicative ability). Note: There is much potential for confusion in understanding the results of this study. In particular we would like to emphasise that when we refer to assessors we are referring to the individuals making judgements of the forms in 2005, and when we refer to doctors (and to pairs of doctors) we are referring to the individuals who applied to medical school in 1990, and whose UCCA forms were used in this study, and to the questionnaire responses made by the same individuals as practising doctors in 2002/3. In particular it should be noted that both the doctors and the assessors each completed a Big Five personality assessment, and each is analysed below. Results The pairs of application forms were judged by 96 assessors, 48 of whom looked at Book 1 and 48 looked at Book 2. A chance null hypothesis predicts that random guessing would result in an assessor making an average of 10 correct judgements out of 20 (50%). In fact overall, of the 1920 judgements made by the 96 judges, 963 (50.2%) were correct, which is not significantly different from 50% (χ2 = .018, 1df, P = .89). The non-significant overall success rate may conceal differences between groups or individual assessors. The four groups of assessors showed similar overall success rates (medical school selectors: 48.0%; doctors 51.8%; medical students 51.4%; psychology students: 51.0% ; oneway ANOVA: F(3,92) = .932, P = .429). The chance expectations of individual assessors can be modelled using a binomial distribution (equivalent to asking how many heads would be expected if a fair coin was tossed 20 times). Figure 2 shows that the scores of individual assessors conform closely to binomial expectations. There is therefore no evidence that some individual assessors can carry out the task at significantly better than chance expectations. Neither was there evidence that the demographic or personality measures of the assessors predicted success at the task (see Additional File 1). Figure 2 Number of happy doctors correctly identified by the 96 assessors, each of whom made a comparative judgement on 20 different pairs of doctors. The horizontal axis shows the number of times the doctor in a pair who was actually the happier was correctly identified as such by an assessor for each of the 20 pairs assessed by an assessor. The blue bars shows the actual numbers of correct judgements out of twenty for each assessor, and the yellow bars the expected distribution under a binomial distribution (left-hand vertical axis). The dashed red line shows the expected cumulative binomial distribution, and the solid line the actual cumulative binomial distribution (right-hand vertical axis). The light blue bars show pairs within the 95% range of the binomial distribution, whereas the dark blue bar show a pair which is outside the approximate 95% range, and hence is significant at about the 5% level. A possible explanation for the inability of assessors to perform above chance expectations is that the task is simply too hard, and assessors are, in effect, forced to make decisions at random. That possibility can be evaluated by looking at the performance of the assessors on the 40 individual pairs of forms. Each pair of doctors was judged by 48 assessors, and therefore the binomial chance expectation is that on average the correct judgement for a particular pair of doctors will be made by 24 out of the 48 assessors. Figure 3 shows the actual number of correct judgements made by the assessors for each of the 40 pairs of doctors, and the binomial expectations (in effect the number of heads expected when a fair coin is tossed 48 times). The distribution is clearly non-binomial, with 21 out of the 40 pairs (52.5%) being outside of the 95% confidence interval expected by chance alone. However, although judgements for 21 of the pairs of doctors were non-random, in 11 pairs of doctors the assessors correctly predicted the happier doctor, but in the other 10 pairs the doctor chosen by the assessors as happier was actually the unhappier doctor. Figure 3 The number of times, for each of the 40 pairs of doctors, that the doctor in each pair of doctors who was actually more satisfied, was identified as happier by each of the 48 assessors who considered that pair of doctors. The horizontal axis shows the number of assessors (0–48) who were correct for a particular pair of doctors, and the light and dark blue bars shows the distribution for each pair of doctors. Note that in this figure the bars correspond to pairs of doctors (and hence sum to 40), whereas in figure 2 they corresponded to assessors (and hence sum to 96). The yellow bars show the expected distribution under a binomial distribution (left-hand vertical axis). The dashed red line shows the expected cumulative binomial distribution, and the solid line the actual cumulative binomial distribution (right-hand vertical axis). The light blue bars show pairs of doctors within the 95% range of the binomial distribution, whereas the dark blue bars show pairs which are outside the 95% range, and hence are significant at the 5% level. Figure 3 confirms that assessors cannot be making random judgements, and that for many of the pairs of doctors there is a clear agreement between the assessors as to which doctor will be the happier (even though that judgement is erroneous for half of the pairs of doctors). A 'consensus score' was calculated to assess how often an individual assessor had concurred with the consensus judgement on the 21 pairs where on aggregate across all assessors there was agreement beyond chance expectations. Eighty-four of the assessors also reported how long they had taken to carry out the task (mean 80.8 minutes; SD 42.6 minutes; median = 60 minutes; range 30–240 minutes). Assessors who had taken longer to carry out the task were less likely to be correct overall (Pearson's r = -.289, p = .008) and more likely to concur with other judges (r = .274, p = .012), both correlations with time remaining significant after partialling out the other measure. It is clear that assessors are unable to carry out the task beyond chance expectations, but they nevertheless concur with one another in their judgements. The question therefore arises as to the nature of the information from the forms which is acting as the basis for the assessors' (erroneous) consensus. We had background measures on the doctors in the pairs and we therefore looked to see whether those measures predicted the judgements of the assessors. It should be emphasised here that we are not looking here at correlations between the actual happiness of doctors and the background measures, but rather between those same background measures and which doctors are thought more likely to be the happier by the assessors. Although personality is known to be a strong correlate of happiness and satisfaction in doctors [3,7], there was no correlation between the difference in Big Five personality scores of the two doctors in each pair and the consensus judgement of the assessors (be it right or wrong). Assessors are not therefore implicitly assessing aspects of personality. We also had measures of educational achievement in the pairs of doctors, including O-level/GCSE grades, A-level grades, and expected A-level grades in those applying pre-A-level. The difference in educational achievement between the doctors in a pair correlated significantly with the consensus judgement of the assessors, with the largest correlation being with expected A-level grades (r = .321, p = .011, N = 62). Since expected A-level grades, and other aspects of educational achievement, are often referred to in referees' reports, it seems likely that assessors are judging that the doctor with the higher, predicted educational achievement will be the happier, more satisfied doctor. Discussion UK medical schools put extensive effort into reading the application forms submitted by applicants, using them as a basis for shortlisting for interview (and in some cases for making offers), and assessing both academic achievement and motivation from them. However the present study suggests that assessors, including experienced medical school selectors, are unable to use application forms to judge which applicants will become happy and satisfied doctors. That does not mean, though, that selectors are making random judgements on the basis of the forms. There is a consensus between assessors, which is greatest amongst those who have spent longest reading the forms. However that consensus seems principally to result from using information about the academic potential of applicants, despite there being good reason to believe that it is not a lack of academic ability which causes the dissatisfaction of unhappy doctors, or that academic ability correlates with happiness [3]. In view of the implications of the present findings for current methods of student selection in the UK (and elsewhere), it would be reassuring if other studies, in other settings, and with different outcome criteria, were to find similar results, and we hope that such studies will be carried out. The implicit use of academic ability as a criterion is consistent with other work [4], which has shown that when raters used a structured pro-forma to assess the presence of various attributes in a resume (e.g., academic ability, social exchanges), then those ratings are related to applicants personality and cognitive abilities [9]. However, these results consider judgements based on pre-defined categories, whereas personal statements are often judged without such a standardized approach – as in UCCA and UCAS forms – and therefore the predictive validity of personal statements for identifying 'at risk' candidates is low. Even should structured personal statement coding be correlated with personality, it might still be more effective to use psychometrically assessed personality scores than personal statements, if it were desired to select on the basis of personality. Such conclusions are, of course, predicated on the basis that selectors use personal statements to identify qualities that mark whether a candidate will be a happy and good doctor. The analyses reported here suggest, though, that academic criteria are still relied on as the main differentiator, which may reflect social and cultural stereotypes of what makes a good doctor. The overall outcome measure in our study, of satisfaction and happiness with medicine as a career, might be criticised on the grounds that what is really required is knowledgeable, effective, doctors who stay in post. If knowledgeable, academically well-qualified doctors alone is all that is required, then probably the best predictor of academic success is educational achievement [3,10], but that is assessed most efficiently, effectively and objectively from measures of educational achievement, rather than indirectly from a referee's report. Academic ability and career satisfaction are not however correlated [3]. There is however good evidence that our unhappy doctors are stressed and burned out (see Additional File 1), that stress and burnout result in less effective work in doctors [11], and that dissatisfied, unhappy doctors are the ones who are most likely to want to leave medicine [1]. Satisfaction is therefore an important outcome variable. Perhaps the most surprising thing about our study is that the task was completely impossible. That might reflect the relative poverty of the information provided on the application forms (for the personal statement the median length is 107 words (mean = 106, SD 22, quartiles 92–120, range = 54–162), and for referees' reports the median length is 398 words (mean = 391, SD 111; quartiles 325–465; range 114–734)). The information provided might also be inaccurate or misleading, either unintentionally, or as part of a process of 'impression management' [5]. Referees have an obligation to the medical school, but also to the person on whose behalf they are making a statement, perhaps ignoring their weaknesses and over-emphasising their strengths, so that some geese appear as swans. It is therefore possible that information obtained differently might be more effective in prediction. The UCCA forms completed in 1990 are now somewhat different from those currently used, and it is possible that modern forms might provide more information which is more useful. However several selectors spontaneously commented that they were surprised by the honesty and the openness of the referees' reports made in 1990, and that more modern statements seemed bland, non-committal and uninformative in comparison. Our results are therefore likely to be valid for modern application forms. We had wondered whether there may be further information about applicants concealed within the subtleties and the nuances of the language used (e.g. the use of negative versus positive emotion words). We therefore used a computerised text analysis program, LIWC [12,13], to assess personal statements and referees' reports on about 70 semantic, syntactic and stylistic measures, but could find no systematic difference between UCCA forms from happy and unhappy doctors. Nevertheless it is possible that there is additional information within the specific content of the forms, which could be indirect indicators of personality, and in particular neuroticism, which is itself a major predictor of stress and burnout. The Schwartz Report into admissions to universities, was critical of the use of selection methods for which there was not demonstrable evidence of reliability or validity [14]. UCCA/UCAS forms, along with A-level grades, are the single most important piece of evidence used in medical school selection, each being carefully read and assessed, and often they are the sole basis for rejection. Applicants, their parents, and their schools therefore put much effort into every detail of the forms. Our study is an assessment of the long-term validity of judgements made from the information on UCCA/UCAS forms. Judgements made are to some extent reliable, as seen by the consensus between assessors (and shown elsewhere [15]). However although many claims are made for the utility of the personal and referee's information provided on the forms, we could find no evidence of long-term predictive validity for an important outcome variable – the judgement of whether or not an applicant will be a happy and satisfied doctor, or instead will be an unhappy, stressed, burned out, dissatisfied doctor who does not enjoy their job and thinks often of leaving for another career. Competing interests The author(s) declare that they have no competing interests. Authors' contributions The 1991 cohort study was originated by ICM. ICM and EF had the idea of transcribing and analysing UCCA forms, and the forms were scanned and transcribed by JL and EF. ICM designed the experimental study, and SI and AC carried out the experiment as a part of research projects submitted as a component of their degree programmes. ICM wrote the first draft of the manuscript, and all authors revised the manuscript. ICM is the guarantor for the study. Pre-publication history The pre-publication history for this paper can be accessed here: Supplementary Material Additional File 1 Supplementary Information – three additional tables and one additional figure Click here for file Acknowledgements We are grateful to Dr Patricia Hughes, Chair of the CHMS Admissions to Medical Schools Group, for her assistance in distributing the questionnaire booklets to medical school selectors, and to the selectors and other assessors who took part in the study. The 2002 follow-up of the 1991 cohort, and subsequent analyses, were funded by the London Deanery. ==== Refs Moss PJ Lambert TW Goldacre MJ Lee P Reasons for considering leaving UK medicine: questionnaire study of junior doctors' comments Brit Med J 2004 329 1263 1265 15469947 Universities and Colleges Admissions Service website McManus IC Smithers E Partridge P Keeling A Fleming PR A levels and intelligence as predictors of medical careers in UK doctors: 20 year prospective study Brit Med J 2003 327 139 142 12869457 Knouse SB Impressions of the resume: the effects of applicant education, experience, and impression management Journal of Business and Psychology 1994 9 33 45 10.1007/BF02230985 Knouse SB Impression management in the resume and its cover letter Journal of Business and Psychology 1994 3 242 249 10.1007/BF01014492 McManus IC Richards P Winder BC Sproston KA Styles V Medical school applicants from ethnic minorities: identifying if and when they are disadvantaged Brit Med J 1995 310 496 500 7888888 McManus IC Keeling A Paice E Stress, burnout and doctors' attitudes to work are determined by personality and learning style: A twelve year longitudinal study of UK medical graduates BMC Medicine 2004 2 29 15317650 10.1186/1741-7015-2-29 Kocsis R Profiling the criminal mind: does it actually work? Lancet 2004 364 s14 s15 15967135 10.1016/S0140-6736(04)17623-X Cole MS Field HS Giles WF Using recruiter assessments of apploicants' resume content to predict applicant mental ability and big five personality dimensions ijsa 2003 11 78 88 McManus IC Powis D Ferguson E James D Richards P Selecting medical students: Why intellectual aptitude tests probably aren't useful for selecting UK school-leaver entrants, and how A-levels could be Brit Med J 2005 Winefield HR Dollard MF, Winefield AH, Winefield HR Work stress and its effects in general practitioners Occupational stress in the service professions 2003 London: Taylor and Francis 191 212 Pennebaker JW Francis ME Booth RJ Linguistic Inquiry and Word Count (LIWC): LIWC 2001 2001 Mahwah, NJ: Lawrence Erlbaum Pennebaker JW Mehl MR Niederhoffer KG Psychological aspects of natural language use: our words, our selves ARP 2003 54 547 577 Admissions to Higher Education Steering Group Fair admissions to higher education: recommendations for good practice 2004 Nottingham: Dept for Education and Skills Publications McManus IC Richards P Reliability of short-listing in medical student selection Med Educ 1989 23 147 151 2716551	16351718	PMC1325230	CC BY	2021-01-04 16:30:56	no	BMC Med Educ. 2005 Dec 13; 5:38	utf-8	BMC Med Educ	2,005	10.1186/1472-6920-5-38	oa_comm
==== Front World J Surg OncolWorld Journal of Surgical Oncology1477-7819BioMed Central London 1477-7819-3-751629318610.1186/1477-7819-3-75ResearchLocal treatment in unresectable hepatic metastases of carcinoid tumors: Experiences with hepatic artery embolization and radiofrequency ablation Meij Vincent [email protected] Johanna M [email protected] Hillegersberg Richard [email protected]öger Robert [email protected] Warner [email protected] Coevorden Frits [email protected] Babs G [email protected] Department of Surgery, Netherlands Cancer Institute/ Antoni van Leeuwenhoek Hospital, Amsterdam, The Netherlands2 Department of Medical Oncology, Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands3 Department of Radiology, Netherlands Cancer Institute/ Antoni van Leeuwenhoek Hospital, Amsterdam, The Netherlands2005 17 11 2005 3 75 75 1 7 2005 17 11 2005 Copyright © 2005 Meij et al; licensee BioMed Central Ltd.2005Meij 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 Hepatic metastases of carcinoid tumors cause incapacitating symptoms, but are usually diffuse and therefore unresectable. In this article we evaluate our experiences with local treatment techniques in the management of carcinoid patients with hepatic metastases and failing systemic treatment. Methods Fifteen consecutive carcinoid patients (11 men and 4 women; median age 60 years; range 45–71 years) were treated with either hepatic artery embolization (HAE) with Ivalon particles or radiofrequency ablation (RFA) (percutaneously or intra-operatively). Follow-up evaluation was performed by CT scan and 24-hours urinary 5-HIAA excretions. Results A total of 18 HAE's was performed in 13 patients, while 10 lesions in 3 patients were treated with RFA. Median follow-up was 12.5 months (2 – 25 months). Median duration of symptoms was 22 months (8 – 193 months). Median overall decrease of 5-HIAA excretion 2 months after HAE was 32% with tumor regression on CT-scan in 4 patients (30%) and improvement of symptoms with a median duration of 15 months in 3 of them (23%). Embolization led to fatal hepatic failure in one patient. The 3 patients treated with RFA showed a decrease of urinary 5-HIAA values of 34, 81 and 93% respectively, with tumor regression in all of them. Improvement of symptoms was reported in 2 patients up to 25 months. Conclusion Liver embolization performed late in the clinical course had limited effect on symptoms and biochemical and radiological parameters. First experiences with RFA are favorable and might encourage to apply RFA more widely in metastatic carcinoid. ==== Body Background Carcinoid tumors are derived from neuroendocrine cells and are usually slowly growing malignant tumors with distinct morphological characteristics. Most frequently, carcinoids are located in the appendix and ileum, but they are also known to arise from other sites, such as the bronchus and pancreas. The carcinoid syndrome is mostly due to hepatic metastases with the release of hormones, such as serotonin, directly into the systemic circulation and is characterized by symptoms of flushing, diarrhea, and wheezing. Eventually right-sided heart-valve fibrosis may develop. There are several treatment options in metastatic disease. A multidisciplinary approach is advocated, in which gastroenterologists, medical oncologists and surgeons all have a role [1]. Systemic treatment is aiming at symptomatic improvement and reduction of hormonal secretion. Somatostatin analogues, interferon-alpha and application of meta-iodobenzylguanidin (MIBG), pharmacological or combined with radioactive MIBG, can result in long-lasting palliation [2-6]. However, tumor reduction is only occasionally described with these treatment modalities and as tumor load increases, symptomatic treatment gradually fails. Local treatment of hepatic metastases of carcinoid is attractive because of the slow and localized growth pattern. When metastases are restricted to one lobe of the liver a hemihepatectomy or segmental resection is advocated. Unfortunately, the metastases are often multiple and diffuse and therefore resection is usually impossible due to anatomical location or number or due to inadequate viable liver tissue that would remain after surgery. If feasible, hepatic resection can be performed safely and provides effective palliation with considerable duration [7]. Another local treatment option is hepatic artery embolization (HAE), which not only may ameliorate symptoms but also might reduce tumor burden. An objective biochemical response of up to 52% and a median duration of effect of 12 months have been reported in cases of failing systemic therapy [8]. Reports on chemoembolization show a slightly better biochemical response and tumor response [9]. Although radio frequency ablation (RFA) is a fairly new technique, various studies have demonstrated its effectiveness in the treatment of unresectable hepatocellular carcinomas and hepatic metastases of colon carcinoma [10,11]. In patients with diffuse hepatic metastasis RFA could achieve local tumor control while remaining enough functional liver tissue. Previous reports on RFA of hepatic carcinoid metastases are limited to case reports and small series but have indicated good local tumor control and acceptable morbidity [8,12-15]. Besides our experience in HAE, we have recently introduced the use of RFA in carcinoid patients. In this article we describe our experience of local treatment techniques in our institutional series of unresectable hepatic metastases of carcinoids and failing systemic therapy. Patients and methods Patients Among 46 patients with carcinoid syndrome treated in the Netherlands Cancer Institute/ Antoni van Leeuwenhoek Hospital between January 2000 and June 2002 we evaluated the data of all patients (n = 15) who were treated with either hepatic artery embolization or RFA. Indications were progressive disease with systemic treatment and multiple hepatic metastases not amenable for resection, because of diffuse localization and bilobar involvement. Somatostatine analogues, interpheron-alpha and meta-iodobenzylguanidin, pharmacological or combined with radioactive MIBG could no longer ameliorate symptoms in these patients. Treatment with RFA was selected in patients where this was technically possible and all metastases could be treated with sufficient remaining functional liver tissue and with safe margins to the portal vein. Before treatment CT scans were performed and urinary 5-hydroxyindole acetic acid (5-HIAA) excretion was measured. Embolization The radiographic procedure was started with a diagnostic arteriography of the celiac axis and the superior mesenteric artery using the Seldinger technique to assess arterial anatomy. The portal vein patency was assessed on pre-embolization CT scan. Thereafter selective catheterisation was performed of the arterial supply of the right (segment 5–8) or left (segment 2–4) lobe pending on localization of the largest metastatic area. Ivalon particles 150–600 μ in polyvinyl alcohol mixed with an iodined contrast medium were injected until stasis was obtained. When tumor was present in both lobes of the liver, both lobes were embolized with an interval of several months starting with the lobe with the largest tumor load. In these cases, evaluation of symptoms and biochemical and radiological signs was assessed after the second embolization. After treatment vital signs, as well as hepatic and renal functions were monitored on the ward regularly. An adequate diuresis was ensured by careful hydration to avoid hepatorenal failure in the post-embolization period Radiofrequency ablation RFA was performed either intraoperatively or percutaneously. A 3.5 or 3.0 LeVeen Needle was used in combination with a 90 watts radio wave generator (Radiotherapeutics, USA). After ultrasound-guided insertion of the needle, sequential increase per minute of the radio wave current was performed until a maximum of 90 watts was achieved. An increase in resistance ultimately resulted in a "roll off", after which a second procedure at the same localization was restarted after a one-minute interval without current, to a maximum of 70% of the first achieved level. Each localization was thus treated, or a combination of multiple passes of the LeVeen needle was used to cover a metastasis larger than 4 cm until complete tumor necrosis was achieved. During follow-up symptoms and measurement of urinary 5-HIAA's were evaluated at least every 3 months in our outpatient clinic. Tumor size and necrosis of tumor were evaluated by a CT scan at 3, 6 and 12 months after treatment. In the patients who were treated with RFA an additional CT scan one month after the procedure was performed. Results Eighteen hepatic artery embolizations were performed in 13 patients and 10 metastases were treated with RFA in 3 patients (table 1). There were 11 men and 4 women with a median age of 60 years (range 45 to 71) at time of treatment. The primary tumor was located in the ileocoecal area in 7 patients, rectum in 1 patient, lung in 3 patients, while in the remaining 4 patients the primary tumor was unknown. Table 1 Characteristics of patients No. Age at local therapy Duration of symptoms (months) Gender Primary tumour Local therapy Location of therapy 1 45 78 F unknown emb. right 2 67 24 M unknown emb. right 3 53 11 M rectum emb. bilat. 4 71 67 M lung emb. right 5 70 63 F lung emb. right 6 60 26 M ileocoecal emb. bilat. 7 62 24 M appendix emb. left 8 60 11 F ileocoecal emb. right 9 61 23 M lung emb. right 10 71 10 M unknown emb. bilat. 11 52 21 M ileocoecal emb. bilat. 12 59 16 M unknown emb. bilat. 13a 55 11 M ileocoecal emb. right 13b 55 17 M ileocoecal RFA left 14 47 8 F ileocoecal RFA bilat. 15 60 193 M ileum RFA bilat. Indications for RFA were a vascular anomaly because of which HAE could not be performed in 1 patient. The other patients treated with RFA had very large lesions among several smaller ones, which made them more appropriate for RFA. The diagnosis of metastatic disease was known during a considerable time before local treatment (median duration of symptoms: 22 months, range: 8–193 months). As shown in table 2, 14 of the 15 patients had elevated urinary 5-HIAA excretion values with a median of 913 μmol/24 hrs (range 35 – 2618 μmol/24 hrs) (normal <40 μmol/24 hrs) prior to embolization. Median follow-up from completion of therapy was 12,5 months (range 6 – 25 months). Seven patients were dead at the time of analysis. Table 2 Biochemical and radiological effects and effects on symptoms No. 5-HIAA before 5-HIAA after 2 months 5-HIAA at follow-up Effect on tumoursize on CT-scan Effect on symptoms Follow-up (months) Complications 1 1736 unknown unknown unknown = 2(†) no 2 35 23 23 progression ↓ 4(†) postembolization syndrome 3 120 160 42 progression = 7(†) hepatic failure 4 1924 1549 1432 progression = 10(†) no 5 111 75 123 progression (after 9 months) = 11(†) postembolization syndrome 6 1662 1564 2022 progression = 12(†) no 7 1740 1617 1604 progression (after 5 m) = 14 abscess 8 913 140 566 regression ↑ 14 postembolization syndrome 9 1961 1531 2050 no change = 14(†) postembolization syndrome 10 72 33 118 regression ↑ 15 no 11 875 720 765 progression = 15 no 12 1529 165 502 regression ↑↑ 22 postembolization syndrome 13a 2618 1458 1186 regression = 6* no 13b 1186 777 1815 regression = 7 no 14 91 6 32 regression ↑ 13 cholestasis 15 277 52 79 regression ↑ 25 no (↓: worsening of symptoms, =: no change, ↑: improvement of symptoms, ↑↑: free of symptoms). Embolization resulted in an overall median decrease of 5-HIAA excretion of 32% (range 0–89%) at first follow-up. In 3 patients, (23%; 95%-confidence interval: 5–54%) the decrease after embolization was more than 50% and met the criteria of biochemical response (patient 8, 10 and 12). In these 3 patients, symptoms improved, including patient 12 who were free of symptoms after bilateral embolization. This symptomatic improvement in the patients with biochemical response was still present at follow-up of 14, 15 and 22 months respectively. In the remaining ten patients only a minimal biochemical response was achieved with no effect on symptoms. Tumor regression on CT scan was present in 4 patients, which corresponded with symptomatic improvement in three of them. In 7 patients CT-scans revealed progression and in one there was no change in tumor size. In one patient radiological evaluation could not be performed as this patient died after two months because of cardiac failure. Postembolization syndrome, characterized by elevated levels of transaminases, lactate dehydrogenase and alkaline phosphatase combined with fever, occurred in 5 patients. All recovered within 10 days. More serious complications occurred in 2 cases. In one patient (nr.7) a necrotic tumor metastasis was infected, for which treatment with intravenous antibiotics and abscess drainage was needed. After embolization of the left artery in another patient (nr.3), embolization of the right hepatic artery was performed after 5 months, which led to fatal hepatic failure 2 months later. Six other patients died during follow-up because of non-intervention related problems after 2 to 14 months: right sided cardiac failure (n = 2), progressive disease (n = 2), carcinoid crisis with renal failure (n = 1) and pneumonia (n = 1). Treatment with RFA induced a biochemical response in 2 out of 3 patients with a 5-HIAA reduction of 81% and 93%. They also experienced a radiological response, and a symptomatic improvement. Patient 14 underwent RFA for one large and several smaller lesions. Patient 15 underwent a resection of segment 6 and 7, because of a large lesion of 7 cm in diameter and RFA of 5 other lesions. In the patient with the longest follow-up the decrease in 5-HIAA excretion after 26 months was still 71% and symptoms were minimal without additional systemic treatment. In patient 13 with a favorable response to embolization of the right hepatic artery, embolization of the left hepatic artery was technically impossible because of a vascular anomaly. He was subsequently treated with multiple passes of RFA percutaneously for two large lesions (6 and 7 cm in diameter) in the let lobe under spinal anesthesia leading to a further reduction in 5-HIAA excretion. However, although CT scans after RFA showed no recurrence of tumor, duration of decreased 5-HIAA values lasted only 3 months and a relief of symptoms could not be obtained. No serious adverse effects occurred after RFA. In one patient (nr.14) post-operative cholestasis was present. A magnetic resonance cholangiopancreatography revealed a minor stenosis of the common bile duct. Symptoms completely recovered after 2 months and hepatic values returned to normal values after 3 months, without additional treatment. Discussion In carcinoid tumors with liver metastases patients suffer from the production of vasoactive hormones. Various systemic treatments are available to diminish the incapacitating symptoms of the carcinoid syndrome. Somatostatin analogues are widely applied and reduce these symptoms in approximately 70%, while a biochemical response is achieved in 42% up to 72% and a duration of up to a year [2,6]. Interferon-alpha has demonstrated similar effects [3]. Application of meta-iodobenzylguanidin (MIBG), in a pharmacological dosage or combined with radioactive MIBG, can give long-lasting palliation [4,5]. Local treatment of the liver, however, remains favorable due to reduction of the tumor load. The metastases are usually not amenable for resection, but HAE and the newer technique of RFA might be effective in the treatment of these patients. Results of local treatment of this tumor are specifically interesting because of its histological characteristics. Arterial hypervasculation of liver metastases from carcinoid tumors argues in favor of hepatic arterial embolization. In addition, it is thought that RFA might have a specific effect on carcinoid tumors, because of the increased conduction of frictional heating in the hypervascular tissue, resulting in more extensive coagulative necrosis. Results of HAE and chemoembolization in earlier reports are variable [8,16-19]. Symptomatic improvement after HAE is reported to occur in 64% to 90% [17-19]. Reports on chemoembolization show a slightly better biochemical and tumor response [9,16]. In our series, however, a symptomatic improvement and a corresponding objective biochemical response are only seen in 3 of 13 (23%) patients treated with HAE. The median decrease of urinary 5-HIAA excretion was 32%. It can be hypothesized that HAE was applied to late in the clinical course. The optimal moment of embolization in the clinical course has been discussed in a Swedish study of 29 patients with carcinoid syndrome, but early embolization appeared to be as effective as late embolization [8]. In addition, side effects in our series were impressive leading to a protracted recovery in several cases. It must be noted that all patients had a considerable duration of symptoms before treatment (median 22; range 8 – 193) and systemic therapy was failing. Remarkably, the 3 patients that did benefit from this treatment had a relatively short duration of symptoms (10, 11 and 16 months). This suggests that embolization in an earlier stage might be more effective. The question is if the same holds true for RFA. RFA might offer a more simple local treatment compared to HAE. The good local tumor control with a satisfactory duration of effect on symptoms has been suggested in several reports of carcinoid patients [12-15]. Although the experience is limited to small series, this may indicate that an early intervention with RFA is favorable. In our report the three patients treated with RFA showed an impressive biochemical response. RFA could ameliorate symptoms in 2 of three patients with duration of over 13 and 25 respectively. Although the measured 5-HIAA values before treatment in these 2 patients were relatively low, they suffered from incapacitating symptoms at this point and a considerable decrease in 5-HIAA values as well as improvement of symptoms could be reached after RFA. Complications are reported to be minimal and include hepatic failure, hepatic abscess and hepatic infarction [10,12]. In our report one case of transient cholestasis occurred after RFA. In our series RFA is a safe and, when performed percutaneously, a minimally invasive procedure which may provide good local tumor control. Although these results are only first experiences, the excellent biochemical response and duration of symptomatic improvement encourage considering this technique as a treatment in cases of metastatic carcinoid syndrome. Difficulty in evaluating local ablation methods lies in the fact that patients with advanced stage carcinoid syndrome all have specific characteristics, e.g. duration and degree of symptoms, localization of hepatic metastases and site of the primary tumor. To further explore the possibilities and indications of local treatment options all these factors should be taken account for in future studies. Conclusion We find these results encouraging to broaden the indication of treatment with RFA and to reconsider the time of treatment with embolization. The treatment of carcinoid syndrome remains a challenge for the physician and all therapeutic options must be considered carefully. Competing interests The author(s) declare that they have no competing interests. Authors' contributions VM carried out the assembly and analysis of the data and drafted the manuscript. JZ and RH participated in the design of the study and provided study material and patients. RK and WP were responsible for the hepatic artery embolizations and percutaneous RFA procedure and collected all the radiological data. FC carried out the intra-operative RFA procedures and conceived of the study. BT participated in the concept and design and coordination of the study. All authors read and approved final version for publication Figure 1 In patient 12 the initial CT-scan (figure 1A) of the liver shows multiple small nodules in both liver lobes. After HAE (figure 1B) the metastases are confluent and the total volume of the liver has decreased remarkably. Symptoms improved and a biochemical response was obvious. Figure 2 This slide of the CT-scan of patient 14 shows one dominant metastasis in the right liver lobe, but several other small nodules are present (figure 2A). After RFA necrosis with typical cystic appearance developed (figure 2B). Figure 3 CT-scan of patient 13 reveals several small and larger metastases in both liver lobes (figure 3A). HAE of the right liver lobe resulted in fair tumor reduction (figure 3B). Two large metastases in the left liver lober were treated by RFA and show characteristic cystic appearance (figure 3C). The end result six months after the RFA treatment shows a significant decrease of tumor mass in both liver lobes (figure 3D). ==== Refs Caplin ME Buscombe JR Hilson AJ Jones AL Watkinson AF Burroughs AK Carcinoid tumor Lancet 1998 352 799 805 9737302 10.1016/S0140-6736(98)02286-7 Kvols LK Moertel CG O'Connell MJ Schutt AJ Rubin J Hahn RG Treatment of the malignant carcinoid syndrome. Evaluation of a long-acting somatostatin analogue N Engl J Med 1986 315 663 666 2427948 Veenhof CH de Wit R Taal BG Dirix LY Wagstaff J Hensen A Huldij AC Bakker PJ A dose-escalation study of recombinant interferon-alpha in patients with a metastatic carcinoid tumor Eur J Cancer 1992 28 75 78 1373635 10.1016/0959-8049(92)90389-J Taal BG Hoefnagel CA Valdes Olmos RA Boot H Beijnen JH Palliative effect of metaiodobenzylguanidine in metastatic carcinoid tumors J Clin Oncol 1996 14 1829 1838 8656251 Zuetenhorst H Taal BG Boot H Valdes Olmos R Hoefnagel C Long-term palliation in metastatic carcinoid tumors with various applications of meta-iodobenzylguanidin (MIBG): pharmacological MIBG, 131I-labelled MIBG and the combination Eur J Gastroenterol Hepatol 1999 11 1157 1164 10524647 Ricci S Antonuzzo A Galli L Orlandini C Ferdeghini M Boni G Roncella M Mosca F Conte PF Long-acting depot lanreotide in the treatment of patients with advanced neuroendocrine tumors Am J Clin Oncol 2000 23 412 415 10955874 10.1097/00000421-200008000-00020 Que FG Nagorney DM Batts KP Linz LJ Kvols LK Hepatic resection for metastatic neuroendocrine carcinomas Am J Surg 1995 169 36 42 7817996 10.1016/S0002-9610(99)80107-X Eriksson BK Larsson EG Skogseid BM Lofberg AM Lorelius LE Oberg KE Liver embolizations of patients with malignant neuroendocrine gastrointestinal tumors Cancer 1998 83 2293 2301 9840528 10.1002/(SICI)1097-0142(19981201)83:11<2293::AID-CNCR8>3.0.CO;2-E Diamandidou E Ajani JA Yang DJ Chuang VP Brown CA Carrasco HC Lawrence DD Wallace S Two-phase study of hepatic artery vascular occlusion with microencapsulated cisplatin in patients with liver metastases from neuroendocrine tumors Am J Roentgenol 1998 170 339 344 9456942 Iannitti DA Dupuy DE Mayo-Smith WW Murphy B Hepatic radiofrequency ablation Arch Surg 2002 137 422 426 11926946 10.1001/archsurg.137.4.422 Wong SL Edwards MJ Chao C Simpson D McMasters KM Radiofrequency ablation for unresectable hepatic tumors Am J Surg 2001 182 552 557 11839316 10.1016/S0002-9610(01)00813-3 Berber E Flesher N Siperstein AE Laparoscopic radiofrequency ablation of neuroendocrine liver metastases World J Surg 2002 26 985 990 12016479 10.1007/s00268-002-6629-5 Wessels FJ Schell SR Radiofrequency ablation treatment of refractory carcinoid hepatic metastases J Surg Res 2001 95 8 12 11120628 10.1006/jsre.2000.5988 Hellman P Ladjevardi S Skogseid B Akerstrom G Elvin A Radiofrequency Tissue Ablation Using Cooled Tip for Liver Metastases of Endocrine Tumors World J Surg 2002 26 1052 1056 12016482 10.1007/s00268-002-6663-3 Siperstein AE Rogers SJ Hansen PD Gitomirsky A Laparoscopic thermal ablation of hepatic neuroendocrine tumor metastases Surgery 1997 122 1147 1154 9426432 10.1016/S0039-6060(97)90221-X Ruszniewski P Rougier P Roche A Legmann P Sibert A Hochlaf S Ychou M Mignon M Hepatic arterial chemoembolization in patients with liver metastases of endocrine tumors. A prospective phase II study in 24 patients Cancer 1993 71 2624 2630 8384072 Schell SR Camp ER Caridi JG Hawkins IF Jr Hepatic artery embolization for control of symptoms, octreotide requirements, and tumor progression in metastatic carcinoid tumors J Gastrointest Surg 2002 6 664 670 12399054 10.1016/S1091-255X(02)00044-6 Carrasco CH Charnsangavej C Ajani J Samaan NA Richli W Wallace S The carcinoid syndrome: palliation by hepatic artery embolization AJR Am J Roentgenol 1986 147 149 154 2424291 Mitty HA Warner RR Newman LH Train JS Parnes IH Control of carcinoid syndrome with hepatic artery embolization Radiology 1985 155 623 626 4001362	16293186	PMC1325231	CC BY	2021-01-04 16:23:35	no	World J Surg Oncol. 2005 Nov 17; 3:75	utf-8	World J Surg Oncol	2,005	10.1186/1477-7819-3-75	oa_comm
==== Front Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1451633669510.1186/1465-9921-6-145ResearchVariation in conserved non-coding sequences on chromosome 5q and susceptibility to asthma and atopy Donfack Joseph [email protected] Daniel H [email protected] Zheng [email protected] Thorsten [email protected] Inna [email protected] Kelly A [email protected] Carole [email protected] Department of Human Genetics, 920 E. 58th Street, The University of Chicago, Chicago, IL 60637, USA2 Perlegen Sciences, Mountain View, CA 94043, USA3 Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA2005 10 12 2005 6 1 145 145 10 9 2005 10 12 2005 Copyright © 2005 Donfack et al; licensee BioMed Central Ltd.2005Donfack 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 Evolutionarily conserved sequences likely have biological function. Methods To determine whether variation in conserved sequences in non-coding DNA contributes to risk for human disease, we studied six conserved non-coding elements in the Th2 cytokine cluster on human chromosome 5q31 in a large Hutterite pedigree and in samples of outbred European American and African American asthma cases and controls. Results Among six conserved non-coding elements (>100 bp, >70% identity; human-mouse comparison), we identified one single nucleotide polymorphism (SNP) in each of two conserved elements and six SNPs in the flanking regions of three conserved elements. We genotyped our samples for four of these SNPs and an additional three SNPs each in the IL13 and IL4 genes. While there was only modest evidence for association with single SNPs in the Hutterite and European American samples (P < 0.05), there were highly significant associations in European Americans between asthma and haplotypes comprised of SNPs in the IL4 gene (P < 0.001), including a SNP in a conserved non-coding element. Furthermore, variation in the IL13 gene was strongly associated with total IgE (P = 0.00022) and allergic sensitization to mold allergens (P = 0.00076) in the Hutterites, and more modestly associated with sensitization to molds in the European Americans and African Americans (P < 0.01). Conclusion These results indicate that there is overall little variation in the conserved non-coding elements on 5q31, but variation in IL4 and IL13, including possibly one SNP in a conserved element, influence asthma and atopic phenotypes in diverse populations. ==== Body Background Comparison of human DNA sequences with those of other mammalian species is a powerful method for identifying functionally important sequence elements in the human genome because sequences with function tend to be evolutionarily conserved whereas those without function tend to accumulate variation over time. In fact, ~50% of the DNA sequences that are evolutionarily conserved between humans and mice lie outside of coding sequences of known genes [1]. Some of these conserved non-coding sequences have been shown to be long-range transcriptional regulatory elements participating in the temporal and tissue-specific expression patterns of genes [2,3]. Previous comparison of a 1 Mb region on human chromosome 5q31, which includes the cytokine genes encoding the T-helper 2 (Th2) cytokines, interleukin (IL)-4, IL-5, and IL-13, with the syntenic murine segment identified highly conserved non-coding sequences [4]. Examination of these conserved non-coding sequences in five additional mammalian species demonstrated that these elements are frequently conserved in all mammals. The longest conserved non-coding sequence, called CNS-1, is located between the IL4 and IL13 genes and showed a high degree of conservation across species [4]. Functional evaluation of CNS-1 in mutant mice revealed its role in the control of the global expression of IL4, IL5 and IL13, suggesting that CNS-1 acts as a coordinate regulator of these three genes [4,5]. This interval on human 5q31 is particularly intriguing because in addition to housing a cluster of genes encoding many Th2 cytokines, linkage to this region has been demonstrated with asthma-related phenotypes in at least six different populations [6-11]. Moreover, variation in the promoter, -589C/T (also referred to as -590C/T) [12], intron 2, +3017G/T [13], and 5'-untranslated region (UTR), +33C/T [14], of the IL4 gene and in the promoter, -1112C/T (also referred to as -1055C/T) [15], and coding region, Arg130Gln (also referred to as Arg110Gln) [16,17], of the IL13 gene have been associated with asthma and atopic phenotypes in many studies (reviewed in ref[18]. However, the specific variation that underlies the linkages described above has not been identified (reviewed in ref. [19]. It is likely, therefore, that additional variation in this interval contributes to susceptibility to both asthma and atopic phenotypes. In the present study, we screened six non-coding elements on 5q31 that are evolutionarily conserved between the human and murine genomes and are thus possible regulatory elements. We studied 10 polymorphisms across this region, including two within and two flanking conserved non-coding elements, and evaluated their relationship to asthma and atopy in members of a large Hutterite pedigree and in well-defined African American and European American patient populations. Methods Sample composition Conserved non-coding elements (Figure 1) were screened for SNPs in DNA from 10 African American and 10 European American unrelated controls, and from 10 individuals who are members of a founder population, the Hutterites. The 10 Hutterites were selected to represent distant branches of their pedigree but without regard to disease status. Figure 1 VISTA plot [24] displaying evolutionarily conserved sequences identified by the comparison of ~48 kb of human 5q31 DNA encoding the IL4, IL13 and KIF3A genes with murine sequences (BAC clone AF276990). On the horizontal axis, conserved sequences are plotted in relation to their position in the human reference sequence; kb distances are shown under the horizontal bar. The height of the peaks on the vertical axis indicates the level of conservation in percent identity between the human reference sequence and the murine sequences. Conserved sequences (>100 bp and >70% identity) defined as coding exons (dark blue), untranslated exons (light blue) and non-coding (red) are shown. The exons in each of three genes are shown as rectangle boxes; only the 3' end (exons 9 through 16) of KIF3A is shown. Six conserved non-coding elements were examined in this study (CNE-A - CNE-F). The SNPs identified or genotyped in this study and their approximate locations are shown. CNE-B corresponds to CNS-1 and CNE-F corresponds to CNS-2 described by Loots et al. [4]. Associations with asthma and atopy were evaluated in a large Hutterite pedigree [9] and in outbred individuals ascertained in Chicago. Six hundred thirty eight Hutterites were evaluated for asthma and atopy, as previously described [9]; 71 had a diagnosis of asthma, 156 were bronchial hyperresponsive to methacholine, and 311 were atopic. The Chicago samples included 205 African Americans and 126 European Americans with asthma and 388 control subjects with a negative personal and family history of asthma (183 African Americans and 205 European Americans). Subjects included in this study reported having had at least three grandparents who were either of African American or European ancestry. Given the allele frequencies observed in these samples (Table 4), we had 80% power to detect a relative risk of ≥ 1.7 in the African Americans and ≥ 2.2 in the European Americans [20]. Evaluation of phenotypes The Hutterites were evaluated for asthma and atopy using previously described protocols [9]. Exposure to cigarette smoke among the Hutterites was rare. The 331 unrelated asthma cases were recruited in Chicago as part of the Collaborative Study on the Genetics of Asthma (CSGA) and met the same diagnostic critieria as that used for the Hutterites [21,22]. Subjects with a history of cigarette smoking (>3 pack-year equivalent) were excluded from these studies. Atopy was defined by skin prick test. No clinical testing was performed on the control subjects. These protocols were approved by The University of Chicago Institutional Review Board; written consent was obtained from all subjects. Identification of conserved sequences An ~40 kb interval on human 5q31 was compared to the syntenic region in the mouse using AVID alignment programs [23] and visualized as a VISTA plot [24]. Conserved non-coding sequences were defined as having every contiguous subsegment of length 100 bp to be ≥ 70% identical to its paired sequence. These regions differ slightly from the earlier study [4] because in that study the CNE calculation was made using PIPMaker and here we used VISTA, which was developed after the Loots study. Identification of polymorphisms Amplified PCR products that included the conserved non-coding elements (Additional File, Table 1) were screened for polymorphisms by denaturing high performance liquid chromatography (DHPLC) [25], which detects nearly 100% of mutations in fragments of 600 bp or less [26-29]. PCR products with variant DHPLC patterns were sequenced; the complement of human BAC clone AC004039.1 was used as the reference sequence for identifying SNPs. Table 1 Distribution of SNPs identified in screening sets by ethnic group. The numbers show how many individuals in each group of 10 with the minor allele. AA = African American, EA = European American, HT = Hutterites. CNE Location in Bac clone AC004039.1 SNP Location in Bac clone AC004039.1 Population AA EA HT CNE-A 48566–48741 - - - - - SNP1-C/T 43038 3 0 0 CNE-B^ 42346–42674 - - - - - CNE-C 32694–33033 SNP2-C/T 32711 6 3 3 SNP3-C/T 31971 1 0 0 SNP4-G/A 31695 1 3 3 CNE-D 31406–31590 - - - - - SNP5-T/C 21794 1 0 0 CNE-E 21595–21737 - - - - - SNP6-C/T 21432 4 6 1 SNP7-G/A 21425 4 6 1 CNE-F* 17615–17863 SNP8-G/C 17713 2 4 1 *Corresponds to CNS-2 [4] ^Corresponds to CNS-1 [4] Genotyping The genotyping methods used in this study are described in Additional File Table 2. In addition to four SNPs in or flanking conserved sequences, we genotyped six known SNPS in the IL4 and IL13 genes to evaluate LD patterns between these genes and the CNEs and evaluate the relative magnitude of their effects. These SNPs were IL13_-1112C/T [15], IL13_+1923 [17], IL13_Arg130Gln (A/G) [16,17], IL4_-589C/T [12], IL4_+3017 [13], and IL4_+8374A/G (previously identified in our lab). Table 2 10-SNP haplotype frequencies in the Hutterites. Haplotypes were constructed manually (see Methods). Only individuals with complete haplotype information for both chromosomes are included (N = 1168 chromosomes). SNPs in the IL13 gene that are associated with IgE and +SPT in the Hutterites are in bold font. IL13 IL4 Intergenic Region Haplotype Frequency -1112C/T +1923C/T Arg130Gln (G→ A) -589C/T SNP2C/T SNP4G/A +3017G/T +8374A/G SNP7A/G SNP8C/G 1 0.660 C C G C C G G A A G 2 0.070 T C G C C G G A A G 3 0.008 C T A C C G G A A G 4 0.021 T T A C C G G A A G 5 0.021 C C G C C G T A G C 6 0.034 T T A C C G T A G C 7 0.019 C T A C C G T A G C 8 0.068 C C G T T A T G G C 9 0.098 T T A T T A T G G C Statistical analysis In the Hutterites, genotyping errors were detected using PEDCHECK [30] and deviations from Hardy-Weinberg equilibrium (HWE) were determined using an application modified to allow for related individuals [31]. To test for associations with SNPs and haplotypes, we used a case-control test developed for large pedigrees, as previously described [32]. Haplotypes comprised of 10 SNPs across the interval were constructed manually by the direct observation of alleles segregating in families. During haplotype construction, missing genotypes were filled in if they could be directly inferred from family data but no inferences were made regarding the haplotype composition when there was more than one possible haplotype. Two locus (pairwise) haplotypes were then generated from the larger 10 SNP haplotypes. We corrected for multiple comparisons using a Bonferonni correction for 4 SNPs and 6 pairwise haplotypes (see Results), and we considered significant P-values to be <0.0125 (0.05/4) and <0.00833 (0.05/6), respectively. Deviations from HWE and differences in allele and genotype frequencies between outbred cases and controls were examined using the program FINETTI [33]. Estimation of haplotype frequencies and testing for associations between cases and controls were conducted using the program FAMHAP [34]; 1,000 permutations were used to assess significance. If empiric P-values were <0.001, 10,000 permutations were performed. We used the Bonferonni correction for multiple comparisons (10 SNPs, P < 0.005; 45 pairwise comparisons, P < 0.0011). This is a conservative correction because these SNPs are not truly independent; some occur in the same gene and some are in LD. On the other hand, we did not correct for the number of phenotypes examined because these are also highly correlated. Within each ethnic group we compared the asthmatic and atopic cases to the non-asthmatic controls. Linkage disequilibrium LD plots were generated in the Chicago samples using publicly available software [35]. Results SNP discovery Six conserved non-coding sequences were identified in the interval between KIF3A and IL13 on human chromosome 5q31 (Figure 1). Of note is that none of the exons in either IL4 or IL13 are conserved between human and mouse, or between human and dog [36]. This is quite unusual (see for comparison, the pattern in KIF3A) and suggests possible divergence of function or accelerated rates of evolution of the human IL-4 and IL-13 proteins between humans and mice/dogs. Eight SNPs, referred to as SNP1-SNP8, were identified within or flanking the six conserved elements (Table 1). One SNP (SNP2) in CNE-C was identical to a previously reported SNP, +33C/T [14], and one SNP (SNP8) was in CNE-F. Six additional SNPs were identified in the sequences flanking CNE-B (SNP1), CNE-D (SNP3, SNP4), and CNE-E (SNP5, SNP6, SNP7). No variation was detected in CNE-B, which corresponds to CNS-1 in the Loots study and was previously shown to coordinately regulate IL4, IL5 and IL13 [4,5,37]. CNS-F, which corresponds to CNS-2 in the Loots study, harbored one variant (SNP8). We note that SNP4 resides within a conserved element in the IL4 gene that was identified by Dubchak and colleagues using human-dog sequence comparisons [36]. Furthermore, other than one rare SNP in Chinese (rs17772853; minor allele frequency 0.01), there is no additional variation in these regions reported in dbSNP [38] or in two previous studies of this region [39,40], suggesting that we identified all common variation in these conserved elements. A description of the eight SNPs and their distribution among the 30 individuals in the screening sample is shown in Table 1. Three SNPs (SNP1, SNP3 and SNP5) were present only in the African American sample. The remaining five SNPs (SNP2, SNP4, SNP6, SNP7 and SNP8) were present in all three groups. SNP6 and SNP7 were the only variants that appeared to be in perfect LD in all three screening samples. Because so few SNPs were discovered in the conserved non-coding elements and because one of the SNPs (SNP7) fell within a conserved element defined using different criteria in another study [40], we genotyped SNP2, SNP4, SNP7 and SNP8, in addition to three known variants each in IL4 and IL13. Patterns of linkage disequilibrium Nine haplotypes, comprised of 10 SNPs, were present in the Hutterites (Table 2). Three groups of SNPs were in perfect LD in this founder population: +1923C/T and Arg130Gln in the IL13 gene; -589C/T, SNP2C/T, SNP4G/A, and +8374A/G in the IL4 gene; and +3017G/T in the IL4 gene with intergenic SNP7A/G flanking CNE-E and SNP8G/C in CNE-F. For the remaining analyses in the Hutterites, therefore, we used only one SNP from each of these three LD groups, selecting the one with the most complete genotype information (+1923C/T, SNP2C/T, and +3017G/T), and one SNP that was not in perfect LD with any other SNP (-1112C/T). The patterns of pairwise LD between the 10 SNPs in the outbred samples are shown in Figures 2a (African Americans) and 2b (European Americans). In both control samples there was little long range LD (r2 ≤ 0.30) between SNPs in the IL4 and IL13 genes and relatively strong LD among SNPs within and in the 3' flanking region of the IL4 gene, similar to the pattern in the Hutterites (Table 2). The LD pattern among the African American cases and controls looked similar. However, among the European Americans, there was more LD between the IL4_-589 promoter SNP and other variants in the IL4 gene in the cases compared with controls. In the controls there was surprisingly little LD between IL4_-589 and other IL4 SNPs. Because there were few pairs of SNPs that showed perfect LD in the outbred samples and they differed between cases and controls, we analyzed all SNPs in the outbred samples. Figure 2 Pairwise LD plots (r2) for cases (lower half) and controls (upper half) in a) African Americans and b) European Americans. SNP studies in the hutterites The minor allele frequencies of SNPs in the Hutterites were: IL13_-1112T, 0.208; IL13_+1923T, 0.173; SNP2-T, 0.156; and IL4_+3017T, 0.226. Genotype proportions were in HWE at all loci (P > 0.01). In the single SNP analyses, there were modest associations between IL13_-1112T and asthma (P = 0.025), BHR (P = 0.028), and allergic sensitization to CR allergens (P = 0.032); and between SNP2-T with allergic sensitization to molds (P = 0.034). None of these were significant after correcting for multiple comparisons. However, highly significant associations were present between variation in the IL13 gene and sensitization to mold allergens (lL13_-1112T, P = 0.00067; IL13_+1923T, P = 0.0074), which remained significant after correcting for multiple comparisons. Moreover, only SNPs in IL13 were associated with total serum IgE, with a highly significant association between high IgE and the IL13_+1923T allele (P = 0.00022) and a more modest association with the lL13_-1112T allele (P = 0.014). Adjusting for allergic sensitization to molds in the analysis reduced the significance of the IL13_+1923T allele, but did not eliminate the association (P = 0.0085). To determine if susceptibility to asthma or atopic phenotypes is determined by combinations of SNPs across this interval or by specific haplotypes, we examined pairwise combinations of the four SNPs (Table 3). Highly significant associations (P < 0.001) with +SPT to mold allergens were observed with haplotypes comprised of either SNP in the IL13 gene (-1112C/T or +1923C/T) and SNP2 in the IL4 gene. Less significant associations were observed with these same pairwise combinations and allergic sensitization to cockroach allergen. However, in all of these analyses, the haplotypes carrying the common alleles at the IL13 SNPs (-1112C and +1923C) were underrepresented in the cases compared with controls, and the two haplotypes carrying the minor alleles at the IL13 SNPs (-1112T and +1923T) were overrepresented in the cases compared with controls, regardless of the allele at SNP2 (i.e., C or T). Therefore, the results of the haplotype analyses suggest that the IL13 SNPs are primarily associated with allergic sensitization to mold and cockroach allergens, and that variation in the IL4 gene is not contributing to this association in the Hutterites. None of the haplotypes were associated with asthma, BHR, or the other atopic phenotypes. Thus, in the Hutterites, variation in the IL13 gene is strongly associated with total serum IgE and allergic sensitization to mold allergens, and to a lesser extent to cockroach allergens, but not to any of the other phenotypes. None of the SNPs in or near conserved non-coding sequences contributed to susceptibility in the Hutterites. Table 3 Results of 2-SNP haplotype analyses in the Hutterites. In this sample, IL13_+1923C/T is in perfect LD (r2 = 1) with IL13_Arg130Gln (G→ A); SNP2 C/T is in perfect LD with IL4_-589C/T, SNP4G/A, and IL4_+8374A/G; IL4_+3017G/T is in perfect LD with SNP7A/G and SNP8C/G (Table 2). Only haplotypes and phenotypes with at least one P-value < 0.05 are shown. The number of cases in each analysis is shown in parentheses. P-values that were significant after adjusting for multiple comparisons are in bold font. Specific IgE Response (+SPT) to Locus 1 Locus 2 Mold (N = 75) Cockroach (N = 148) IL13_--1112 IL13_+1923 0.0063 0.042 SNP2 0.00020 0.0022 IL4_+3017 0.033 0.0074 IL13_+1923 SNP2 0.00033 0.0017 IL4_+3017 0.017 0.042 Table 4 Allele and genotype frequencies in case and control samples. The number of individuals in each sample is shown in parentheses. Not all individuals in the samples were genotyped for all SNPs. Only SNP2-T showed a modest association with allergic sensitization to mold allergens in European Americans (Puncorrected = 0.04). European Americans African Americans SNP All Asthma +SPT Any Allergen +SPT Molds Controls All Asthma +SPT Any Allergen +SPT Molds Controls (N = 126) (N = 82) (N = 46) (N = 205) (N = 205) (N = 160) (N = 68) (N = 183) IL13_-1112T 0.259 0.264 0.275 0.213 0.418 0.409 0.375 0.407 CC 0.576 0.586 0.550 0.613 0.337 0.336 0.391 0.360 CT 0.330 0.300 0.350 0.348 0.489 0.510 0.469 0.467 TT 0.094 0.114 0.100 0.039 0.174 0.154 0.140 0.173 IL13_+1923T 0.245 0.282 0.238 0.227 0.666 0.664 0.672 0.660 CC 0.575 0.535 0.575 0.584 0.087 0.094 0.094 0.133 CT 0.359 0.366 0.375 0.376 0.495 0.483 0.469 0.413 TT 0.066 0.099 0.050 0.040 0.418 0.423 0.437 0.454 IL13_Gln (T) 0.221 0.254 0.272 0.202 0.192 0.191 0.161 0.169 Arg/Arg (C/C) 0.618 0.577 0.625 0.620 0.643 0.639 0.695 0.686 Arg/Gln (C/T) 0.324 0.338 0.325 0.355 0.330 0.340 0.288 0.290 Gln/Gln (T/T) 0.058 0.085 0.05 0.025 0.027 0.021 0.077 0.024 IL4_-589T 0.193 0.197 0.163 0.163 0.677 0.678 0.687 0.643 CC 0.670 0.662 0.700 0.702 0.115 0.128 0.125 0.133 CT 0.274 0.282 0.275 0.270 0.397 0.389 0.375 0.447 TT 0.056 0.056 0.025 0.028 0.478 0.483 0.5 0.420 CNE-C_SNP2-T 0.196 0.223 0.257 0.146 0.405 0.419 0.404 0.382 CC 0.667 0.600 0.541 0.730 0.333 0.316 0.333 0.382 CT 0.294 0.354 0.405 0.247 0.524 0.530 0.526 0.471 TT 0.049 0.046 0.054 0.023 0.143 0.154 0.141 0.147 SNP4-A 0.176 0.186 0.187 0.175 0.399 0.412 0.415 0.359 GG 0.692 0.661 0.656 0.675 0.340 0.323 0.321 0.415 GA 0.264 0.305 0.313 0.300 0.525 0.531 0.528 0.452 AA 0.044 0.034 0.031 0.025 0.135 0.146 0.151 0.132 IL4_+3017T 0.327 0.339 0.387 0.308 0.850 0.867 0.825 0.805 GG 0.490 0.452 0.382 0.478 0.045 0.061 0.067 0.051 TG 0.367 0.419 0.471 0.429 0.209 0.217 0.217 0.288 TT 0.143 0. 129 0.147 0.093 0.746 0. 727 0.716 0.661 IL4_+8374G 0.161 0.164 0.135 0.159 0.275 0.274 0.259 0.257 AA 0.708 0.687 0.730 0.703 0.526 0.533 0.554 0.541 AG 0.261 0.299 0.270 0.275 0.398 0.387 0.375 0.404 GG 0.031 0.014 0 0.022 0.076 0.080 0.071 0.055 SNP7-G 0.279 0.318 0.319 0.317 0.636 0.638 0.614 0.568 AA 0.548 0.485 0.472 0.458 0.156 0.152 0.123 0.213 AG 0.347 0.394 0.417 0.450 0.416 0.421 0.526 0.460 GG 0.105 0.121 0.111 0.092 0.428 0.427 0.351 0.327 CNE-F_SNP8-C 0.323 0.339 0.386 0.293 0.618 0.623 0.598 0.577 CC 0.131 0.113 0.143 0.081 0.393 0.401 0.328 0.335 CG 0.384 0.452 0.486 0.422 0.450 0.444 0.541 0.484 GG 0.485 0.435 0.371 0.497 0.157 0.156 0.131 0.181 Studies in outbred case-control samples Allele and genotype frequencies of the 10 SNPs in 337 subjects with asthma and 388 non-asthmatic controls are shown in Table 4 by ethnicity and phenotype. Genotypes were in HWE in the African American and European American control samples (P > 0.01). In the single SNP analyses, there was only a modest association between SNP2-T and allergic sensitization to mold allergens in European Americans (P = 0.04), which was not significant after adjusting for multiple comparisons. However, pairwise combinations of SNPs in the IL4 gene were significantly associated with asthma and allergic sensitization, primarily in the European American sample (Table 5). In that sample, nearly all of the haplotypes that were associated with asthma and the one most strongly associated with atopy included the IL4_-589T allele and other SNPs in the IL4 gene (SNP2-T, SNP4-A, IL4_+3017T, IL4_+8374G, SNP8-C). A haplotype comprised of the IL13_+1923T and IL13_130Gln alleles was also strongly associated with asthma in this sample. All but one of the seven associations remained significant after adjusting for multiple comparisons. In the African Americans, the frequencies of the IL13_-1112T/IL13_+1923T and IL13_-1112T/IL4_+3017T haplotypes were increased in cases with allergic sensitization to mold (P = 0.009 and 0.005, respectively), although this was not significant after correcting for multiple comparisons. However, because some of the controls may have been SPT+ to mold allergens, this is a conservative test. Similar to the Hutterites, there were no associations with asthma or SPT to any allergen or with combinations of SNPs in the IL4 gene in the African Americans. Table 5 Results of 2-SNP haplotype analysis in a) European Americans cases and controls, and b) African American cases and controls. Empiric P-values are based on 1,000 or 10,000 permutations; only haplotypes with at least one P-value < 0.05 are shown. P-values that were significant after adjusting for multiple (45) comparisons are in bold font; ns, not significant (P ≥ 0.05). Locus 1 Locus 2 Asthma vs. Controls +SPT to Any Allergen vs. Controls +SPT Molds vs. Controls a) European Americans IL13_-1112 IL4_+3017 ns ns 0.0128 IL13_+1923 IL13_130 <10-4 0.0011 0.0121 IL13_130 SNP2 ns 0.0295 0.0101 IL4_-589 SNP2 <10-4 0.0004 0.0026 SNP4 <10-4 <10-4 0.0005 IL4_+3017 0.0003 0.0030 0.0269 IL4_+8374 <10-4 <10-4 0.0002 SNP7 0.0021 0.0279 0.0335 SNP8 0.0002 0.0072 0.0155 SNP2 SNP4 ns 0.0240 0.0042 IL4_+3017 ns 0.0131 <10-4 IL4_+8374 ns 0.0252 0.0054 SNP7 ns ns 0.0019 SNP8 ns ns 0.0128 IL4_+3017 SNP7 ns ns 0.014 SNP7 SNP8 ns ns 0.014 b) African Americans IL13_-1112 IL13_+1923 ns ns 0.009 IL4_+3017 0.034 ns 0.005 Discussion and conclusions Cross-species comparisons are powerful tools for identifying potential functional elements in non-coding DNA [3,4,36,41-44]. However, it is unknown whether conserved non-coding elements in the human genome harbor variation that contributes to inter-individual differences in susceptibility to common diseases. To address this question, we surveyed variation in six conserved non-coding elements in the Th2 cytokine gene cluster on chromosome 5q31 to determine whether such variation, if it exists, is associated with susceptibility to asthma-related phenotypes. Only one of these conserved non-coding elements, CNS-1 (CNE-B in our study), has been shown to have regulatory properties: the deletion of CNS-1 in transgenic mice results in the reduction of human IL-4, IL-5 and IL-13 producing cells [5,37]. Similar to our results, neither Noguchi et al. [39] nor Banerjee et al. [40] found sequence variation in CNS-1 in 48 individuals of Japanese origin [39] or in 17 individuals of African origin and 23 individuals of European origin [40]. These results combined with ours indicate that CNS-1 is highly conserved among humans and is under strong selective constraints, consistent with its role as a cis-acting regulatory element. CNS-2 (CNE-F in our study) was also among the most conserved non-coding elements identified in a comparison of human 5q31 DNA with conserved syntenic mouse sequences, second only to CNS-1 [4]. We found one SNP in this element (SNP8), similar to the study of Banerjee [40]. However, this variant was not associated with asthma or atopy in the Hutterites or outbred case-control samples. However, we note that SNP8 in CNE-F (CNS-2) is in very strong LD with IL4_+3017, which was associated with IgE levels in a previous study in Caucasian subjects [13]. We did not find any variation in CNE-E, although one rare and two common SNPs (SNP5 and SNP6, SNP7, respectively) were identified just outside the boundaries of this element. SNP7 was in a conserved element defined by Banerjee, but this SNP was also not associated with asthma or atopy in our study. Only SNP2 (+33C/T) in CNE-C was associated with asthma and atopy, and only when considered in combination with other SNPs in the IL4 gene. This SNP was previously associated with IgE levels in Japanese (P < 0.05) [14]. However, our results indicate that either combinations of SNPs in and near the IL4 gene act synergistically to influence susceptibility, or other variation on a haplotype that includes the -589-T, SNP2-T, SNP4-G, +3017-T, +8374-G, and SNP8-C alleles influences susceptibility. In either case, the variation in IL4 that influences asthma and atopy resides in non-coding regions. Similarly, the -589-T and +3017-T alleles, which have been associated with asthma and/or atopy in other studies [12,13,45-50], do not by themselves or in combination with each other account for the associations observed in this study. Lastly, we identified an association between variation in the IL13 gene and allergic sensitization to mold allergens in the Hutterites, which was also present, albeit to a lesser degree, in two outbred populations. Associations of other atopic phenotypes with two functional polymorphisms [15,51] in IL13 have been reported previously [15-17,52-55], but this is the first report of a specific association with +SPT to molds. Haplotypes comprised of SNPs in the IL13 gene were also associated with +SPT to mold allergens in the African American and European American samples, suggesting that either these SNPs interact to confer risk or additional variation in this gene also contributes. In addition, the +1923T and/or 130Gln alleles were also very strongly associated with total serum IgE (as a quantitative trait) in the Hutterites. The association with IgE was only partially accounted for by mold sensitization, indicating a role for this gene in IgE mediated immune responses, consistent with studies in other populations [16,17,54,55]. The fact that we identified associations between variation in the IL13 gene and atopy in all three populations (and with asthma in the European Americans), but between variation in the IL4 gene and asthma only in the European Americans, reflects the complexity of genetic susceptibility to asthma and atopy. It is notable that allele frequencies at SNPs across this interval differed considerably between the African American and European American samples (Table 4). For example, the minor allele in the European American sample was the more common (major) allele in the African American sample at five loci (IL13_+1923, IL4_-589, IL4_+3017, SNP7, CNE-F_SNP8). At nearly all other loci, the allele frequencies were more even (i.e., closer to 50%) in the African American than in the European American sample. Furthermore, although the overall pattern of LD was similar in the African American and European American control subjects (Figure 2), there was more LD between the -589C/T alleles with alleles at other IL4 SNPs in the European American cases compared with controls. The latter is the expected pattern at a disease locus [56], and is consistent with the highly significant associations that we observed between IL4 haplotypes and asthma in the European Americans. These differences in allele frequencies and LD patterns may have reduced our power detect associations in the African American sample, particularly with respect to untyped SNPs that might be in LD with IL4_-589. Alternatively, the observation that no one SNP or combination of SNPs is penetrant in all populations may reflect the modifying effects of background genes or environmental exposures on risk [57, 58]. This possibility is supported by a genome-wide linkage study of asthma in which different linkage signals were detected in Caucasian and African American families, despite the fact that both groups were evaluated using identical protocols and ascertained at the same centers [10,21]. These results highlight the challenges in elucidating the genetic architecture of complex diseases, which is likely to differ among individuals with different environmental exposures and different genetic backgrounds, some of which is captured by racial/ethnic ancestry. In summary, these data indicate that the conserved non-coding elements on human chromosome 5q31 in the interval including the IL13 and IL4 genes do not contain variation that influences disease risk among individuals. SNP2 (+33C/T), in a conserved element (CNE-C) in the IL4 gene, may influence susceptibility in combination with other variation in IL4, or may merely be in LD with other variation in the gene that influences susceptibility to asthma and atopic phenotypes. Additional studies are required to differentiate between these alternatives, to fully characterize the functional variants in this region that influence disease risk, and to provide a model for understanding the role of non-coding variation on gene function and disease susceptibility. List of abbreviations AA African Americans CNE conserved non-coding element CSGA Collaborative Study on the Genetics of Asthma DHPLC denaturing high performance liquid chromatography EA European Americans HWE Hardy-Weinberg equilibrium IgE immunoglobulin E LD linkage disequilibrium IL4 interleukin 4 IL13 interleukin 13 PCR polymerase chain reaction SNP single nucleotide polymorphism SPT skin prick test Th2 T-helper 2 Competing interests The author(s) declare that they have no competing interests. Authors' contributions Dr. Donfack and Mr. Schneider designed primers and performed all of the DHPLC, sequencing and genotyping analyses. Dr. Tan and Dr. Kurz performed the LD studies and all data analyses. Dr. Dubchak and Dr. Frazer performed the VISTA analyses, defined the conserved region, and provided unpublished sequence data. Dr. Ober conceived and designed the study and, with Dr. Donfack, wrote the manuscript. All authors contributed comments to various drafts of the manuscript. Supplementary Material Additional File 1 Table 1. Primer sequences and PCR conditions. Table 2. Genotyping methods. Click here for file Acknowledgements We thank Harvey Dytch, Natasha Phillips, Rebecca Anderson, Lin Pan, Lauren Weiss, Sarah Diacon, Fionnuala Hickey, and Dina Newman for technical assistance. This work was supported by NIH grants HL49596, HL56399, and HL66533 to C.O. and The University of Chicago General Clinical Research Center (NIH grant M01 RR00055). ==== Refs Frazer KA Sheehan JB Stokowski RP Chen X Hosseini R Cheng JF Fodor SP Cox DR Patil N Evolutionarily conserved sequences on human chromosome 21 Genome Res 2001 11 10 1651 1659 11591642 Nobrega MA Ovcharenko I Afzal V Rubin EM Scanning human gene deserts for long-range enhancers Science 2003 302 5644 413 14563999 Boffelli D Nobrega MA Rubin EM Comparative genomics at the vertebrate extremes Nat Rev Genet 2004 5 6 456 465 15153998 Loots GG Locksley RM Blankespoor CM Wang ZE Miller W Rubin EM Frazer KA Identification of a coordinate regulator of interleukins 4, 13, and 5 by cross-species sequence comparisons. Science 2000 288 136 140 10753117 Mohrs M Blankespoor CM Wang ZE Loots GG Afzal V Hadeiba H Shinkai K Rubin EM Locksley RM Deletion of a coordinate regulator of type 2 cytokine expression in mice Nat Immunol 2001 2 842 847 11526400 Meyers DA Postma DS Panhuysen CIM Evidence for a locus regulating total serum IgE levels mapping to chromosome 5 Genomics 1994 23 464 470 7835897 Marsh DG Neely JD Breazeale DR Ghosh B Freidhoff LR Schou C Beaty TH Genetic basis of IgE responsiveness: relevance to the atopic diseases Int Arch Allergy Immunol 1995 107 25 28 7613143 Martinez FD Solomon S Holberg CJ Graves PA Baldini M Erickson RP Linkage of circulating eosinophils to markers on chromosome 5q Am J Resp Care Med 1998 158 1739 1744 Ober C Tsalenko A Parry R Cox NJ A second-generation genomewide screen for asthma-susceptibility alleles in a founder population Am J Hum Genet 2000 67 1154 1162 11022011 Xu J Meyers DA Ober C Blumenthal MN Mellen B Barnes KC King RA Lester LA Howard TD Solway J Langefeld CD Beaty TH Rich SS Bleecker ER Cox NJ Genomewide screen and identification of gene-gene interactions for asthma-susceptibility loci in three U.S. populations: Collaborative Study on the Genetics of Asthma Am J Hum Genet 2001 68 6 1437 1446 11349227 Yokouchi Y Nukaga Y Shibasaki M Noguchi E Kimura K Ito S Nishihara M Yamakawa-Kobayashi K Takeda K Imoto N Ichikawa K Matsui A Hamaguchi H Arinami T Significant evidence for linkage of mite-sensitive childhood asthma to chromosome 5q31-q33 near the interleukin 12 B locus by a genome-wide search in Japanese families Genomics 2000 66 2 152 160 10860660 Rosenwasser LJ Klemm DJ Dresback JK Inamura H Mascali JJ Klinnert M Borish L Promoter polymorphisms in the chromosome 5 gene cluster in asthma and atopy Clin Exp Allergy 1995 25 (Suppl 2) 74 78 8590350 Basehore MJ Howard TD Lange LA Moore WC Hawkins GA Marshik PL Harkins MS Meyers DA Bleecker ER A comprehensive evaluation of IL4 variants in ethnically diverse populations: association of total serum IgE levels and asthma in white subjects J Allergy Clin Immunol 2004 114 1 80 87 15241348 Suzuki I Hizawa N Yamaguchi E Kawakami Y Association between a C+33T polymorphism in the IL-4 promoter region and total serum IgE levels Clin Exp Allergy 2000 30 1746 1749 11122213 van der Pouw Kraan TCTM van Veen A Boeije LCM van Yurl SAP de Groot ER Stapel SO Bakker A Verweij CL Aarden LA van der Zee JS An IL-13 promoter polymorphism associated with increased risk of allergic asthma Genes and Immunity 1999 1 61 65 11197307 Heinzmann A Mao XQ Akaiwa M Kreomer RT Gao PS Ohshima K Umeshita R Abe Y Braun S Yamashita T Roberts MH Sugimoto R Arima K Arinobu Y Yu B Kruse S Enomoto T Dake Y Kawai M Shimazu S Sasaki S Adra CN Kitaichi M Inoue H Yamauchi K Tomichi N Kurimoto F Hamasaki N Hopkin JM Izuhara K Shirakawa T Deichmann KA Genetic variants of IL-13 signalling and human asthma and atopy Hum Mol Genet 2000 9 549 559 10699178 Graves PE Kabesch M Halonen M Holberg CJ Baldini M Fritzsch C Weiland SK Erickson RP von Mutius E Martinez FD A cluster of seven tightly linked polymorphisms in the IL-13 gene is associated with total serum IgE levels in three populations of white children J Allergy Clin Immunol 2000 105 506 513 10719301 Hoffjan S Nicolae D Ober C Association studies for asthma and atopic diseases: a comprehensive review of the literature Respir Res 2003 4 1 14 28 14748924 Hoffjan S Ober C Present status on the genetics of asthma Curr Opin Immunol 2002 14 6 709 717 12413520 Schlesselman JJ Case-Control Studies 1982 New York , Oxford University Press 354 CSGA A genome-wide search for asthma susceptibility loci in ethnically diverse populations Nature Genetics 1997 15 389 392 9090385 Lester LA Rich SS Blumenthal MN Togias A Murphy S Malveaux F Miller ME Dunston GM Solway J Wolf R Samet JM Marsh DG Meyers DA Ober C Bleecker ER Ethnic differences in asthma and associated phenotypes: Collaborative Study on the Genetics of Asthma J Allergy Clin Immunol 2001 108 357 362 11544453 Bray N Dubchak I Pachter L AVID: A global alignment program Genome Res 2003 13 1 97 102 12529311 Frazer KA Pachter L Poliakov A Rubin EM Dubchak I VISTA: computational tools for comparative genomics Nucleic Acids Res 2004 32 Web Server issue W273 279 15215394 Oefner PJ Underhill PA Comparative DNA sequencing by denaturing high-performance liquid chromatography (DHPLC) Am J Hum Genet Suppl 1995 57 A266 Han SS Cooper DN Upadhyaya MN Evaluation of denaturing high performance liquid chromatography (DHPLC) for the mutational analysis of the neurofibromatosis type 1 ( NF1) gene Hum Genet 2001 109 5 487 497 11735023 Lin D Goldstein JA Mhatre AN Lustig LR Pfister M Lalwani AK Assessment of denaturing high-performance liquid chromatography (DHPLC) in screening for mutations in connexin 26 (GJB2) Hum Mutat 2001 18 1 42 51 11438992 Roberts PS Jozwiak S Kwiatkowski DJ Dabora SL Denaturing high-performance liquid chromatography (DHPLC) is a highly sensitive, semi-automated method for identifying mutations in the TSC1 gene J Biochem Biophys Methods 2001 47 1-2 33 37 11179759 Ravnik-Glavac M Atkinson A Glavac D Dean M DHPLC screening of cystic fibrosis gene mutations Hum Mutat 2002 19 4 374 383 11933191 O'Connell JR Weeks DE PedCheck: A program for identification of genotype incompatibilities in linkage analysis Am J Hum Genet 1998 63 259 266 9634505 Bourgain C Abney M Schneider D Ober C McPeek MS Testing for hardy-weinberg equilibrium in samples with related individuals Genetics 2004 168 4 2349 2361 15371359 Bourgain C Hoffjan S Nicolae R Newman D Steiner L Walker K Reynolds R Ober C McPeek MS Novel case-control test in a founder population identifies p-selectin as an atopy-susceptibility locus Am J Hum Genet 2003 73 3 612 626 12929084 http://ihg.gsf.de/linkage/download/finetti.zip Becker T Knapp M Maximum-likelihood estimation of haplotype frequencies in nuclear families Genet Epidemiol 2004 27 1 21 32 15185400 Dubchak I Brudno M Loots GG Pachter L Mayor C Rubin EM Frazer KA Active Conservation of Noncoding Sequences Revealed by Three-Way Species Comparisons Genome Research 2000 10 1304 1306 10984448 Lee GR Fields PE Flavell RA Regulation of IL-4 gene expression by distal regulatory elements and GATA-3 at the chromatin level Immunity 2001 14 447 459 11336690 Noguchi E Nukaga-Nishio Y Jian Z Yokouchi Y Kamioka M Yamakawa-Kobayashi K Hamaguchi H Matsui A Shibasaki M Arinami T Haplotypes of the 5' region of the IL-4 gene and SNPs in the intergene sequence between the IL-4 and IL-13 genes are associated with atopic asthma. Hum Immunol 2001 62 1251 1257 11704288 Banerjee P Bahlo M Schwartz JR Loots GG Houston KA Dubchak I Speed TP Rubin EM SNPs in putative regulatory regions identified by human mouse comparative sequencing and transcription factor binding site data Mamm Genome 2002 13 10 554 557 12420132 Shamsher MK Chuzhanova NA Friedman B Scopes DA Alhaq A Millar DS Cooper DN Berg LP Identification of an intronic regulatory element in the human protein C (PROC) gene Hum Genet 2000 107 458 465 11140943 Pennacchio LA Olivier M Hubacek JA Cohen JC Cox DR Fruchart JC Krauss RM Rubin EM An apolipoprotein influencing triglycerides in humans and mice revealed by comparative sequencing Science 2001 294 5540 169 173 11588264 Rogers HJ Bate N Combe J Sullivan J Sweetman J Swan C Lonsdale DM Twell D Functional analysis of cis-regulatory elements within the promoter of the tobacco late pollen gene g10 Plant Mol Biol 2001 45 577 585 11414616 Koch MA Weisshaar B Kroymann J Haubold B Mitchell-Olds T Comparative genomics and regulatory evolution: conservation and function of the Chs and Apetala3 promoters Mol Biol Evol 2001 18 1882 1891 11557794 Walley AJ Cookson WO Investigation of an interleukin-4 promoter polymorphism for associations with asthma and atopy. J Med Genet 1996 33 689 692 8863163 Kawashima T Noguchi E Arinami T Yamakawa-Kobayashi K Nakagawa H Otsuka F Hamaguchi H Linkage and association of an interleukin 4 gene polymorphism with atopic dermatitis in Japanese families J Med Genet 1998 35 502 504 9643293 Noguchi E Shibasaki M Arinami T Takeda K Yokouchi Y Kawashima T Yanagi H Matsui A Hamaguchi H Association of asthma and the interleukin-4 promoter gene in Japanese Clin Exp Allergy 1998 28 449 453 9641571 Sandford AJ Chagani T Zhu S Weir TD Bai TR Spinelli JJ Fitzgerald JM Behbehani NA Tan WC Pare PD Polymorphisms in the IL4, IL4RA, and FCERIB genes and asthma severity J Allergy Clin Immunol 2000 106 135 134 10887316 Burchard EG Silverman EK Rosenwasser LJ Borish L Yandava C Pillari A Weiss ST Hasday J Lilly CM Ford JG Drazen JM Association between a sequence variant in the IL-4 gene promoter and FEV(1) in asthma Am J Respir Crit Care Med 1999 160 3 919 922 10471619 Zhu S Chan-Yeung M Becker AB Dimich-Ward H Ferguson AC Manfreda J Watson WT Pare PD Sandford AJ Polymorphisms of the IL-4, TNF-alpha, and Fcepsilon RIbeta genes and the risk of allergic disorders in at-risk infants Am J Respir Crit Care Med 2000 161 5 1655 1659 10806171 Vladich FD Brazille SM Stern D Peck ML Ghittoni R Vercelli D IL-13 R130Q, a common variant associated with allergy and asthma, enhances effector mechanisms essential for human allergic inflammation J Clin Invest 2005 115 3 747 754 15711639 Liu X Nickel R Beyer K Wahn U Ehrlich E Freidhoff LR Bjorksten B Beaty TH Huang SK An IL13 coding region variant is associated with a high total serum IgE level and atopic dermatitis in the German multicenter atopy study (MAS- 90) J Allergy Clin Immunol 2000 106 1 Pt 1 167 170 10887320 Howard TD Whittaker PA Zaiman AL Koppelman GH Xu J Hanley MT Meyers DA Postma DS Bleecker ER Identification and association of polymorphisms in the interleukin-13 gene with asthma and atopy in a Dutch population. Am J Respir Cell Mol Biol 2001 25 377 384 11588017 Leung TF Tang NL Chan IH Li AM Ha G Lam CW A polymorphism in the coding region of interleukin-13 gene is associated with atopy but not asthma in Chinese children Clin Exp Allergy 2001 31 10 1515 1521 11678850 DeMeo DL Lange C Silverman EK Senter JM Drazen JM Barth MJ Laird N Weiss ST Univariate and multivariate family-based association analysis of the IL-13 ARG130GLN polymorphism in the Childhood Asthma Management Program Genet Epidemiol 2002 23 4 335 348 12432502 Hayes MG Tsuchiya T del Bosque-Plata L Cox NJ Patterns of linkage disequilibrium in calpain-10 Am J Hum Genet 2004 73S 387 Marchini J Donnelly P Cardon LR Genome-wide strategies for detecting multiple loci that influence complex diseases Nat Genet 2005 37 4 413 417 15793588 Ober C Perspectives on the past decade of asthma genetics J Allergy Clin Immunol 2005 116 2 274 278 16083779	16336695	PMC1325232	CC BY	2021-01-04 16:23:24	no	Respir Res. 2005 Dec 10; 6(1):145	utf-8	Respir Res	2,005	10.1186/1465-9921-6-145	oa_comm
==== Front Nutr Metab (Lond)Nutrition & Metabolism1743-7075BioMed Central London 1743-7075-2-331630068210.1186/1743-7075-2-33ResearchPhosphoenolpyruvate carboxykinase and the critical role of cataplerosis in the control of hepatic metabolism Hakimi Parvin [email protected] Mark T [email protected] Jianqi [email protected] David F [email protected] Ronald A [email protected] Satish C [email protected] Lea [email protected] Shirley M [email protected] Richard W [email protected] Department of Biochemistry, Case Western Reserve University School of Medicine, Cleveland, OH, USA2 Department of Pediatrics, Case Western Reserve University School of Medicine, Cleveland, OH, USA3 Department of Genetics, Case Western Reserve University School of Medicine, Cleveland, OH, USA4 Schwartz Center for Metabolism and Nutrition, MetroHealth Medical Center, Cleveland, OH, USA5 Department of Biochemistry, Hebrew University-Hadassah Medical School, Jerusalem, Israel6 Department of Molecular Biology, Princeton University, Princeton, NJ, USA2005 21 11 2005 2 33 33 21 10 2005 21 11 2005 Copyright © 2005 Hakimi et al; licensee BioMed Central Ltd.2005Hakimi 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 metabolic function of PEPCK-C is not fully understood; deletion of the gene for the enzyme in mice provides an opportunity to fully assess its function. Methods The gene for the cytosolic form of phosphoenolpyruvate carboxykinase (GTP) (EC 4.1.1.32) (PEPCK-C) was deleted in mice by homologous recombination (PEPCK-C-/- mice) and the metabolic consequences assessed. Results PEPCK-C-/- mice became severely hypoglycemic by day two after birth and then died with profound hypoglycemia (12 mg/dl). The mice had milk in their stomachs at day two after birth and the administration of glucose raised the concentration of blood glucose in the mice but did not result in an increased survival. PEPCK-C-/- mice have two to three times the hepatic triglyceride content as control littermates on the second day after birth. These mice also had an elevation of lactate (2.5 times), β-hydroxybutyrate (3 times) and triglyceride (50%) in their blood, as compared to control animals. On day two after birth, alanine, glycine, glutamine, glutamate, aspartate and asparagine were elevated in the blood of the PEPCK-C-/- mice and the blood urea nitrogen concentration was increased by 2-fold. The rate of oxidation of [2-14C]-acetate, and [5-14C]-glutamate to 14CO2 by liver slices from PEPCK-C-/- mice at two days of age was greatly reduced, as was the rate of fatty acid synthesis from acetate and glucose. As predicted by the lack of PEPCK-C, the concentration of malate in the livers of the PEPCK-C-/- mice was 10 times that of controls. Conclusion We conclude that PEPCK-C is required not only for gluconeogenesis and glyceroneogenesis but also for cataplerosis (i.e. the removal of citric acid cycle anions) and that the failure of this process in the livers of PEPCK-C-/- mice results in a marked reduction in citric acid cycle flux and the shunting of hepatic lipid into triglyceride, resulting in a fatty liver. ==== Body Background The factors that regulate the transcription of the gene for the cytosolic form of phosphoenolpyruvate carboxykinase (GTP) (EC 4.1.1.32) (PEPCK-C) have been studied in great detail over the years [1]. However, its metabolic role in the various tissues in which the gene is expressed has received far less attention. Discussion of the function of PEPCK-C in mammals has largely been confined to its role in gluconeogenesis, with only a brief mention of alternative possibilities. PEPCK-C has become an important marker for hepatic gluconeogenesis; studies of the mechanisms involved in diabetes and related diseases, often include an analysis of the alterations in the mRNA for PEPCK-C. What is seldom mentioned, however, is the fact that PEPCK-C activity is present in a number of tissues that do not make glucose, including white and brown adipose tissue, mammary gland during lactation, small intestine, brain, lung, muscle and a number of others (see [2] for a review of this subject). Besides its role in gluconeogenesis, PEPCK-C is involved in glyceroneogenesis in adipose tissue [3] and liver [4,5]. This pathway is an abbreviated version of gluconeogenesis and results in the synthesis of glyceride-glycerol from precursors other than glucose or glycerol [6,7]. PEPCK-C can also catalyze the formation of oxalacetate, a citric acid cycle intermediate; it has been suggested that the enzyme "refills" the cycle during periods of biosynthesis (a process known as anaplerosis). However, the major biosynthetic tissues such as the liver and adipose tissue have considerable activity of pyruvate carboxylase, which itself generates oxalacetate. PEPCK-C is also the major cataplerotic enzyme, (cataplerotic enzymes remove citric acid cycle anions that are formed by the entry of the carbon skeletons of amino acids into the cycle). Both gluconeogenesis and glyceroneogenesis are cataplerotic processes, which use intermediates of the citric acid cycle for their specific biosynthetic process. As we will show in this paper, cataplerosis is a critical metabolic function. Finally, Hahn and Nowak [8] proposed that in brown adipose tissue a cycle operates between the mitochondria and the cytosol in which PEPCK-C converts GTP to GDP, to insure a continued citric acid cycle flux (GDP is a substrate for succinyl CoA synthase). Ablating the gene for PEPCK-C in mice provides an opportunity to better assess the metabolic role of the enzyme in mammalian tissues. The gene for PEPCK-C has been deleted specifically in the livers of mice by She et al [9], who noted that the mice had a greatly reduced rate of hepatic gluconeogenesis from a variety of precursors. However, the animals did not develop hypoglycemia, even after as much as 48 hrs of fasting and could be induced to develop diabetes [10]. The absence of PEPCK-C in the liver also resulted in a decrease in the level of glycogen in both the liver and muscle; there was also diminished whole-body glucose turnover. Mice in which the gene for PEPCK-C has been totally deleted in all tissues die in the first two days after birth with profound hypoglycemia. These findings indicate that the kidney (the second gluconeogenic organ) can make sufficient glucose to provide for the needs of the animal during fasting and that it responds to the hormonal alterations characteristic of diabetes by increasing renal glucose output. A surprising finding from the studies of She et al [9] was the observation that the ablation of the gene for PEPCK-C caused the mice to develop a fatty liver. As we will report here, the fatty liver appears as early as the end of the first day after birth and is accompanied by a 5-fold elevation in the concentration of triglyceride in the blood. While not expected by conventional metabolic logic, the accumulation of triglyceride in the liver could be the result of a failure of the development of glyceroneogenesis in adipose tissue or to a loss of gluconeogenesis (the major cataplerotic pathway in the liver) in the absence of PEPCK-C. In this paper we assess the impact of a loss of PEPCK-C in the perinatal period and evaluate the various functions suggested for this enzyme in energy metabolism. Methods Materials Restriction enzymes, Taq DNA polymerase and proteinase K, NAD, NADH, lactate dehydrogenase, and malate dehydrogenase were purchased from Roche Applied Science (Indianapolis, IN). QuickPrep total RNA kit, [5-14C]-sodium glutamate, [U-14C]-glucose, and [2-14C]-sodium acetate was purchased from Amersham Biosciences Corp. (Piscataway, NJ). Triglyceride (GPO) Liquid Reagent Set was purchased from Pointe Scientific, Inc. (Lincoln Park, MI). The NEFA kit was from Wako Chemicals USA, Inc. (Richmond, VA). The embryonic stem cells used in this study were a generous gift from Janet Rossant, University of Toronto. Generation and maintenance of PEPCK-C-/- mice The gene for PEPCK-C was deleted in the mice as described in figure 1. A targeting vector was constructed in which the region of the PEPCK-C gene extending from a Pst I site at ~-300 bp in the PEPCK-C gene promoter to an Xho I site in exon three of the structural gene, was substituted with the PGK_βgal-neoR gene (figure 1A). The neomycin resistance gene (neo) and the diphtheria toxin A chain (DTA) were used as positive and negative selection markers, respectively. The targeting vector was linearized by Not I digestion and electroporated (25 μg of DNA) into E14 embryonic stem cells. Targeted clones that underwent homologous recombination were identified, using Southern blot hybridization of digested genomic DNA from ES cells. The probe used was a PEPCK-C cDNA sequence composed of exons one to six of the mouse PEPCK-C gene (figure 1A). Two out of eight hundred of the G418 resistant clones, that had been screened, showed correct homologous recombination. These clones were subsequently injected into C57BL/6 blastocytes to produce chimeric mice, which were then crossed with C57BL/6 mice to establish germ-line founders. The resulting offspring were genotyped by using Southern blotting or PCR. Figure 1 Generation of PEPCK-C-/- mice. Panel A. Diagram of the targeting vector, which is aligned with endogenous sequence of PEPCK-C gene. The sequence between Pst I and Xho I, which covers part of the gene promoter and the exons 1 and 2 in endogenous gene, was replaced with Neo resistance gene (PGK_βgal-neoR) in the targeting vector. H, Hind III; P, Pst I; X, Xho I. PGK_βgal-neoR, the phosphoglycerate kinase gene promoter drives a fusion gene of βgal and neoR. DTA, the diphtheria toxin A chain. Panel B. Genotyping of the PEPCK-C-/- mice by Southern blotting. Genomic DNA was digested with Hind III and hybridized with a PEPCK-C cDNA probe composed of exons 1~6 of the rat PEPCK-C gene (black bars). For the wild type allele of PEPCK-C, 1.3-kb and 2.75-kb fragments were detected; only one fragment (2.5 kb) was excepted for the targeted allele. Metabolic studies PEPCK-C-/- mice and control littermates at one to three days of age were killed by decapitation. Their livers and blood were collected. Livers slices (5 to 10 mg) were prepared and incubated in a 10-ml Erlenmeyer flask, containing 1 ml of reaction buffer, which was composed of Krebs Ringer bicarbonate buffer at pH 7.4, 1 % defatted bovine serum albumin, and one of following substrates: 1 μCi [U-14C] glucose (5 mM), 1 μCi [2-14C] sodium acetate (2 mM), or 1 μCi [U-14C] sodium glutamate (1 mM). The flask had a thin rubber stopper that served both as a cap to close the flask and to hold a plastic bucket. The incubation was carried out at 37°C for 2 h in an atmosphere of O2/CO2 (95%/5%), in a shaking water bath. At the end of the incubation period, 200 μl of hyamine hydroxide was injected through the rubber stopper into the hanging bucket and 0.5 ml of 10% perchloric acid was injected into the incubation medium, to insure the complete liberation of CO2. After shaking for an addition 30 min, the tissue was removed, rinsed in saline and transferred to 2 ml of chloroform-methanol (2:1 v/v) for the extraction of total lipids. The lipid was extracted and its radioactivity determined. The hanging buckets containing the hyamine hydroxide were dissolved in liquid scintillation fluid and the radioactivity measured using a liquid scintillation spectrometer. Body composition Mice were killed at two days of age and carcass analysis performed as described by Leischner et al. Histology of the Liver Segments of livers from 2-day-old mice were fixed in 10% formalin (Sigma, St. Louis, MO) at 4°C. The tissues were embedded in paraffin and stained with hematoxylin and eosin by the Pathology Core Facility (Case Western Reserve University, Cleveland, OH). The photographs were taken at 400× magnification. DNA Analysis DNA was isolated from the tails of the mice by lysis overnight at 55°C in a buffer containing 50 mM KCl, 10 mM Tris-HCl, pH 8.3, 2.5 mM MgCl2, 0.1% gelatin, 0.45% Nonidet P-40, 0.45% Tween20, and 24 mg/ml of proteinase K. The DNA was digested with Hind III, and the resulting fragments were separated by electrophoresis on 1% agarose gel, transferred to Gene Screen Plus®, and hybridized to a cDNA probe for PEPCK-C. RNA analysis Total RNA was extracted from the livers of one and two-day old mice using a QuickPrep total RNA kit (Amersham Pharmacia Biotech) by a modified acid-phenol/guanidine thiocynate procedure [11] and Northern blotting was performed as described in detail previously [12]. Briefly, 20 μg of total RNA was separated by electrophoresis on an agarose gel, transferred to a Gene Screen Plus membrane and hybridized with a 1.0kb Sma I fragment of the 3'end of the PEPCK-C cDNA. The DNA probe was labeled by using [α-32P]-dCTP. Biochemical analysis The following metabolites were measured in the blood of mice killed up to two days after birth. Blood glucose was determined using an Encore® Glucometer. The concentrations of triglyceride, β-hydroxybutyrate, and urea nitrogen in the blood were determined by Veterinary Diagnostic Services in Marshfield Laboratories (Marshfield, WI), using standard clinical procedures. The concentrations of amino acids in plasma were measured with a high-pressure liquid spectrophotometer equipped with a fluorescent detector, using a o-phthaldehyde derivative and pre-column derivatization [13]. Livers of mice were used to determine the content of glycogen [14] or were freeze-clamped and metabolites extracted as described previously [15]. Spectrophotometric methods were used to determine the concentrations of malate [16], lactate [17], pyruvate [18] and ammonia [19] in the liver extracts or in the blood. The concentration of hepatic triglyceride was determined using the triglyceride reagent set from Pointe Scientific, Inc. (Lincoln Park MI). Briefly, segments of liver were saponified in ethanol-KOH, the sample diluted 1:5 and one ml of triglyceride reagent added to 10 μl of diluted sample and incubated at 37°C for 5 min. The reaction product was analyzed using spectrophotometer (A500) and the concentration determined by linear regression using standards treated with one ml of triglyceride reagent. The activity of PEPCK-C in the liver cytosol of two-day-old PEPCK-C-/- mice and control littermates was determined as described by Ballard and Hanson [20] and the activity expressed as units/g tissue, where one unit equals one μmole of substrate converted to product per min. Statistical analysis All data are reported as means ± the SE. The statistical analysis was performed using SigmaPlot from Systat Software, Inc. (Point Richmond, CA). Results A line of mice was created (PEPCK-C-/- mice) in which the gene for PEPCK-C was deleted by inserting the neo resistance gene, driven by the phosphoglycerate kinase gene promoter, into the PEPCK-C gene, thereby deleting a region of the gene from the Pst I site at -569 of the gene promoter to the XhoI site within the third exon (figure 1). The PEPCK-C-/- mice had no detectable mRNA for PEPCK-C in the liver and kidney (the only tissues tested) (figure 2A) and no activity for the enzyme was noted in their livers (figure 2B). PEPCK-C-/- mice were present in litters at the expected Mendalian ratio at birth, however, all of the PEPCK-C-/- mice died within two to three days after birth. The animals had no distinguishing phenotype at birth, suckled normally and had milk in their stomachs (figure 2C). However, by one day after birth PEPCK-C-/- the mice failed to gain weight (figure 2D) and had about half the total body fat content of control littermates (Table 1). Mice heterozygous for the PEPCK-C gene deletion (PEPCK-C+/- mice) had only half the activity of the enzyme in their livers (figure 2B) and lived normally, with no observable metabolic abnormalities. Table 1 Metabolic alterations in PEPCK-C-/- mice at two days of age. Body composition PEPCK-C+/+ (n) PEPCK-C-/- (n) % Water 80.3 ± 0.30 3 79.26 ± 0.22 3 % Fat 3.48 ± 0.36 3 1.85 ± 0.57 3 % Lean Mass 16.42 ± 0.63 3 18.88 ± 0.78 3 Metabolites in liver Triglyceride (μmoles/g) 4.17 ± 1.00 5 9.58 ± 1.75 5 Lactate (μmoles/g) 0.19 ± 0.02 3 0.37 ± 0.09 3 Malate (μmoles/g) 0.24 ± 0.05 3 2.84 ± 1.58 3 Metabolites in plasma Triglyceride (mg/dl) 23.7 ± 6.3 3 36.3 ± 21.5 2 Ammonia (μmoles/ml) 0.14 ± 0.02 5 0.45 ± 0.28 4 BUN (mg/dl) 29.0 ± 7.0 2 65.5 ± 28.5 2 Lactate (μmoles/ml) 0.28 ± 0.13 3 0.72 ± 0.47 3 β-Hydoxybutyrate (μmoles/ml) 5.83 ± 0.21 3 17.95 ± 0.70 3 The values are expressed as the mean ± S.E. for the number of animals shown in parenthesis Figure 2 Characterization of the PEPCK-C-/- mice. Panel A. A representative Northern blot for mRNA isolated from the liver and kidney of two-day-old mice. Panel B. PEPCK-C activity was determined in the livers of two-day-old mice. The results are expressed as the mean ± S.E. for three animals in each group. Panel C. Photograph of two-day-old PEPCK-C-/- mice. Panel D. Growth retardation of PEPCK-C-/- mice. Body weights of neonates were measure at 6, 30, and 37 h after birth. Three wild type mice and three PEPCK-C-/- mice were used. The effect of a deletion of the gene for PEPCK-C on glucose homeostasis is evident from the changes in the concentration of glucose in the blood of PEPCK-C-/- mice, as compared to control littermates. PEPCK-C activity first appears in the liver at birth [20]. Thus, a deletion of the gene for PEPCK-C should not have a significant effect on glucose homeostasis before birth, since the glucose needed for fetal development is provided by the mother. One day after birth the levels of glucose in the blood of PEPCK-C-/- mice was only 60% of that in controls, and by day three the concentration of glucose in the blood of these mice was 12.3 mg/dl (figure 3A). It is likely that a failure of gluconeogenesis, caused by the ablation of PEPCK-C [10] contributed to the low level of glucose in the blood. The concentration of hepatic glycogen was determined at fetal day 19 (one day before birth) and at day two after birth in both PEPCK-C-/- mice and controls (figure 3B). There was 10% less glycogen in the livers of PEPCK-C-/- mice delivered one day before birth, as compared to control animals. By two days after birth the concentration of hepatic glycogen was markedly decreased in both PEPCK-C-/- mice and controls, but the overall level of glycogen mobilization was much greater in the PEPCK-C-/- mice (figure 3B). This is likely due to the need of he animal to rely the mobilized hepatic glycogen as a source of glucose for fuel metabolism. Interestingly, the administration of glucose by intraperitoneal injection did not rescue the mice from death at day two after birth (data not shown), suggesting that factors other than a disruption of glucose homeostasis were responsible for the death of the mice (i.e. the great accumulation of hepatic triglyceride or a failure of cataplerosis). Figure 3 Alterations in glucose homeostaisis in PEPCK-C-/- and control mice during the perinatal period. Panel A Hypoglycemia in PEPCK-C-/- mice. The concentration of blood glucose was measured in mice at one, two and three days after birth. The results are expressed as the mean ± S.E. for from three to six animals. Panel B. Increased mobilization of hepatic glycogen. The hepatic glycogen level was analyzed at the age of fetal day 19 (day -1) and neonatal day (day +2). The results are expressed as the mean ± S.E. for three animals in each group. Despite having less body fat, PEPCK-C-/- mice had a marked hepatic lipid accumulation, which was visible when the liver was removed from the animal for biochemistry analysis. Histological analysis of the livers of animals at two days of age indicated extensive lipid infiltration of the liver (figure 4). This accumulation of lipid progressed until the mice died between two and three days after birth. The metabolic profile of the PEPCK-C-/- mice at day two after birth is shown in Table 1. The concentration of triglyceride in the livers of these mice was twice that of controls; this was accompanied by a similar, almost 2-fold difference in the level of triglyceride in the blood of the PEPCK-C-/- mice. The concentration of β-hydroxybutyrate in the blood of PEPCK-C-/- mice was three times that of control mice. In addition, the concentration of ammonia in the blood of the PEPCK-C-/- mice was elevated 3-fold over that noted in the blood of control mice, while blood urea nitrogen (BUN) was increased by 2-fold (Table 1), suggesting an increased generation of amino nitrogen from the breakdown of amino acids. Figure 4 Morphology of the livers of PEPCK-C-/- and control mice. Livers of two-day-old pups were analyzed with H&E staining. The arrow indicates fat accumulation in the liver of PEPCK-C-/- mice. Since the deletion of the gene for PEPCK-C should alter the levels of intermediates in related metabolic pathways, we determined the level of malate and lactate in the liver by freeze-clamping the livers of PEPCK-C-/- mice and controls at two days after birth. Intermediates were extracted from the frozen livers and their concentrations measured (Table 1). The concentration of malate was increased by 10-fold in the livers of PEPCK-C-/- mice and the lactate concentration was elevated 2.5-fold. We did not distinguish between the malate in the cytosol and the mitochondria, but it is likely that there is an increase in the concentration of this citric acid cycle intermediate in both compartments of the hepatocyte. This finding of an elevated concentration of malate in the liver suggests that in absence of PEPCK-C the rate of oxalacetate conversion to PEP is limited and this intermediate is converted instead to malate, which accumulates. A similar increase in the concentration of malate in the livers of PEPCK-C-/- mice was reported by She et al. [9]. It is not intuitive that the ablation of PEPCK-C, which is normally classified as a gluconeogenic enzyme, would result in profound fatty livers in the mice. However, PEPCK-C is a major cataplerotic enzyme, so that its absence in the liver would be predicted to alter the rate of citric acid cycle flux, since a major route for the disposal of carbon skeletons from the break down of amino acids would be blocked. This is demonstrated in experiments in which the rate of oxidation of acetate, glucose and glutamate to CO2 and their conversion to total lipid by liver slices from PEPCK-C-/- mice was compared with control animals. The generation of 14CO2 from [2-14C]-acetate was 30% of the rate of oxidation of this compound by liver slices from control mice and the rate of hepatic lipid synthesis by PEPCK-C-/- mice was negligible (Table 2). The same pattern was noted for the metabolism of [U-14C]-glucose. We also determined the rate of oxidation of [5-14C]-glutamate, which enters the citric acid cycle as α-ketoglutarate and must leave as malate or some other four-carbon anion to insure continuing carbon flux through the cycle. Gluconeogenesis is the most important route of disposal of the carbon skeletons that enter the citric acid cycle; in the absence of PEPCK-C the rate of oxidation of [5-14C]-glutamate to 14CO2 was markedly diminished (Table 2). Table 2 The conversion of glucose, acetate and glutamate to CO2 and lipid by liver slices from PEPCK-C-/- and control mice. CO2 Total Lipid Substrates PEPCK-C+/+ (n) PEPCK-C-/- (n) PEPCK-C+/+ (n) PEPCK-C-/- (n) Glucose-U-14C 0.92 ± 0.26 (8) 0.30 ± 0.04 (5) 0.153 ± 0.009 (3) 0.063 ± 0.002 (2) Acetate-2-14C 12.36 ± 0.88 (4) 2.72 ± 0.20 (5) 0.097 ± 0.018 (7) 0.002 ± 0.000 (2) Glutamate-5-14C 5.15 ± 0.65 (2) 1.75 ± 0.65 (2) 1.778 ± 0.091 (4) 0.983 ± 0.098 (3) The values are expressed as μmoles of substrate converted to product/hour/g liver and represent the mean ± SE of the mean for the number of determinations shown in parenthesis. Livers slices from two-day-old mice were incubated for 2 h in Krebs Ringer bicarbonate buffer, pH 7.4, containing substrates at the concentrations outlined in the Methods, and the rates of 14CO2 production and 14C incorporation into total lipid determined. The radioactivity was quantified using a liquid scintillation spectrometer. The synthesis of non-essential amino acids from citric acid cycle intermediates, such as oxalacetate (from aspartate and asparagine) and α-ketoglutarate (from glutamate and glutamine) is potentially capable of relieving the accumulation of intermediates in the cycle that occurs in the livers of PEPCK-C-/- mice [21]. We have measured the level of amino acids in the blood of PEPCK-C-/- mice and controls (figure 5). At two days after birth, there was a marked increase in a number of the non-essential amino acids in the blood of PEPCK-C-/- mice. The levels of alanine, aspartate, asparagine, glutamate, glutamine and glycine in the blood of the PEPCK-C-/- mice were two to eight-fold higher than noted in control animals. This suggests that the carbon skeletons of these amino acids, which would normally enter the citric acid cycle, are prevented from being readily metabolized and the amino acids accumulate in the blood. It is also likely that some fraction of the non-essential amino acids were synthesized de novo from citric acid cycle intermediates, which accumulate in the liver due to the absence of PEPCK-C. This suggestion is supported by the increased concentration of ammonia in the blood of the PEPCK-C-/- mice (see Table 1), which favors the formation of non-essential amino acids in the liver, especially glutamine, which was elevated five times over that noted in the blood of control animals. The concentration in the blood of a number of the essential amino acids, such as lysine, methionine, phenylalanine and tryptophan was not altered by ablation of PEPCK-C, while the levels of leucine, isolucine and valine were increased. The elevation in the level of ornithine, together with the elevated levels of both ammonia and urea in the blood of the PEPCK-C-/- mice suggests elevated urea cycle flux to dispose of amino nitrogen from the breakdown of amino acids. Figure 5 The concentration of amino acids in the blood of PEPCK-C-/- mice and controls. The levels of amino acids were determined in mice at two days after birth. The results are expressed as the mean ± S.E. of three controls and four PEPCK-C-/- mice. Discussion It is clear from this and other studies [9,10] that PEPCK-C is critical for life; its absence in the mouse results in death within the first two to three days after birth. We are aware of only one reported example of PEPCK-C deficiency in humans. In 1976, Vidnes and Sovik [22] reported the absence of PEPCK-C in the liver of an infant, which had persistent hypoglycemia and an early death, despite the presence of PEPCK-M. The rarity of PEPCK-C deficiency in humans further supports the physiological significance of this enzyme. The developmental profile of hepatic PEPCK-C is of interest in this regard. PEPCK-C is absent in the liver of all mammals during fetal development and appears dramatically at birth [20]. In contrast, the enzyme is present in the kidney [23] and may appear in other tissues before birth, but the exact pattern of development in tissues other than the kidney and liver has not been determined. As an example, the time course of development of PEPCK-C in white adipose tissue is currently unknown, despite the important role that PEPCK-C plays in this tissue [3,7]. Considerable PEPCK-M activity is present in mammalian liver before birth [24] but it is the appearance of PEPCK-C at birth and the marked alteration in the hepatic redox state, which occurs during this period [25], that are generally considered to be the important events in the initiation of hepatic gluconeogenesis. Since the role of PEPCK-C in gluconeogenesis is well established in all species [26], it is reasonable to assume that one of the causes of the death of the PEPCK-C-/- mice is the profound hypoglycemia which occurs at two to three days after birth. However, providing glucose to the mice by intraperitoneal injection did not rescue the animals. Mice with a liver-specific deletion of the gene for PEPCK-C survive the perinatal period, maintain normal glucose homeostasis during fasting and can be made diabetic [10]. Whole body glucose turnover in these mice is only slightly decreased, when compared to control animals. This suggests that the kidney can assume the function of the liver and make enough glucose for the needs of the animal. In addition, the liver can synthesize glucose from glycerol, a process that by-passes the reactions in the pathway of gluconeogenesis that require PEPCK-C. Since only 5–10% of the total PEPCK is mitochondrial in rodent species [27], it is unlikely that PEPCK-M plays a significant role in hepatic gluconeogenesis in mice lacking hepatic PEPCK-C. Mice lacking hepatic PEPCK-C have lower levels of liver glycogen two days after birth. As noted in Table 1, newborn PEPCK-C-/- mice have considerably less glycogen than control littermates, presumably due to the need to mobilized glycogen to support the blood glucose levels in the absence of hepatic gluconeogenesis. However, She et al [9] reported that mice lacking PEPCK-C in the liver had a greatly reduced rate of glycogen synthesis in both the liver and the muscle. The marked accumulation of fat in the livers of the PEPCK-C-/- mice strongly suggests that there is another critical metabolic role for the enzyme (other than gluconeogenesis) which, when missing, can markedly alter the metabolic capacity of the animal. The concentration of hepatic triglyceride in mice two days after birth is twice that of control animals. The biochemical mechanisms responsible for the accumulation of triglyceride in PEPCK-C-/- mice are complex. In the present paper, we present evidence that in the absence of PEPCK-C hepatic cataplerosis is greatly reduced, resulting in a reduction of the ability of the liver to appropriately oxidize acetyl CoA to CO2 in the citric acid cycle. This causes an increase in the synthesis of ketone bodies and a build up of fatty acids in the liver, with the subsequent synthesis of triglyceride. Cataplerosis describes reactions involved in the disposal of citric acid cycle intermediates generated by the entry of compounds into the cycle during the breakdown of amino acids and other metabolites (i.e. propionyl CoA), or by the carboxylation of pyruvate to oxalacetate [28]. The biological necessity for cataplerosis resides in citric acid cycle dynamics. The cycle oxidizes acetyl CoA to CO2 but cannot fully oxidize four or five carbon intermediates. This requires a series of reactions that are capable of efficiently removing citric acid cycle anions before they accumulate in the mitochondria. It is now recognized that PEPCK-C functions in cataplerosis in many tissue, not just in the more intensively studied tissues, such as the liver and kidney. For example, the net generation of alanine from glutamine, as occurs in the small intestine, involves PEPCK-C and cataplerosis. Glutamine carbon enters the citric acid cycle as α-ketoglutarate, is converted in the cycle to malate, which leaves the mitochondria, is oxidized to oxalacetate and then decarboxylated to PEP by PEPCK-C. Pyruvate kinase then converts the PEP to pyruvate, the pyruvate is transaminated by alanine aminotransferase to alanine, which is then released into the blood and transported to the liver. It has been estimated that as much as 40% of the alanine used for gluconeogenesis by the human liver during starvation is derived from the glutamine in the small intestine [29]. Alternatively, the pyruvate can be converted to acetyl CoA in the mitochondria for the generation of energy via the citric acid cycle. A similar role for PEPCK-C has been proposed for the conversion of other amino acids to alanine in the muscle [30]. Deleting the gene for PEPCK-C in tissues of the mouse, where it is normally present, will have the effect of altering cataplerosis and profoundly interfering with energy metabolism. The results of the present study suggest that in the absence of PEPCK-C in the liver the rate of removal of intermediates from the citric acid cycle is greatly reduced. This is illustrated in figure 6. The high concentration of malate (10 times that noted in controls) in the livers of PEPCK-C-/- mice is most likely due to the lack of removal of oxalacetate in the absence of PEPCK-C. As malate accumulates, the rate of citric acid cycle flux is markedly reduced, resulting in a decrease in both the oxidation of acetyl CoA to CO2 and its conversion to fatty acid in the livers of PEPCK-C-/- mice, as well as a reduction in the rate of oxidation of glucose and glutamate to CO2 (see Table 2). The acetyl CoA not utilized in the citric acid cycle is converted to ketone bodies, which increase in the blood of the PEPCK-C-/- mice (Table 1). The decreased oxidation of acetyl CoA in the citric acid cycle results in the shunting of fatty acids into the synthesis of triglyceride, with the subsequent accumulation of fat in the liver. As would be predicted, the absence of PEPCK-C results in an increase in the level of a number of non-essential amino acids, including aspartate, asparagine, glutamate, glycine, alanine and glutamine that are synthesized from citric acid cycle anions. When hepatic PEPCK-C is ablated, the synthesis and release of these amino acids from the liver serves as a major cataplerotic process. An alteration in citric acid cycle flux would thus contribute to the observed accumulation of these amino acids in the blood. Clearly, the failure of the liver to carry out cataplerosis has far reaching effects on energy metabolism in the animal. It should be noted, however, that genetic disorders in humans in which the activity of pyruvate carboxylase is reduced results in the development of a fatty liver, suggesting that a failure of anaplerosis can also result in hepatic fat accumulation. Figure 6 The role of PEPCK-C in cataplerosis in the liver. The reactions of the citric acid cycle are presented, with the major end-products shown in gold boxes; the large red box represents the mitochondrial membrane. The concentrations of amino acids, β-hydroxybutyrate, glucose and triglyceride were determined from blood samples, while the levels of malate and pyruvate were determined from freeze-clamped liver. The ablation of PEPCK-C (shown by a red bar) results in a 10-fold increase in the concentration of malate in the liver (we did not distinguish between malate in the cytosol and the mitochondria) and a build-up of cycle intermediates [23]. This leads to a decrease in the rate of citric acid cycle flux and the resultant accumulation of acetyl CoA, which is subsequently converted to ketone bodies and released by the liver. The rate of fatty acid oxidation in the PEPCK-C-/- mice is also markedly decreased, resulting in an increase in triglyceride synthesis from these fatty acids that leads to the development of a fatty liver. There is also a marked increase in the concentration of amino acids in the blood that were generated from citric acid cycle intermediates. The increased rate of flux of intermediates leaving the citric acid cycle is denoted by heavy arrows. Mice with a liver-specific ablation of PEPCK-C have been used by Burgress et al [21] to determine the importance of this enzyme in citric acid cycle flux in the intact animal. They noted that after a 24 h fast virtually all of the newly synthesized glucose was from glycerol and the formation of glucose from citric acid cycle intermediates was negligible. Flux through the citric acid cycle was 10 to 40-fold lower than noted in the livers of control mice, which correlated with an accumulation of citric acid cycle intermediates. They concluded that in the absence of PEPCK-C there was a failure of cataplerosis, leading to a decreased citric acid cycle flux, decreased fatty acid oxidation and an accumulation of liver lipid. This finding is in accord with that described in this paper using mice in which the gene for PEPCK-C is deleted in all tissues and underscores the key role of PEPCK-C in the normal functioning of the citric acid cycle in mammals. In addition to altering cataplerosis, the deletion of the gene for PEPCK-C will result in a decrease in glyceroneogenesis in tissues of the mouse. At birth the mouse ingests milk that is relatively high in fat (13%) and low in carbohydrate (3%) when compared with some other mammalian species (i.e. human). This dietary fat, while a good source of energy, is deposited in white adipose tissue during the early perinatal period. The relatively low level of carbohydrate in the diet suggests that glyceroneogenesis in adipose tissue could play an important role in controlling the rate of triglyceride deposition in this tissue during suckling. Studies in our laboratory, using in vivo tracer methodology, have shown that glyceroneogenesis in adipose tissue of the rat is the predominant source of glyceride-glycerol, even when the animals were fed a diet high in sucrose (Nye, Hanson and Kalhan, unpublished data). If glyceroneogenesis is playing an important role in the deposition of triglyceride in the adipose tissue of the newborn mouse, the absence of PEPCK-C in the tissue could explain the very high level of triglyceride accumulation noted in the livers of the PEPCK-C-/- mice, since fatty acids not re-esterifed to triglyceride in adipose tissue, could end up being converted to triglyceride in the liver. In support of this suggestion, the tissue-specific ablation of PEPCK-C gene expression in adipose tissue of mice resulted in the development of a fatty liver. Conclusion The studies presented here and the recent work on mice with a liver-specific deletion in the gene for PEPCK-C [9,10,21], emphasize the important role played by this enzyme in a number of metabolic process other than gluconeogenesis. It is likely that PEPCK-C is critical for cataplerosis, especially in those tissues in which significant levels of biosynthesis occur. The dynamic nature of the citric acid cycle and its role in both energy generation and biosynthetic process requires the coordinated activity of PEPCK-C to insure a balance between carbon input into the cycle and carbon outflow. The fact that this critical function of PEPCK-C has not been appreciated for so long, underlines the importance of genetically modified animal models as tools for better understanding the control of metabolism. As an example, the metabolic function of PEPCK-M has been virtually ignored over the years despite the fact that it represents 50% of the activity of PEPCK in humans. This neglect is ironic, since Utter and Kurahashi discovered PEPCK-M in the livers of chickens [31] and used it to delineate the pathway of gluconeogenesis. Our understanding of the biology of PEPCK will not be complete until we directly determine the metabolic role of PEPCK-M in mammalian species. In this regard, the major limiting factor is the fact that the animal species most easily genetically modified (the mouse) has only a marginal activity of PEPCK-M in any tissue. It should be possible, however, to introduce a transgene containing the gene PEPCK-M into tissues of PEPCK-C-/- mice. This would allow us to assess the metabolic function of PEPCK-M in the absence of PEPCK-C; these studies are in progress. Competing interests The author(s) declare that they have no competing interests. Authors' contributions PH, made genetically modified mice and carried out molecular studies; MJ, DL and DAC provided guidance in aspects of the preparation of genetically modified of the mice; JY, assisted in the preparation and analysis of the data for the manuscript; SCK, provided amino acid analysis and data interpretation and helped in the preparation of the manuscript; SMT, participated in the development of the genetically modified mice; LR, participated in the experimental design and helped in the preparation of the manuscript; RWH, conceived of the study and worked on the experimental design, guided aspects of the work and wrote the manuscript. Acknowledgements This research was supported by DK 58620 and DK 25541 (RWH) by HD11089 (SCK) and by HD37934 to RAC, from the National Institutes of Health and by a United States-Israel Binational Grant 9100268 (L R). ==== Refs Chakravarty K Cassuto H Reshef L Hanson RW Factors that control the tissue-specific transcription of the gene for phosphoenolpyruvate carboxykinase-C Crit Rev Biochem Mol Biol 2005 40 129 154 15917397 10.1080/10409230590935479 Hanson RW Patel YM Meister A P-enolpyruvate carboxykinase: the gene and the enzyme Advances in Enzymology 1994 69 New York, John Wiley and Sons 203 281 Reshef L Hanson RW Ballard FJ A possible physiological role for glyceroneogenesis in rat adipose tissue J Biol Chem 1970 245 5979 5984 5484457 Botion LM Brito MN Brito NA Brito SR Kettelhut IC Migliorini RH Glucose contribution to in vivo synthesis of glyceride-glycerol and fatty acids in rats adapted to a high-protein, carbohydrate-free diet Metabolism 1998 47 1217 1221 9781624 10.1016/S0026-0495(98)90326-2 Kalhan SC Mahajan S Burkett E Reshef L Hanson RW Glyceroneogenesis and the source of glycerol for hepatic triacylglycerol synthesis in humans J Biol Chem 2001 276 12928 12931 11278297 10.1074/jbc.M006186200 Reshef L Olswang Y Cassuto H Blum B Croniger CM Kalhan SC Tilghman SM Hanson RW Glyceroneogenesis and the triglyceride/fatty acid cycle J Biol Chem 2003 278 30413 30416 12788931 10.1074/jbc.R300017200 Hanson RW Reshef L Glyceroneogenesis revisited Biochimie 2003 85 1199 1205 14739071 10.1016/j.biochi.2003.10.022 Hahn P Novak M Development of brown and white adipose tissue J Lipid Res 1975 16 79 91 1168684 She P Shiota M Shelton KD Chalkley R Postic C Magnuson MA Phosphoenolpyruvate carboxykinase is necessary for the integration of hepatic energy metabolism Mol Cell Biol 2000 20 6508 6517 10938127 10.1128/MCB.20.17.6508-6517.2000 She P Burgess SC Shiota M Flakoll P Donahue EP Malloy CR Sherry AD Magnuson MA Mechanisms by which liver-specific PEPCK knockout mice preserve euglycemia during starvation Diabetes 2003 52 1649 1654 12829628 Chirgwin JM Prezybyla AE MacDonald RJ Rutter WJ Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease Biochemistry 1979 18 5294 6000 518835 10.1021/bi00591a005 McGrane MM deVente J Yun J Bloom J Park EA Wynshaw-Boris A Wagner T Rottman FM Hanson RW Tissue-specific expression and dietary regulation of a chimeric PEPCK/bGH gene in transgenic mice J Biol Chem 1988 263 11443 11451 2841327 Turnell DC Cooper JD Rapid assay for amino acids in serum or urine by pre-column derivatization and reversed-phase liquid chromatography Clin Chem 1982 28 527 531 7067099 Lo S Russell JC Taylor AW Analysis of glycogen in small samples J Appl Physiol 1970 28 234 236 5413312 Garber AJ Hanson RW The interelationships of varous pathways forming gluconeogenic precursors in guinea pig liver mitchondria. J Biol Chem 1971 246 589 595 5542673 Gutmann I Wahlefeld AW Bergmeyer HU L-Malate: Determination with malate dehydrogenase and NAD Methods of Enzymatic Analysis 1974 3 New York City, Academic press 1585 1589 Gawehn K Bergmeyer HU Bergmeyer HU D-(-)-Lactate Methods of Enzymatic Analysis 1974 3 New York City, Academic Press 1492 1495 Czok R Lampecht W Bergmeyer HU Pyruvate, phosphoenolpyruvate and D-glycrate-2-phosphate Methods of Enzymatic Analysis 1974 3 New York City, Academic Press 1446 1451 Kun E Kearney EB Bergmeyer HU Ammonia Methods of Enzymatic Analysis 1974 4 Second Edition New York, Academic Press 1802 1806 Ballard FJ Hanson RW P-enolpyruvate carboxykinase and pyruvate carboxylase in the developing liver Biochem J 1967 104 866 871 6049928 Burgess SC Hausler N Merritt M Jeffrey FM Storey C Milde A Koshy S Lindner J Magnuson MA Malloy CR Sherry AD Impaired tricarboxylic acid cycle activity in mouse livers lacking cytosolic phosphoenolpyruvate carboxykinase J Biol Chem 2004 279 48941 48949 15347677 10.1074/jbc.M407120200 Vidnes J Sovik O Gluconeogenesis in infancy and childhood III. Deficiency of the extramitochondrial form of hepatic phosphoenolpyruvate carboxykinase in a case of persistent neonatal hypoglycemia Acta Pediatr Scand 1976 65 307 312 Zarzoli A Turkenkopf IJ Mueller VL Gluconeogenesis in developing rat kidney cortex Biochem J 1969 111 181 185 4303362 Arinze IJ On the development of P-enolpyruvate carboxykinase and gluconeogenesis in guinea pig liver. Biochem Biophys Res Comm 1975 65 184 189 1147983 10.1016/S0006-291X(75)80077-5 Philippidis H Ballard FJ The development of gluconeogenesis in rat liver: in vivo experiments Biochem J 1969 113 651 657 5386185 Hanson RW Mehlman MA Gluconeogenesis: Its role in mammalian species. 1976 New York, John Wiley and Sons Hanson RW Garber AJ P-enolpyruvate carboxykinase: I. Its role in gluconeogenesis Amer J Clin Nutr 1972 25 1010 1021 4342753 Owen OE Kalhan SC Hanson RW The key role of anaplerosis and cataplerosis for citric acid cycle function J Biol Chem 2002 277 30409 30412 12087111 10.1074/jbc.R200006200 Felig P Owen OE Wahren J Cahill GFJ Amino acid metabolism during prolonged starvation J Clin Invest 1969 48 584 594 5773094 Snell K Duff DA The release of alanine by rat diaphragm muscle in vitro. Biochem J 1977 162 399 403 849291 Utter MF Kurahashi K Mechanism of action of oxalacetate carboxylase from liver. J Amer Chem Soc 1953 75 785 787 10.1021/ja01099a522	16300682	PMC1325233	CC BY	2021-01-04 16:24:11	no	Nutr Metab (Lond). 2005 Nov 21; 2:33	utf-8	Nutr Metab (Lond)	2,005	10.1186/1743-7075-2-33	oa_comm
==== Front BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1501626643210.1186/1471-2164-6-150 Methodology ArticleProposed methods for testing and selecting the ERCC external RNA controls External RNA Controls Consortium [email protected] Members of the External RNA Controls Consortium are listed in Appendix 12 Send correspondance to Laura H Reid2005 2 11 2005 6 150 150 9 8 2005 2 11 2005 Copyright © 2005 Reid and External RNA Controls Consortium; licensee BioMed Central Ltd.2005Reid and External RNA Controls Consortium; 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 External RNA Control Consortium (ERCC) is an ad-hoc group with approximately 70 members from private, public, and academic organizations. The group is developing a set of external RNA control transcripts that can be used to assess technical performance in gene expression assays. The ERCC is now initiating the Testing Phase of the project, during which candidate external RNA controls will be evaluated in both microarray and QRT-PCR gene expression platforms. This document describes the proposed experiments and informatics process that will be followed to test and qualify individual controls. The ERCC is distributing this description of the proposed testing process in an effort to gain consensus and to encourage feedback from the scientific community. On October 4–5, 2005, the ERCC met to further review the document, clarify ambiguities, and plan next steps. A summary of this meeting and changes to the test plan are provided as an appendix to this manuscript. ==== Body Background External RNA controls consortium The External RNA Control Consortium (ERCC) is an ad-hoc group with approximately 70 members from private, public, and academic organizations. The group was initiated in 2003 to develop a set of external RNA control transcripts that can be used to assess technical performance in gene expression assays. The external RNA controls will be added after RNA isolation, but prior to cDNA synthesis. They are being designed to evaluate whether the results for a given experiment are consistent with defined performance criteria. All ERCC work is intended to apply to quantitative, real-time reverse transcriptase polymerase chain reaction (QRT-PCR) assays as well as one-color and two-color microarray experiments. The ERCC has worked together to define the desired properties of the transcripts, general protocols for their application, and an analysis scheme for performance assessment. In December 2003, the group developed a specification document that was discussed and refined in a public workshop at the National Institute of Standards and Technology (NIST) [1]. Protocols for the use of external RNA controls in clinical applications are included in the Molecular Methods 16-P document from the Clinical and Laboratory Standards Institute, and were developed in a formal, accredited, open, consensus forum including several ERCC members. The analysis approach was developed in a public workshop at NIST in June 2004, and is based upon the measurement of pooled transcripts at known concentrations. In the past year, the ERCC has refined specifications, generated and collected control sequences, evaluated optimal polyadenylated (polyA) tail length and identified a path forward for access and distribution of the controls. We are now initiating the Testing Phase of the project as described in this document. Purpose of this document During ERCC Testing, candidate external RNA controls will be evaluated in both microarray and QRT-PCR gene expression platforms. This document describes the proposed experiments and informatics process that will be followed to authoritatively test and qualify individual controls. Qualification of a control sequence is consistent with the ISO 9000 definition of validation: "Confirmation, through the provision of objective evidence, that requirements for a specific intended use or application have been fulfilled." Based on the results of the test and qualification experiments, the ERCC will select a set of external RNA controls that perform consistently across technologies and platforms. DNA clones of the controls, basic informatics tools and appropriate documentation will be available to the public. Commercial products (e.g. primer sets and pools of transcripts) may also be developed and made available as Certified Reference Materials. The ERCC is committed to open access and inclusive practices. We are distributing this description of the proposed testing process in an effort to gain consensus from the scientific community and to confirm the value of the final products. We hope it will be carefully reviewed by clinical and research laboratory scientists. On October 4–5, 2005, NIST hosted an ERCC Testing Workshop as an open forum to encourage feedback from the community and to invite volunteers to participate in the testing phase tasks. Comments from the meeting are posted on the NIST site [2] and included in Appendix 2 of this document. Interested parties should contact Dr. Janet Warrington [3] for further information. Specific aims The five specific aims of the testing project will result in production of a well-characterized, tested set of controls with demonstrated acceptable performance on major microarray platforms and commonly used QRT-PCR methods. The specific aims are as follows. 1. Design and produce the reagents necessary for the testing plan, including candidate external RNA controls, prototype arrays, primer sets for QRT-PCR and informatics tools for managing and annotating the testing results. 2. Qualify the reagents. Identify and minimize cross reactivity and potential interactions between the external controls, the probes used for their detection and background RNA molecules. 3. Qualify the assay by collecting performance data in multiple gene expression assays. Identify the linear range, sensitivity and reproducibility of individual candidate controls. Define performance criteria and select a candidate reference set of external RNA controls for future testing. 4. Qualify the product by using the candidate reference set of external RNA controls in typical microarray and QRT-PCR assays. Extend analyses to multiple RNA backgrounds and testing sites. Confirm the performance quality of the set of transcripts and the informatics tools used for their analysis. Finalize the reference set and ERCC products. 5. Distribute DNA clones of the external RNA controls and informatics tools to general scientific community. Publish report on ERCC materials, testing data and analysis methods. Write application use and general protocols. Definitions, assumptions and limitations In this document, the term "external RNA controls" refers to unlabeled, polyadenylated sense transcripts that are added to an RNA sample prior to processing and used to measure technical performance of the assay system. Although labeled transcripts will be generated during the testing effort to characterize the hybridization dynamics of each control, the final ERCC product is not intended to be used as a "hybridization control" in microarray experiments. Use of external controls The external RNA controls (sometimes referred to as "spikes") are intended to be added to total RNA samples before initiating the gene expression experiment. For microarray assays, a pool of multiple external RNA controls can be added to a single sample, then labeled and hybridized in parallel with the target. If each of the added spikes is introduced at a different concentration, a calibration curve can be constructed to indicate the linear range of the hybridization. Different pools of external RNA controls can be used in separate hybridizations (one-color system) or in different channels of the same hybridization (two-color system). For QRT-PCR assays, a single external RNA control can be added to the RNA sample and amplified simultaneously with the target sequence (two-color system) or independently in a replicate well (one-color system). Current external RNA controls in microarray assays Several sets of external RNA controls are currently available for microarray research [4-9], but few probes corresponding to the controls are available on commercial microarrays and there is no consistency across platforms. As a result of the ERCC effort, a standard set of approximately 96 well-characterized, external RNA controls will be available. Many different array manufacturers are participating in the ERCC process and have committed to including probes complementary to the ERCC external RNA controls on their future microarray products. Current external RNA controls in QRT-PCR assays There are two RNA transcript quantification methods for QRT-PCR assays, absolute and relative quantification. Absolute quantification is very useful and accurate when QRT-PCR is used for measuring one or a few targets, but it often requires generation of template standards for each target and it may become cumbersome especially when many targets are analyzed [10,11]. When QRT-PCR is used for gene expression profiling, or measuring dozens of transcripts simultaneously, relative quantification is often used. In this case, users rely on internal genes for normalization, usually a "housekeeping" gene or other transcript whose expression level is thought to be invariant. This approach has its own limitations. For example, an accurate relative comparison of gene expression requires similar primer affinities for both the target and the internal control, which are often difficult to achieve. Also, recent reports [12-17] have shown that the expression levels of several housekeeping genes change in response to drug treatments or environmental changes, such as stress. These fluctuations raise questions about the general utility of a single universal set of genes as internal controls. Finally, the results of QRT-PCR reactions containing QRT-PCR inhibitors in RNA preparations are cumbersome to correct by using an internal gene transcript. External RNA controls would provide an alternative method for normalization in relative quantification assays. The combination of external and internal or invariant controls is recommended as the most robust approach to optimal experimental control. Dose dependence, tissue specificity and degradation issues impact consistency of information and can be controlled for using a combination of controls. In addition, internal controls are more appropriately custom designed for analyte control in assays with a clearly defined intended use and therefore are by definition assay-dependent. The external RNA controls will provide a complementary resource to internal and assay-specific custom controls. Clone collection The ERCC product will be a set of clones that have been well characterized for performance on multiple microarray and QRT-PCR platforms. Candidate clones are submitted to the ERCC and evaluated as described in this document. Based on the testing results, the ERCC will select clones that perform acceptably on all participating platforms to be used as the reference set. At this time, a number of organizations have submitted 140 candidate transcripts to be tested. As shown in Table 1, the candidate external RNA clones are either synthesized from unique sequences (i.e., artificial) or are derived from genes in several non-mammalian species. Inserts are 500 – 2000 bp with a 20–30 bp polyadenylated tail. These clones are given freely without intellectual property rights. The ERCC welcomes additional submissions. Interested parties should contact Dr. Janet Warrington for further information [3]. Table 1 Summary of external RNA control clone library Number Affiliation of Contributor Genus species Length of RNA 1 – 28 Affymetrix B. subtilis 700–2,000 29 – 40 Affymetrix Artificial Sequences 500–1,900 41 – 43 USDA-ARS-NCAUR Bos taurus 500 44 – 46 USDA-ARS-NCAUR Glycine max 500 47 Ambion Lamda phage 1,000 48 – 53 Ambion Artificial Sequences 750–1,000 54 – 61 Ambion E. coli 750–2,000 62 – 82 Stanford University Methanococcus 500–750 83 – 85 Agilent Technologies Artificial Sequences 500 86 – 90 GE Healthcare E. coli 1,000 91 – 140 Affymetrix/Ambion/Atactic Artificial Sequences 1,000 ERCC products This work culminates in the selection and distribution of materials that support the use of external RNA controls in expression assays. At the completion of the testing phase, the ERCC will release three products for use by the scientific community: • DNA clones and sequences of the reference set of external RNA controls • Basic informatics tools • Publication of data, test results, protocols Several additional reagents and documents will be developed during the testing phase. Many of the documents will be issued electronically to the general scientific community via NIST [2] as supplemental information validating the control production process. Laboratory reagents will be used exclusively by ERCC members during testing. Some of the reagents will be donated by participating ERCC institutions, including array manufacturers, reagent manufacturers and NIST. Methods Testing strategy The testing work is divided into five sequential phases to coincide with the specific aims listed above. Milestones will be used to indicate the completion of each phase (Table 2). Table 2 Summary of testing phases Testing Phase Specific Aim Milestone 1 – Design & Development Generate Reagents Distribution for prototype testing 2 – Prototype Testing Validate Reagents Initial data collected, acceptance criteria established 3 – Proof of Concept Validate Assay Candidate set of ERCC clones 4 – Functional Testing Validate Product Final set of ERCC clones 5 – Performance Review Distribute Product Symposium The proposed methods and specific tasks associated with each phase of testing are described in the following sections. The testing will be performed in parallel by three working groups: microarray group, QRT-PCR group and informatics group. A fourth group responsible for RNA production will be contracted by NIST during phase 1. All of the ERCC members will participate in the documentation and publication tasks. Tasks for transcript and other reagent production subgroup This section describes the production and quality control processes for the candidate external RNA clones used in testing. NIST will be soliciting these manufacturing tasks through standard governmental procurement process. Phase 1 – design and development • Collect clones DNA clones for candidate external RNA controls will be collected and stored in a NIST laboratory. Preliminary sequence information will also be collected from each clone contributor. Each clone will be assigned a unique identification number that coincides with all bioinformatic nomenclature specification requirements. A list of all candidate external RNA controls with vector, insert characteristics for the DNA plasmids and E. coli clones (strains carrying plasmids) will be compiled. This file will also contain the potential plasmids, which will be used for the production of the external RNA controls. A portion of the DNA plasmid or E. coli stock will be sent to the selected manufacturer(s) for the production of the ERCC reagents to be used in the testing project. • Verify DNA sequence Each plasmid will be grown in a small scale (~10 ml culture), purified and used for DNA sequence verification. Primer extension reactions will be designed to obtain full-length insert sequences. Both strands will be sequenced. These sequence files will be compared to the sequence submitted by the clone contributor. It will be important to verify both the presence of a polyA tail and the 3' restriction site to be used for linearization during in vitro transcription (IVT) template preparation. Ambiguities will be noted and will be further investigated by re-sequencing and database comparisons. The ERCC Informatics subgroup will determine the "correct" sequences to be used for microarray probe and QRT-PCR primer design. Once all of the sequences have been collated and verified, these plasmids will be used for the next phase of production. • Produce bacterial stocks and plasmids All of the clones will be grown in 1 liter cultures to generate 1–2 mg of plasmid DNA. Samples of the bacterial cultures will be saved to create two sets of triplicate glycerol stocks properly labeled and stored at -80°C. One set will be sent to NIST and one set will remain at the manufacturer's site until the completion of the project, at which time the second set will be also be returned to NIST. Each purified plasmid will undergo a set of quality control tests to measure DNA concentration, confirm purity and verify the ability to be linearized at a restriction site located 3' of the polyA tail region. This last confirmation ensures the usability of the clones in the IVT reactions. The sequence will be re-verified using a single pass primer extension from the RNA polymerase promoter sequence (e.g., T7, T3). • Digest plasmids Each plasmid will be linearized by restriction digestion using the appropriate amount of DNA needed to produce the specified amount of transcript. The plasmids will be inspected by high quality agarose gel electrophoresis to ensure complete digestion. The linear plasmids will be purified and concentrations will be normalized to 1 mg/ml. • Test IVT template Approximately 1 ug of each linear plasmid will be used in a 20 ul IVT reaction. The resulting RNA transcript will be diluted and inspected on an Agilent 2100 Bioanalyzer. This analysis will determine whether the size of the RNA transcript matches the expected length for each plasmid and whether any of the IVT reactions produced aborted products. At this stage, the observed transcript RNA yields will be used to calculate the efficiency of IVT reactions for each plasmid. Transcription problems will be addressed at this stage and will help dictate reaction volumes needed for the large scale IVT reactions in the next phase of production. • Produce RNA transcripts The IVT reactions will be scaled-up to produce the desired specified amount of RNA transcript, with considerations of RNA recovery and reasonable overage. Immediately after IVT, a small amount of RNA will be analyzed on the Agilent 2100 Bioanalyzer. If the transcript RNA is the expected size, it will enter into a large scale RNA purification method. The final purified RNA will be normalized to a concentration of ~1 mg/ml. These will be known as the "Production RNA Transcript Stocks". • Verify quality of transcript stocks The quality of the Production RNA Transcript Stocks will be assessed for several characteristics, including concentration, purity, integrity, stability and size. Concentration will be determined using rigorously developed standard operating procedures for UV absorbance. Purity has two aspects: the integrity of the transcript (i.e., percentage of full length transcript) and the stability of the transcript (i.e., presence of low levels of nucleases). The quality of the stock transcripts will also be confirmed by measuring the length of the RNA to determine if it matches the expected size. The Agilent 2100 Bioanalyzer will be used for analysis of integrity, stability and nucleotide length. • Prepare individual stocks of RNA transcripts The individual RNA transcripts will be diluted in RNA Storage Buffer (citrate buffer pH 6.3) and normalized to 100 ng/ul at 1000 nt in length. For example, RNA transcripts that are 750, 1200, and 2000 nt in length will be diluted to 75, 120 and 200 ng/ul respectively. To determine the molar concentration of 1000 nt transcript, a formula or software script will be used to calculate exact molecular weight (MW) given RNA sequence. The RNA (single-stranded) molecular weight will be calculated for the phosphorylated, protonated form of the molecule using the following formula: MW = (#A × 329.21) + (#C × 305.18) + (#G × 345.21) + (#U × 306.17) + 18.02 Stocks of individual RNA transcripts will be available in two forms: 96- well plates or screw-top tubes. Each RNA will be diluted and 50 ul aliquots of each normalized concentration will be transferred to a pre-specified well of a 96-well plate. Wells will be spot-checked for volume. Plates will be sealed with tape. Barcode tracking will be utilized to identify lots and all other tracked information. Plates will be labeled, dated and stored at -80°C until final packaging and shipping. The RNA transcripts will also be distributed in individual tubes at an equal molar concentration (to be determined). • Prepare pools of RNA transcripts A series of RNA mixtures will be made according to the experimental plans. A description of all the pools required for microarray testing is presented in Table 3. The QRT-PCR testing generally relies on a series of dilutions of equal molar pools of the RNA transcripts (pools 0 and 11 in Table 3). Table 3 Description of pools and experiments in microarray testing Pool External RNA Clones Background RNA No. of Arrays 0 1 to 144 none 0 1 1 to 48 pre-labeled none 3 2 49 to 96 pre-labeled none 3 3 97 to 144 pre-labeled none 3 4 1 to 144pre-labeled none 0 5 1 to 72 (high conc.) 73 to 144 (low conc.) human 3 6 1 to 72 (low conc.) 73 to 144 (high conc.) human 3 7* 1 to 96 (diff. conc.) human 12 8* 1 to 96 human 12 9* 1 to 96 human 12 10* 1 to 96 human 12 11 1 to 96 human 0 12 1 to 96 human 3 13 1 to 96 human 3 14 1 to 96 human 3 Total 75 Pools 7–10 may be further diluted to expand the concentration range tested. Array count is per one-color platform and does not include background RNA negative control samples. Pools 0 and 11 will also be used in QRT-PCR assays. • Manufacture pre-labeled cRNA The RNA transcripts will be reverse transcribed with an oligo(dT)-T7 promoter primer and second strand synthesis will be performed to create cDNA templates for the synthesis of pre-labeled cRNA for each external RNA control. To allow testing in multiple microarray platforms, three types of cRNA labels will be used: biotin, DIG and amino-allyl. This labeled cRNA will be purified and quantified as above, normalized to equal molar concentrations, and pooled as shown in Table 3 to construct accurate standard signal intensity response curves for each oligonucleotide probe feature corresponding to the external RNA controls. Tasks for microarray subgroup This section describes the testing methods necessary to develop a set of external RNA controls that can be used to assess the technical performance of microarray experiments. It is anticipated that many of the microarray vendors will participate in the ERCC testing, including Applied Biosystems, Affymetrix, Agilent, GE Healthcare and Illumina. Spotted, non-commercial microarrays will also be contributed by the USDA and NCI. This collection of microarray products include both one-color platforms, which hybridize one labeled target to a single microarray, and two-color platforms, which hybridize two targets with different labels to a single microarray. The microarray testing phases are designed to evaluate the candidate external RNA controls on both one-color and two-color platforms using the same pools of transcripts. Many of the experiments are designed to initially generate sufficient, quality data using relatively few microarrays and pools of transcripts. They may be expanded to a larger range of transcript concentrations and/or a greater number of replicate samples, as desired. A description of all required external RNA control pools is presented in Table 3. This plan assumes that phase 1 begins with 144 candidate external RNA controls and that a putative set of approximately 96 clones has been selected for further characterization by phase 3. Phase 1 – design and development • Design probes and generate commercial microarrays After sequence verification, each array manufacturer will design probes targeting all of the candidate external RNA controls and generate microarrays to be used during testing. • Design probes and generate non-commercial microarrays Long DNA oligonucleotide (70-mer) probes will be designed by a joint effort of USDA and TIGR upon completion of sequence collection and in silico validation. The oligonucleotides will be synthesized with 5' amine modifications through a custom synthesis and spotted on glass slides. Phase 2 – prototype testing The goals of the Prototype Testing phase, addressed in two experiments, are to ensure that the RNA transcript stocks and the probes to detect each of the external RNA controls exhibit a basic level of functionality. • Test probes for cross-hybridization The first experiment is designed to check for unacceptable levels of cross-hybridization between the full set of candidate probes. To observe hybridization characteristics apart from labeling efficiencies, cRNA targets for each external RNA control will be pre-labeled with biotin, DIG or amino allyl molecules and pooled before hybridization. Assuming 144 candidate external RNA controls, the pre-labeled targets will be split into three pools (e.g., controls 1–48 in pool 1, controls 49–96 in pool 2 and controls 97–144 in pool 3) and a fourth pool of all pre-labeled transcripts will be generated. For one-color platforms, each pool of targets will be hybridized separately in triplicate against the test arrays. For two-color platforms, the pool 1–3 transcripts will be coupled to Cy3, while the pool 4 transcripts will be coupled to Cy5. Each test pool of Cy3-labeled transcripts will be hybridized in triplicate against reference pool 4 of Cy5-labeled transcripts. Each pool will initially be tested in the absence of labeled background cRNA. The expected result is that probes will only show a significant signal when the target they were designed against is in the pool hybridized to the array (i.e., 1 of the 3 pools). Examination of the relative signals for the probes whose targets were hybridized to the arrays should allow identification of probes that show potential cross-reactivity to another sequence included in the same pool. As a negative control, three arrays will be hybridized to cRNA targets generated from at least one representative background human sample in the absence of any external RNA controls. This representative background RNA sample will be from a human tissue (or set of tissues) chosen from the Microarray Quality Control Project [18] currently underway at the FDA-NCTR. As described in the previous section, RNA from other species can also be tested for cross hybridization. Probes with minimal cross hybridization will be incorporated into the set of 96 external RNA controls that are further characterized in phase 3. • Confirm labeling and dose response abilities The second experiment in the Prototype Testing phase will test whether the external RNA controls can be labeled in the presence of a complex background of total RNA. It will also demonstrate their ability to detect known differences in transcript abundance between two pools. One pool will contain external RNA controls in one of two concentrations (e.g., controls 1–72 at 1:10,000 and controls 73–144 at 1:40,000) in the representative human background total RNA. A second pool will be created in which the concentrations are reversed. Both pools will go through three independent target preparation reactions and hybridizations. For one-color platforms, each labeled pool of transcripts will be hybridized to a separate microarray. For two-color platforms, targets from both pools can be labeled with different Cy dyes and hybridized to the same microarray. The observed ratios across the two pools will be compared to the expected 4-fold change. External RNA controls that do not label or give the expected response (something reasonably close to a 4:1 intensity ratio) will be removed from the pool of candidate external RNA controls. Phase 3 – proof of concept • Perform modified latin square and graeco-latin square experiments During the Proof of Concept phase, specific acceptance metrics will be determined and the performance of each external RNA control will be tested over a range of concentrations. 1:5,000,000 to 1:1,000. The experimental design for this phase of array testing must meet the following three criteria: 1) Require a minimal number of arrays and pools; 2) Introduce transcripts as series of pools with balanced cRNA load that are made at a central site, rather than in individual testing labs; and 3) Use the same pools on both one-color and two-color platforms. To best achieve these objectives, this phase of array testing is based on modified versions of a Latin Squares design for one-color platforms and a Graeco-Latin Square design for two-color platforms [19]. Illustrations of these types of experiments are given in Figure 1. Panels A and B describe a 4 × 4 experiment where four different transcripts are tested at four different concentrations. Panels C, D and E show how the same 4 × 4 experiment can be accomplished on two different platforms using the same four pools of transcripts. Figure 1 Illustrations of latin square and graeco-latin square designs. "A1" to "A4" number the 4 arrays used in the experiment, "G1" to "G4" number the 4 transcripts being studied and "L1" to "L4" denote 4 different concentrations for each transcript. The four pools of transcripts are labeled "W" to "Z". "g" and "r" note the gene concentrations or pools used in the green or red channel, respectively of a two-color experiment. Phase 3 will characterize a putative set of clones that are selected based on their performance in phase 2. In this discussion, we will assume the set includes 96 clones. These external RNA controls will be split into four groups of 24 controls each (groups A through D in Table 4). Four pools of external RNA controls will be created such that for each pool, each of the four groups of controls will be at a different concentration (i.e., a different relative mass). For simplicity, multiple transcripts are present at the same concentration in this modified design, rather than each transcript at a different concentration. This design can be referred to as a modified (or semi-) latin square experiment. Table 4 Concentration of controls in dilution 1 pools for modified latin square experiments Pool Concentration Group A (Controls 1–24) Concentration Group B (Clones 25–48) Concentration Group C (Clones 49–72) Concentration Group D (Clones 73–96) Pool 7 125 1 5 25 Pool 8 25 125 1 5 Pool 9 5 25 125 1 Pool 10 1 5 25 125 Concentration is given as mass ratios, so that "125" represents 1:125,000 or 1 ng of RNA transcript per 125 ng of background RNA where the spike amount is adjusted for its length. The four pools will be spiked into four independent target preparation reactions in the presence of the complex background human total RNA. A single experiment will consist of triplicate hybridizations of these four samples (see Table 5). For one-color platforms, each pool will be hybridized to a separate microarray. For two-color platforms, one pool labeled with Cy3 and another pool labeled with Cy5 will be hybridized to the same microarray. As a negative control, three arrays will be hybridized to cRNA targets generated from the background human sample without any external RNA controls. During analysis, data from both the one-color and two-color arrays will be normalized using the distribution of signals from the external RNA control probes, rather than to signals from the probes that hybridize to the background RNA. This normalization approach eliminates the need for dye swap experiments with two-color arrays. Table 5 Modified latin square hybridization setup Controls Group Pool 7 Pool 8 Pool 9 Pool 10 A Conc. 1 Conc. 2 Conc. 3 Conc. 4 B Conc. 4 Conc. 1 Conc. 2 Conc. 3 C Conc. 3 Conc. 4 Conc. 1 Conc. 2 D Conc. 2 Conc. 3 Conc. 4 Conc. 1 • Expand the range of concentrations tested A benefit of this simplified design is that it measures the performance of a large number of external RNA controls with only 12 arrays (plus the three negative control arrays), allowing for wider participation during this phase of testing, instead of limiting participation to those facilities that are able to run the potentially hundreds of arrays required for complete Latin Square or Graeco-Latin Square experiments. The potential drawback to this design is that it allows for measurement of only a limited number of target concentrations. For those interested in measuring the external RNA control performance across a wider range of target concentrations, the experiment can be expanded without the requirement of additional pools. This expansion can be accomplished by diluting each of the pools further to establish other concentration ranges, prior to introducing them into the background human total RNA. Examples of three such dilution ranges are provided in Table 6. In this illustration, the four pools described in Table 3 are diluted 2-fold, 4-fold or 40-fold to generate Dilution 2, Dilution 3 and Dilution 4 pools, respectively. Sixteen possible pools for testing are generated, which are labeled based on their pool and dilution numbers: P7-D1; P7-D2; P7-D3; P7-D4; P8-D1; P8-D2; P8-D3; P8-D4; P9-D1; P9-D2; P9-D3; P9-D4; P10-D1; P10-D2; P10-D3; and P10-D4. Table 6 Concentration of controls in dilution pools for expanded range experiments Pool Conc. A Dil. 2 Conc. B Dil. 2 Conc. C Dil. 2 Conc. D Dil. 2 Conc. A Dil. 3 Conc. B Dil. 3 Conc. C Dil. 3 Conc. D Dil. 3 Conc. A Dil. 4 Conc. B Dil. 4 Conc. C Dil. 4 Conc. D Dil. 4 Pool 7 250 2 10 50 500 4 20 100 5,000 40 200 1,000 Pool 8 50 250 2 10 100 500 4 20 1,000 5,000 40 200 Pool 9 10 50 250 2 20 100 500 4 200 1,000 5,000 40 Pool 10 2 10 50 250 4 20 100 500 40 200 1,000 5,000 Concentration is given as mass ratios, so that "250" represents 1:250,000 or 1 ng of RNA transcript per 250 ng of background RNA, where the spike amount is adjusted for its length. Concentrations of the initial stock, "Dilution 1" pools are shown in Table 4. Participants can choose to test a few or many different ranges. Testing all four ranges (i.e., the original stock plus Dilutions 2–4) would consume 48 arrays (plus the three negative control arrays) and would measure each of the 96 potential external RNA controls at 16 concentrations ranging from an estimated mass ratio of 1:1,000 down to 1:5,000,000. For one-color arrays, each pool would be hybridized to a separate array, so that 48 microarrays would be required to generate triplicate sets of data. For two-color platforms, one pool labeled with Cy3 and another pool labeled with Cy5 will be hybridized to the same microarray. From these experiments we will determine the concentration range over which each target responds linearly. Additionally, we will determine the limit of detection, linear range, and resolvable fold-change across the linear range for each external RNA control. As shown in Table 7, the simultaneous hybridizations on two-color platforms enable differential expression evaluations of red/green ratios from 1/125 (0.008) to 125. Table 7 Expected red:green ratios in two-color hybridizations Array Pool in Green Channel Pool in Red Channel Group A Ratio Group B Ratio Group C Ratio Group D Ratio 1 P7-D1 P10-D1 0.008 5 5 5 2 P8-D1 P9-D1 0.2 0.2 125 0.2 3 P9-D1 P8-D1 5 5 0.008 5 4 P10-D1 P7-D1 125 0.2 0.2 0.2 5 P7-D2 P8-D2 0.2 125 0.2 0.2 6 P8-D2 P7-D2 5 0.008 5 5 7 P9-D2 P10-D2 0.2 0.2 0.2 125 8 P10-D2 P9-D2 5 5 5 0.008 9 P7-D3 P9-D3 0.04 25 25 0.04 10 P8-D3 P10-D3 0.04 0.04 25 25 11 P9-D3 P7-D3 25 0.04 0.04 25 12 P10-D3 P8-D3 25 25 0.04 0.04 13 P7-D4 P7-D4 1 1 1 1 14* P8-D4 P8-D4 1 1 1 1 15* P9-D4 P9-D4 1 1 1 1 16* P10-D4 P10-D4 1 1 1 1 Self-to-self hybridizations Pools in each channel are labeled based on their pool and dilution numbers in Table 6. Groups are sets of transcripts at the same concentration as defined in Table 4. Phase 4 – functional testing • Test dose response curve pools A typical dose response curve (DRC) experiment consists of testing each external RNA control at multiple concentrations, one per array, requiring several arrays. In contrast, a single array DRC experiment consists of multiple concentrations and multiple external RNA controls per concentration, all tested on a single array. Because the entire experiment is contained within a single array, each external RNA control is measured only at a single concentration. Therefore, the intensities from all of the probes are used together to construct the DRC across the entire concentration range. During the Functional Testing phase, the target metrics determined during the Proof of Concept phase will be used to create a single-array DRC external RNA control pool. An example single array DRC pool is shown in Table 8. Table 8 Example single array DRC pool Concentration No. of Targets 1,000 2 2,000 2 4,000 2 5,000 0 10,000 4 20,000 4 25,000 0 40,000 10 50,000 0 100,000 12 125,000 0 200,000 12 250,000 12 500,000 12 1,000,000 12 5,000,000 12 Total 96 Concentration is given as mass ratios, so that "1,000" represents 1:1,000 or 1 ng of RNA transcript per 1,000 ng of background RNA where the spike amount is adjusted for its length To keep the ratio of external RNA controls to background RNA as low as possible, the higher concentrations of external RNA controls (e.g., 1:1,000) are under-represented compared with the lower concentrations (e.g., 1:100,000). Additionally, external RNA controls originating from the same species will be evenly distributed across the concentration range to minimize the effects of removing them from experiments performed with background RNA from the same species. Up to three pooling schemes will be tested, each with triplicate hybridizations. For one-color platforms, each labeled DRC pool of transcripts will be hybridized to a separate microarray. For two-color platforms, targets from different DRC pools can be labeled with different Cy dyes and hybridized to the same microarray (e.g., pool 12 and 13 or pool 13 and 14). • Perform forced failure experiments Once the final single array DRC pool has been chosen, a small series of forced failure experiments will be carried out to mimic how the system will respond to common failure modes (e.g., incorrect temperature, incorrect buffer composition, extreme washing stringency, etc.). Phase 5 – performance review • Review and publish the experimental results Results will be shared within the ERCC for review and evaluation to determine if additional data or redesign is indicated. • Repeat experiments in different labs Tasks for QRT-PCR subgroup The QRT-PCR testing phases are designed to use the same pools of transcripts developed for microarray testing and to evaluate the candidate external RNA controls on multiple QRT-PCR platforms. There are two commonly used methods of detecting the QRT-PCR amplicons: DNA-binding dyes (e.g., SYBR Green I) and target-specific probes with fluorescent 5' exonuclease activity (e.g., TaqMan®). The testing plan will generate data on both platforms, using the same primer sets. The DNA-binding detection assays will use a single tube, one enzyme (rTth) and SYBR Green I based protocols. The target-specific detection assays will use a two-step protocol developed for the TaqMan® system. Phase 1 – design and development • Design QRT-PCR primers and detection probes All primers for external RNA controls will be designed using commonly used software. Two sets of QRT-PCR primers (one at 3' and one at 5' end) will be finalized and used to determine the entirety and specificity of RNA transcripts. For target-specific detection assays, a fluorescent 5' exonuclease detection probe will also be designed. The sequences of all QRT-PCR reagents and their locations on the RNA transcripts as well as the sizes of amplicons and their predicted Tm will be provided to the community. All primers and detection probes will be blasted against Genbank for potential cross-reactivity. • Define cycling parameters With the DNA-binding detection assays, the concentration and ratio of primers must be optimized to prevent fluorescent signal from "primer dimers" or nonspecific amplicons. Experiments will be performed to develop amplification parameters for ABI 7900HT and Stratagene MX3000 using a single tube, one enzyme (rTth) and SYBR Green I based QRT-PCR protocol. With the target-specific detection assays, the primer and detection probe concentrations and thermal cycling conditions will be as suggested by the manufacturer. cDNA will be generated using standard kits and random primers prior to amplification. • Assist in verifying quality of transcript stocks QRT-PCR will be used to assess the quality of the RNA transcripts produced for each of the candidate external RNA controls. Individual transcripts will be evaluated for entirety, stability, and DNA contamination. The specificity of each primer set and detection probe will be determined by amplifying individual transcripts in equal molar pools of the candidate controls. QRT-PCR can also be used to verify transcript ratios in pools with variable concentrations of the candidate controls. Phase 2 – prototype testing • Test primers for cross-reactivity Each set of QRT-PCR primers as well as any detection probes will be tested for cross-reactivity with other external RNA transcripts by amplifying its perspective RNA transcript from three different RNA sources. First, individual transcripts will be amplified from a pool containing multiple RNA transcripts at equal molar concentration (pool 0 in Table 1). Second, individual transcripts will be amplified from the same pools of multiple RNA transcripts at equal molar concentration introduced into a human total RNA background. The human background RNA increases the complexity of the reaction and may be based on a reference RNA sample, for example a pool of RNA derived from 10 different human tissues. Third, individual transcripts will be amplified from the human total RNA that lacks external RNA controls. All amplicons will be examined by gel electrophoresis or melting dissociation curves. The Ct values of amplification for each transcript under other RNA background should be very similar to those with only pure RNA transcripts. If any QRT-PCR reagent gives unexpected products, the primers and/or the detection probe will be re-designed and re-tested. • Optimize efficiency of QRT-PCR A 10-fold serial dilution from 108 copies to 1 copy will be made for a pool of all candidate RNA transcripts in an equal molar concentration. The RNA dilutions will be used for examining the amplification efficiency of at least one QRT-PCR primer set, and for some platforms the corresponding detection probe, for each RNA transcript. Ideally, with a given dilution of the pool, all QRT-PCR primer sets will amplify their perspective targets with similar efficiency. Otherwise, the primer sets and detection probes with a low efficiency should be re-designed and re-tested. • Determine limit of detection Since all molecular weights and sequences are known, the physical copies of each RNA transcript in the pool of all candidate RNA transcripts can be calculated. One RNA transcript will be chosen as a standard for concentration determination and the relative concentrations of other RNA transcripts will be determined based on this standard. All possible efforts should be made to determine the copy number of each transcript by QRT-PCR close to its physical copy number of the RNA transcript in the pool. The lowest concentration of each transcript in the pool of all candidate RNA transcripts, which can be detected by QRT-PCR amplification, can be considered as the limit of detection for that transcript. • Assist in verifying quality of pools QRT-PCR will be used to assess the ratios of different RNA transcripts in different pools. Note that due to different amplification efficiencies in RT and PCR, there may be a discrepancy between the ratios determined by QRT-PCR and those measured by physical quantity. Phase 3 – proof of concept • Establish acceptance criteria for RNA transcripts Based on the phase 2 results, a set of up to 96 external RNA controls will be selected for further testing. Using the 10-fold serial dilutions of a pool of these 96 RNA transcripts (pool 11 in Table 1), limit of quantification, linearity, precision and accuracy for each of RNA transcript will be determined, and the acceptance criteria of assay performance for each RNA transcript will be established. Those criteria should be established for the pure RNA transcripts as well as for the pure RNA transcripts under a background of other total RNA, such as human RNA reference. Phase 4 – functional testing • Compare QRT-PCR platforms and instruments The QRT-PCR testing will generate data from both DNA-binding and target-specific detection assays on multiple thermal cycling instruments, including ABI 7000, ABI 7900HT and Stratagene MX3000. The results will be useful to evaluate the technical performance between those different instruments and platforms. • Compare microarray and QRT-PCR platforms Assay performance of a set of selected external RNA transcripts or pools with the same or different concentrations will be evaluated by microarray and QRT-PCR. The results will be very useful to guide the comparison of experimental results generated by microarray and QRT-PCR. • Optimize external RNA concentrations External RNA controls can be used for monitoring the amplification processes or for detecting the presence of possible inhibitors in QRT-PCR amplification. Optimal concentrations of external RNA transcripts should be established so that the concentrations of the introduced external RNA controls are low enough to detect the presence of possible inhibitors, but not too low that the variation of the assay would be indistinguishable from the low assay performance caused by inhibitors. Different concentrations of a given external RNA transcript will be spiked in total human RNA. The optimal concentration of an external RNA transcript would give a Ct about 30–32. Clinical RNA specimens extracted from different tissues and purified by different sample preparation methods can be used for evaluating the optimal concentrations of external RNA transcripts. • Conduct forced failure experiments To establish the utility of external RNA controls for detecting potential QRT-PCR inhibitors, a few known QRT-PCR inhibitors, such as ethanol and guanidine thiocyanate, will be introduced into QRT-PCR reactions along with external RNA controls. An impaired QRT-PCR reaction of external RNA controls could indicate the presence of possible QRT-PCR inhibitors in the RNA preparation. • Evaluate multiplex QRT-PCR assays Multiplex assays are possible when different reporter dyes are coupled to different target-specific probes so that cleavage of the multiple probes can be detected in a single PCR. These multi-color assays may also be performed. Phase 5 – performance review • Review and publish the experimental results • Repeat experiments in different labs To establish the reproducibility of the research and diagnostic utilities of external RNA controls, the experiments described in phase 4, multi-site testing will be performed. Tasks for informatics subgroup The ERCC is committed to delivering an informatics approach that can be used to establish the technical performance of an expression measurement through the analysis of the measured response of external spiked-in RNA controls. This approach will be implemented and delivered in an open-source manner, such as R code for a bioconductor package. The informatics requirements for ERCC development will include sequence bioinformatics links to public annotation databases. Sequence informatics will be delivered via these public resources. The scope of ERCC informatics activity will not include examination of the application of external RNA controls for purposes other than evaluation of the technical performance of an expression measurement. Such applications as normalization to spike-in controls, or calibration, or evaluation of selectivity and specificity, while potentially useful, are left to assay development efforts for specific intended use applications. Phase 1 – design and development • Develop sequence bioinformatics The sequence data will need to be managed in a sustainable fashion using established tools and approaches. At this stage, sequence bioinformatics includes both a nomenclature system and a maintainable archive of sequences. The nomenclature system requires unique identifiers as well as consistent annotation for RNA control sequences and the related PCR reagents and microarray probes across various platforms. A preliminary nomenclature system is presented in Table 9. Table 9 Illustration of a nomenclature system Reagent Nomenclature Legend RNA Transcript ERCC-nnnnn-vv nnnnn = unique 5-digit sequence number vv = 2 digit version number PCR primer/probe microarray probe ERCC-nnnnn-vv- pppp-lll-aaa pppp = 4-digit positional location relative to the 0th base at the 5' end of the transcript sequence lll = primer/probe length aaa = nucleic acid sequence of the initial triplet of the primer/probe (complement of the RNA sequence at pppp, pppp+1, pppp+2) The sequence archive requirements are as follows: 1) data will be stored in FASTA format; 2) where appropriate, annotation will include unambiguous cross-references to gene identifiers in public databases; and 3) maintainability and version control. • Explore analytical approaches The Analytical Informatics work in this phase is exploratory and will establish prototype tools for the potential analysis approaches. The major tasks for the analytical informatics work at this phase are: 1) identify appropriate QRT-PCR analysis approaches; 2) investigate various analysis approaches for microarray technical performance, including evaluation of preprocessing strategies (e.g., data normalization); 3) develop prototype tools that can be used for the exploration of the analysis approaches; and 4) test prototype tools using existing or synthetic data. The analysis approaches that will be explored in this phase will include: Approach 1 – chi square fit of a linear portion of a calibration curve from the external RNA controls (Figure 2). Figure 2 Illustration of chi square fit. Panel A. The distances from a straight-line fit (arrows) are calculated. Panel B. The Chi square fit of the distances is then determined. Approach 2 – ratio of observed/known concentration ratios between nearest-neighbor spikes versus concentration of low spike in pair (Figure 3). Figure 3 Illustration of spike performance. Approach 3 – linear range 1. To characterize an external RNA transcript response we can describe signal in response to target concentration by the following general function. Here, S – signal from microarray element, a – variable incorporating parameters such as number of probes per microarray element, number of fluorescent label molecules per target molecule, power of light etc., φ – a function describing thermodynamics of hybridization isotherm and this function has a vector of parameters () which is unique for each probe, c – concentrations of spike-in target, b-background due to dark detector counts and sources of light other than labeled targets. 2. Let μs(c) be a function describing relationship between expected signal and spike concentration. Also, let εS(c) be a function describing relationship between standard deviation of signal and spike concentration. 3 Define signals S and S (S < S*) as reliably resolved if Define concentration levels and ( <) as reliably resolved if expected signals produced by these concentrations are reliably resolved. Define concentration fold change – as reliably resolved if concentration levels and are reliably resolved. Define linear range as where and are the ends of the longest interval of target concentrations such that f (ct) ≤ 2 ∀ ct ∈ (, ). Thus, the linear range is the largest continuous interval of target concentrations such that a two fold increase of concentration anywhere within this interval gives rise to reliably resolved expected signal levels. 4. Using these assumptions we will explore several models to determine linear range and minimal reliable resolved concentration fold change. Approach 4 – fitting the spike-in probe behavior on an individual array to a reference model of known, acceptable performance (Figure 4). Figure 4 Illustration of model data (including modeled noise). The values of m and b that were input into the model were m = 0.85 and b = 0.08. The noise model is realistic, in that it includes both constant (scanner) and proportional (chemical) noise. For 1-color arrays signal intensity can be described as: Log2I = target concentration + probe affinity + background + E A reference dataset can be used to estimate probe affinity and background for each of the probes or probe sets. This should allow for a more accurate estimation of the target concentration of the external RNA controls. For 2-color arrays we can use the following model to analyze spike-in ratios: Where m and b are the slope and intercept from a least-squares fit. This model has the nice feature of physically reasonable and appealing interpretations of the slope and intercept. The intercept is the log of the bulk normalization constant for the external RNA control (i.e., the model is self-normalizing, independent of the normalization of the background samples carrying the spike-ins). The slope measures ratio flattening (i.e., reduction of the absolute value of the log ratio from its expected value, due to such effects as cross-hybridization of other targets to the spike-in probes). The model clearly differentiates between failures to observe expected ratios that are due to normalization and effects that are due to ratio flattening. This is important, because the two effects are often confused and misdiagnosed. Approach 5 – ANOVA modeling. There are several potential approaches to analyzing the variance in the experiment: • The variability between repeated probes on the same array. This will give an indication of the variability due to spatial distribution and spot deposition. • Other ANOVA models are possible depending on the experimental design and the variables captured in the experiment. Approach 6 – additional robust measures for performance metrics will be explored. Phase 2 – prototype testing • Develop and test prototype analytic code Implementations for the analytical approaches will be developed in the prototyping phase. These implementations will be tested against modeled/simulated data, as well as against existing data, where possible. As data become available from the QRT-PCR and microarray testing activities, those data will be used to evaluate the prototype analytical implementations. The analyses may be implemented with a variety of numeric and graphical tools (e.g., spreadsheets or proprietary statistical tools). Phase 3 – proof of concept • Qualify analytic models At this stage of the project, data will become available from the array and QRT-PCR testing activities. These data will be used to test the prototype implementations of the analysis approaches, and those approaches will be refined as appropriate. Refinements will be tested and qualified. At the end of this stage, the approaches will be sufficiently refined to move forward to implementation for functional testing. Phase 4 – functional testing • Prototype open source code for analytics This stage of the project will focus on developing implementations of the analysis approaches in R as a bioconductor package. This package should be validated to work with array data from the variety of microarray platforms being tested. Refinement of the graphical display of analysis results will emphasize platform-to-platform consistency where possible (so similar analytical graphs are presented for 1- and 2-color, and the variety of platforms). The QRT-PCR analysis approach may be developed as a separate bioconductor package, with the emphasis likely to be on a graphical display of performance measures as a time series. • Develop web-based interface for analytics Development of web-based implementations of the analysis implementations will be investigated, and common data formats for input and output will be specified and implemented. • Qualify final analytic strategy Phase 5 – performance review • Publish bioconductor package The bioconductor package will be published and released to the community. A manuscript will accompany publication of the package, describing the analytical approaches embodied in the package, and demonstrating performance with the validation data. All relevant performance data from the testing and development of the ERCC analytical informatics will be collated and published at this time. • Publish sequence database The reference sequence database will be published and made available on the web in both flat-file formats and a common sequence database format (FASTA). • Publish relevant performance data Performance data from the testing will be published. Collate performance data (from Test Reports), including probe performance data, in backgrounds. Test plan tasks for ERCC • Identify least burdensome path for collection and distribution of data • Organize analysis jamboree • Plan publication of results and timing for publication(s) • Organize controls symposium Conclusion The tasks described in this document have been designed and reviewed by many ERCC members. They represent our best consensus at this time, but may not be the final format of testing. A summary of the issues discussed at the ERCC Testing Workshop on October 4–5, 2005 is provided in Appendix 2 and copies of the presentations are posted on the NIST website [2]. Additional comments are welcome and should be sent to Dr. Janet Warrington [3]. Abbreviations ERCC, External RNA Control Consortium QRT-PCR, quantitative, real-time reverse transcriptase polymerase chain reaction NIST, National Institute of Standards and Technology polyA, polyadenylated IVT, in vitro transcription MW, molecular weight DRC, dose response curve Cy, Cyanine Dye Ct, Cycle Threshold Appendix Appendix 1: Members of the external RNA controls consortium (Table 10) Table10 Members of the external RNA controls cosortium Anne Bergstrom Lucas Agilent Technologies, Inc. Anne R. Kopf-Sill NuGEN Technologies, Inc Bin Chen Centers for Disease Control and Prevention Bud Bromley ViaLogy Corp. Carole Foy LGC Ltd Cecelia S. Hinkel Centers for Medicare Medicaid Services Cecilie Boysen ViaLogy Corp. Chunmei Liu Affymetrix Inc. Daya Ranamukha-arachchi FDA/CDRH/OSEL Division of Biology Elizabeth Wagar UCLA Ernest S. Kawasaki NCI/NIH Federico M. Goodsaid CDER/FDA Friederike Wilmer QIAGEN GmbH Gavin Fischer Stratagene Gretchen L. Kiser GE Healthcare Helen C. Causton Clinical Sciences Centre/Imperial College Microarray Centre James C. Fuscoe NCTR/FDA James D. Brenton University of Cambridge Janet A. Warrington Affymetrix, Inc. Jesus Soriano ATCC John Coller Stanford University John D. Burrill Applied Biosystems Kate Rhodes Cyntellect Incorporated Kathleen F. Kerr University of Washington Kathryn C. Zoon NIAID/NIH Kathy Lee Applied Biosystems Laura H. Reid Expression Analysis, Inc. Leming Shi NCTR/FDA Marc Salit NIST Mary Satterfield NIST Matthew Marton Rosetta Inpharmatics, LLC Maureen Cronin Genomic Health, Inc. Michael P. Conley Enzo Life Sciences, Inc. Mickey Williams Roche Mike Fero Stanford University Mike Wilson Ambion, Inc. Natalia Novoradovskaya Stratagene Patrick Gilles Invitrogen Paul K. Wolber Agilent Technologies, Inc. Pranvera Ikonomi American Type Culture Collection Raj Puri FDA/Center for Biologics Evaluation and Research Richard P. Beyer University of Washington Richard Shippy GE Healthcare Robert Setterquist Ambion, Inc. Rosalie K. Elespuru FDA/CDRH/OSEL Division of Biology Shawn C. Baker Illumina, Inc. Stephen A. Chervitz Affymetrix, Inc. Steven R. Bauer FDA/Center for Biologics Evaluation and Research Steven Russell University of Cambridge Tamma Kaysser-Kranich GE Healthcare Theo K. Bammler University of Washington Thomas B. Ryder Affymetrix, Inc. Timothy J. Sendera GE Healthcare Uwe Scherf CDRH/FDA Xiaolian Gao Atactic Technologies Xiaoning Wu Roche Molecular Systems, Inc. Xu Guo Affymetrix, Inc. Z. Lewis Liu USDA-ARS-NCAUR Appendix 2: Summary of ERCC test plan workshop The ERCC reviewed feedback on the Proposed Testing Plan at a NIST-hosted workshop on October 4–5, 2005. The meetings were attended by more than 50 participants, including ERCC members from Europe (Belgium, Germany, UK) and the US. During the first day of the meeting, summaries of the test plan were presented by representatives from each of the four subgroups. Suggested improvements and possible testing issues were discussed as listed below: Reagent production Perhaps the ERCC set should include different types of controls (less than 100 bp or no polyA tail) that would support new labeling technologies or gene expression applications. Many members liked this possibility, and although it is outside the scope of the current ERCC initiative, it may develop into future ERCC activities. A request was made for early release of the submitted but not confirmed sequence of the external RNA controls. Another ERCC deliverable describing how to validate the quality of the external RNA controls was suggested. Microarrays The external RNA control concentrations used during phase 2 cross hybridization experiments as well as the exact definition of cross-reactivity were discussed. It was noted that identifying the specific transcript responsible for cross hybridization observed in a pool of 24 candidates may require many subsequent hybridizations with individual controls. Early definitions of the concentrations in the Latin square and dose response curve experiments are needed, with an emphasis on accommodating the ranges of both channels in two-color arrays. After a thorough discussion, it was decided to include pre-labeled external RNA controls in phase 2 of testing. Many of the microarray technology developers indicated that the test arrays might include multiple probes or probe sets in order to identify the best performing sequences. This format was acceptable, as long as participants agreed not to change probe sequence between phases. QRT-PCR The proposed test plan uses dilutions of equimolar pools of 96–140 the candidate RNA controls in the QRT-PCR assays. At the meeting, it was decided that plates of individual RNA transcripts would be more useful during the initial testing in phases 1–4. It was recognized that only 1–3 external RNA controls at a time would be used in QRT-PCR assays due to limitations in the number of fluorophors that can be detected. There was also some discussion that QRT-PCR assay optimization may require more than 2 primer sets described in the proposed test plan. A new application for the controls in the validation of thermal cyclers and other QRT-PCR instruments was discussed. Informatics The ERCC will need a plan for data storage and distribution. NCI, NCBI, and FDA representatives volunteered to provide a data repository for Test Plan data. A working group will be formed to identify needs of the ERCC and identify which of these data repositories would be a good fit. Participants were directed to the CLSI MM16-P document for a list of possible performance metrics. Preliminary review of these metrics could be initiated now using already available gene expression data on many of the candidate external RNA controls. There was some discussion on whether the external RNA controls should be optimized and selected to have the best possible results (e.g. perfect dose response curve, no cross hybridization) or left imperfect to better reflect typical probes or probe sets on the microarrays. After nine hours of review and discussion, consensus was achieved on the ERCC Test Plan by a show of hands with no major outstanding issues. On the second day of the meeting, preliminary resource requirements were identified and the scope of testing (sample number etc.) was better defined. Key features of this discussion are summarized below: 1. In an effort to efficiently utilize the RNA reagents, early stages of testing (Phases 1–4) will be performed only at developer sites and limited to one site per platform. We estimate that up to 10 microarray and 4 QRT-PCR commercial developers will participate. If resources permit, 3 or 4 noncommercial development sites (e.g. NCI or USDA) will be included to represent home-made, spotted microarrays. 2. Phase 5 will be divided into two parts with separate goals. In Phase 5A, the last experiment of Phase 4 will be repeated at user sites to confirm that results can be reproduced using the same materials and protocols as at the developer sites. In Phase 5B, a set of the best performing external RNA controls will be distributed to user sites and incorporated into typical experiments with their materials and protocols. The intent is to establish performance of the external RNA controls in a broad variety of routine applications and protocols. One possible Phase 5B experiment would be to repeat a previous gene expression experiment to validate that similar differential expression results are observed with and without the inclusion of external RNA controls. 3. A tentative list of criteria for technology developers and user test sites was developed. It was decided that all sites should have the following attributes: 1) Demonstrated experience in gene expression platform; 2) Ability to contribute materials and labor; 3) Commitment to deposit all testing data in publicly available database; and 4) Agreement to meet ERCC timeline. In addition, developer sites should have demonstrated capacity to design and create the necessary reagents as well as a commitment to follow the test plan and collect data on all candidate clones. User test sites may be asked to contribute a novel application to the ERCC goals and to provide sample material for the complex RNA background. It is not necessary to have contributed sequences to be considered as a test site. 4. While the external RNA controls are likely to be useful in RNA samples from a variety of species, testing will emphasize human, mouse and rat RNA. A Stratagene representative offered to provide universal reference RNA samples from these three species to be used as the complex RNA background in Phases 1–4. Test microarrays will include probes or probe sets for both the candidate external RNA controls as well as genes included on their typical human, mouse and rat arrays (as determined by each developer). This design will enable global normalization methods and aid in cross hybridization experiments. The developers may elect to pre-screen their test arrays using RNA from a variety of species. Action items and next steps will be further discussed in the monthly ERCC conference calls. Note 3Certain commercial equipment, instruments, or materials are identified in this document. Such identification does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the products identified are necessarily the best available for the purpose. ==== Refs Cronin M Ghosh K Sistare F Quackenbush J Vilker V O'Connell C Universal RNA reference materials for gene expression Clin Chem 2004 50 1464 1471 15155546 10.1373/clinchem.2004.035675 National Institute of Standards and Technology Dr. Janet Warrrington mailto: [email protected] Affymetrix GeneChip® Poly-A RNA Control Kit GE Healthcare/Amersham Biosciences Codelink™ Whole Genome Controls Stratagene SpotReport™ Alien™ Array Validation System Choe SE Boutros M Michelson AM Church GM Halfon MS Preferred analysis methods for Affymetrix GeneChips revealed by a wholly defined control dataset Genome Biology 2005 6 R16 15693945 10.1186/gb-2005-6-2-r16 Kuhn K Baker SC Chudin E Lieu M-H Oeser S Bennett H Rigault P Barker D McDaniel TK Chee MS A novel, high-performance random array platform for quantitative gene expression profiling Genome Res 2004 14 2347 2356 15520296 10.1101/gr.2739104 Lockhart DJ Dong H Byrne MC Follettie T Gallo MV Chee MS Mittmann M Wang C Kobayashi M Horton H Brown EL Expression monitoring by hybridization to high-density oligonucleotide arrays Nature Biotechnology 1996 14 1675 1680 9634850 10.1038/nbt1296-1675 Applied Biosystems ABI Prism 7770 Sequence Detection System User Bulletin #2 Innis MA Gelfand DH Sninsky JJ PCR Applications: Protocols for Functional Genomics 1999 Academic Press Vandecasteele SJ Peetermans WE Merckx R Van E Quantification of expression of Staphylococcus epidermidis housekeeping genes with Taqman quantitative PCR during in vitro growth and under different conditions J Bacteriol 2001 183 7094 7101 11717267 10.1128/JB.183.24.7094-7101.2001 Radonic A Thulke S Mackay IM Landt O Siegert W Nitsche A Guideline to reference gene selection for quantitative real-time PCR Biochem Biophys Res Commun 2004 313 856 862 14706621 10.1016/j.bbrc.2003.11.177 Janssens N Janicot M Perera T Bakker A Housekeeping genes as internal standards in cancer research Mol Diagn 2004 8 107 13 15527325 Jeong YJ Choi HW Shin HS Cui XS Kim NH Gerton GL Jun JH Optimization of real time RT-PCR methods for the analysis of gene expression in mouse eggs and preimplantation embryos Mol Reprod Dev 2005 71 284 9 15806558 10.1002/mrd.20269 Lee PD Sladek R Greenwood CMT Hudson TJ Control genes and variability: absence of ubiquitous reference transcripts in diverse mammalian expression studies Genome Research 2002 12 292 297 11827948 10.1101/gr.217802 van de Peppel J Kemmeren P vanBakel H Radonjic M vanLeenen D Holstege FCP Monitoring global messenger RNA changes in externally controlled microarray experiments EMBO Reports 2003 4 387 393 12671682 10.1038/sj.embor.embor798 MicroArray Quality Control Project Box GEP Hunter WG Hunter JS Statistics for Experimenters 1978 New York: John Wiley & Sons	16266432	PMC1325234	CC BY	2021-01-04 16:32:47	no	BMC Genomics. 2005 Nov 2; 6:150	utf-8	BMC Genomics	2,005	10.1186/1471-2164-6-150	oa_comm
==== Front BMC Fam PractBMC Family Practice1471-2296BioMed Central London 1471-2296-6-491635171410.1186/1471-2296-6-49Research ArticlePrimary healthcare provision and Chronic Fatigue Syndrome: a survey of patients' and General Practitioners' beliefs Thomas Marie A [email protected] Andrew P [email protected] Centre for Occupational and Health Psychology, School of Psychology, Cardiff University, 63 Park Place, Cardiff, UK2005 13 12 2005 6 49 49 15 8 2005 13 12 2005 Copyright © 2005 Thomas and Smith; licensee BioMed Central Ltd.2005Thomas and Smith; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background The current study was conducted as part of a research project into the evaluation and assessment of healthcare provision and education in Chronic Fatigue Syndrome (CFS). One aim of the study was the development of informative and educational literature for both General Practitioners (GP) and sufferers. Issues such as diagnosis, management and treatment of the syndrome should be included in information booklets written by healthcare professionals. It was important to begin the process by assessing the level of specialist knowledge that existed in typical GP surgeries. This data would then be compared to data from CFS patients. Method 197 survey booklets were sent to CFS sufferers from an existing research panel. The patients approached for the purpose of the study had been recruited onto the panel following diagnosis of their illness at a specialised CFS outpatient clinic in South Wales. A further 120 booklets were sent to GP surgeries in the Gwent Health Authority region in Wales. Results Results from the study indicate that the level of specialist knowledge of CFS in primary care remains low. Only half the GP respondents believed that the condition actually exists. Conclusion Steps are recommended to increase the knowledge base by compiling helpful and informative material for GPs and patient groups. ==== Body Background A patient with Chronic Fatigue Syndrome (CFS) is described as one suffering unrelenting, debilitating fatigue (for a period of six months or more) which is unresolved by rest. This fatigue is not the result of normal physical activity and can cause both mental and physical impairment to the sufferer. Furthermore, the fatigue experienced is not as a result of an ongoing medical condition. Chronic Fatigue Syndrome remains a poorly understood condition and still poses problems in terms of causality, diagnosis and management for clinicians and researchers alike [1]. The myriad symptoms of the syndrome also present major diagnostic problems for primary healthcare providers. Unfortunately, lack of specialised knowledge (within the healthcare system) and scepticism on the part of some often leads to a breakdown in trust and confidence between patient and physician. This problem was highlighted in an investigation of perceptions in patients with CFS who had been referred to a specialised clinic [2]. 68 patients completed a survey assessing their satisfaction with the medical care offered at the clinic. Two-thirds of the sample expressed feeling dissatisfied with the quality of care received during their illness. Furthermore, these patients were more likely to describe delay, dispute or confusion over diagnosis. Many of these same patients had received a psychiatric diagnosis for their symptoms which they rejected. In addition, this sub-group of patients perceived doctors as dismissive, sceptical or lacking in knowledge of CFS and felt that advice given was inadequate or conflicting. In contrast, satisfied patients believed doctors to be sympathetic and supportive of their condition. A major conclusion drawn from this study was the importance of understanding and effective communication between doctor and patient in dealing with CFS. Patients, it would seem, preferred GPs who, although admitting a lack of knowledge on the subject, offered empathy and support. There are many avenues open to sources of information on CFS for doctors and patients alike. Unfortunately, too often they offer inaccurate and conflicting advice. In an age of increasing access to the internet, an Australian survey [3] reviewed 225 websites over a two week period. Widely differing views were found from websites offering information with regard to treatment for CFS. There was, however, general agreement that graded exercise and avoidance of prolonged rest were the most successful management strategies for sufferers. 64% of the sites offering advice had a named author. However, only a quarter to a third of the sites reviewed advised readers to clarify the information proffered with an appropriate health physician or avoided the inclusion of inaccurate statements. The report concluded that physicians should provide guidance for patients as to which internet sites to trust. It also recommended that GPs should be made fully aware of the nature of the information being accessed by patients. Misinformation leading to possible distress for the patient could then be avoided. The strategies, employed by physicians in Sweden, to categorise, diagnose and treat CFS and fibromyalgia (FM) patients was investigated by Åsbring and Närvänen in 2003 [4]. During their study twenty-six physicians, all of whom had knowledge of working with CFS or FM, completed a semi-structured interview. Results from the study suggested that there was a discrepancy between the ideal role that the physician wished to fulfil and the reality of everyday work involving interaction with CFS patients. The physicians were concerned that their lack of specialist knowledge prevented them from providing proper healthcare support for their patients. This, the authors' concluded, led to the professional role being questioned by the patient. The point was also raised that some physicians were viewing CFS as a less serious illness than those conditions deemed to have 'disease status'. Further to this, scepticism was expressed on the part of the physicians as to the actual existence of CFS. Indeed, further studies have supported these findings [5]. It can be inferred from these studies, therefore, that there is continued diversity within the primary healthcare setting. Views ranging from questioning the existence of the syndrome to differing modes of diagnosis and management have been recently reported [5]. As an example of this, the Department of Public Health and Primary Care at the University of Hull conducted a national survey of General Practitioners (GP) and their beliefs regarding CFS [6]. Their research produced a comprehensive report on the current state of affairs within the United Kingdom. 300 questionnaires were sent out to GP surgeries in ten Regional Health Authorities in England, Scotland and Wales. Five of the Authorities surveyed had specialised centres for CFS, five did not. The five regions that did not have specialised centres for CFS were matched as closely as possible to those that did. One conclusion from the study indicated that although the GPs in the areas with specialised CFS clinics were more likely to belief that the condition existed, there was no difference in their 'propensity to diagnose' than those in areas without specialised services. This suggests that although GPs in the areas with specialised centres were aware of the clinics' existence, there was limited flow of specialist knowledge from the centres to primary care. GPs' perceptions of patients with CFS have been studied in comparison with other syndromes. In 2004, Raine et al. [7] compared GP beliefs regarding patients with CFS to those with Irritable Bowel Syndrome (IBS). Their findings indicated that the attitudes of GPs to either CFS or IBS dictated subsequent management of the illness. The research concludes that these perceptions would ideally need to change to facilitate successful treatment implementation. The aetiology, diagnosis, management and treatment for patients with CFS remain unclear and the need for further research into this condition is vital. This paper does in some cases set out to replicate the findings of other studies but was conducted as part of a wider research project investigating healthcare evaluation and patient education in CFS. Objectives of the present study The aim of the current study was to investigate the opinions of CFS sufferers themselves, regarding diagnosis and treatment, and compare them to the current thinking of GPs from a different Health Authority but within the same geographical region. We could then address the question of whether the situation had changed in the light of two important reports being in the public domain, namely those by the Royal Colleges [8] and the National Task Force on CFS/ME [1]. Information from the study could then be used to develop educational literature for both GPs and patients regarding CFS diagnosis and management. Methods Ethical approval for this study was granted by the Gwent Health Authority. All data were coded to ensure the anonymity of both the patients and the GP Surgeries taking part. Design The study took the form of a simple survey proforma and the questions designed as a preliminary point of reference for the production of educational literature for patient and GPs. Participants Patient sample Patient recruitment was from an existing research panel. All of these volunteers had been diagnosed using the Centre for Disease Control (CDC) criteria for CFS [9] at a specialised outpatient clinic some years previously. 197 CFS sufferers were surveyed by postal questionnaire. Primary care sample 120 questionnaire booklets were distributed by members of staff at the Gwent Health Authority Headquarters into the official postbags for the area's GP practices. Neither the GPs nor patients were sent reminders to return the booklets following the first mail shot. Procedures Questionnaires Two short booklets were compiled, by the authors, to glean us much comparable data between the patients and GPs as possible. The booklets complied for the GPs were done so in as concise a manner as possible in order to maximise response rates in a profession where time is limited. Patient booklets elicited similar information in order to establish comparability with the GP sample. However, patients were also required to comment on any therapy they might have received and their current state of health. In this way it was hoped that data collected from the research panel regarding past diagnosis and management could be compared to the up-to-date information given by the GPs. In this way we would be collecting data relative to our research based on work from previous studies [6]. Patient questionnaires First and most importantly, the patients were asked if they were still suffering from CFS, and for how long the condition had presented itself. The questionnaire then went on to illicit information regarding the level of primary healthcare received. This included the diagnostic tests and treatment options offered. Patients were also asked if any of the management/treatments offered were successful and asked to rate their health status using a previously validated current state of health measure [10]. This measure assesses the severity of their illness on a 5–item scale ranging from 'worse than at any stage' to 'almost completely recovered'. GP questionnaire The GPs surveyed were asked two fundamental questions: (a) did they believed that there was a single entity called Chronic Fatigue Syndrome (often known as Myalgic Encephalomyelitis), and, if so, (b) had they ever diagnosed patients with this illness. If the respondent answered 'no' to both of the above questions they were asked to return the survey. GPs who answered 'yes' were then asked to supply details of diagnostic criteria [9,11] and management regimes offered to their patients. They were also asked if their surgery carried any information booklets for patients and if so their source. Both patients and GPs were asked if they would be prepared to comment on literature complied by healthcare professionals in the future. Data analysis Descriptive statistics were performed on the categorical and continuous data. Open-ended questions were collated and categorised. Results 92 patient questionnaires were completed and returned giving a 48% response rate. A further 21 were returned to sender leaving 84 unaccounted for. Of the questionnaires distributed to the GPs, 45 were returned, two of which were blank giving a 39% response rate. Patient survey Of the 92 patient respondents, 78 reported that they were still suffering from CFS (84.8%), indicating a 15.2% recovery rate for the sample. The mean illness duration for the group was 13.14 years (range = 3 to 32 years, s.e.m = 0.63). When asked to rate their current state of health, 2.2% of the sample reported feeling 'worse than at any stage' of their illness, 16.3% reported feeling 'bad', 32.6% were feeling 'bad with some recovery', and 33.7% were 'recovering with occasional relapses'. 51.6% of the sample indicated that their GP had diagnosed their condition, taking on average 6.58 (range = 2 to 20 appointments, s.e.m = 0.78) appointments to do so. When asked if the patient believed that they were suffering from CFS before their GP's diagnosis, 52.6% stated that this was the case. The patients were then asked to whom they had turned to for information on CFS (other than their GPs). The majority of respondents (52.2%) had contacted the ME Association. However, 63% stated that they had gained information form 'other sources'. These included 'friends or colleagues with CFS' (17.2%) and newspaper or magazine articles (62.1%). The patient sample was then questioned about diagnosis and management. 82.6% of those surveyed reported that their GP had conducted investigative tests to exclude other diseases. These tests included: (a) a full range of blood tests (90.9%), (b) a test for the Epstein-Barr virus (51.9%), (c) tests for other viral infections (48.1%), and (d) 'other' tests (18.2%). Those described as 'other tests' included thyroid function tests (n = 4), hormone function tests (n = 3), rheumatoid factor (n = 3), ECG (n = 1), lung function (n = 1) and MRI scan (n = 1). In terms of CFS management, 59.8% of the patients had been offered treatment by their GP, the most popular being antidepressant therapy (92.7%) or analgesics (not including 'over the counter' medicines) (36.4%). Of the 51 patients who had taken antidepressant medication, 18 reported that this form of therapy had made their symptoms worse, twenty-two reported no change in their symptoms, nine said that the therapy had improved their symptoms and one patient reported that antidepressant therapy had returned them to normal health. One patient did not respond to this follow-up question. Twenty patients reported that they had been prescribed pain relief to manage their symptoms. One reported that analgesics had made their symptoms worse; thirteen reported no change in their symptoms and six felt that the analgesic medication had made their symptoms better. Other management strategies offered by the GPs in the survey included Cognitive Behaviour Therapy (CBT), Grade Exercise Therapy (GET), Occupational Therapy (OT) and Counselling. 33 of the 92 respondents had also been referred to hospital consultants (other than the one who later confirmed the diagnosis of CFS). Table1 lists the outpatient departments attended by the survey responders before being referred to the specialised clinic. Table 1 Outpatients departments attended by CFS sufferers before attending a dedicated CFS clinic. Outpatient Department Number of Attendees General Medicine 16 Cardio/Thoracic 3 Psychological Medicine 3 Immunology 2 Dentistry 1 ENT 1 Neurology 1 Virology 1 In addition to these data two patients attended Homeopathy clinics and another received a private consultation. The remaining two patients did not offer a response to the question. 77.2% of the survey respondents had tried 'alternative' therapies to alleviate their symptoms spending on average £981. One respondent reported spending as much as £7000 on alternative therapies. Finally, 89% of the patients sampled said that they would be willing to comment on information booklets aimed at CFS diagnosis and treatment. GP survey Of the 45 GP respondents, 55.8% believed that the condition called CFS existed and 67.4% of these had diagnosed patients with CFS. On average, 6.2 (s.e.m = 0.97) separate appointments were required to diagnose the condition. None of the GPs who completed the survey used the CDC or Oxford criteria for CFS, preferring to either conduct investigative tests to rule out other illness (68.8%) and/or refer the patient on to tertiary care (65.6%). When considering the sub-group of GPs who reported diagnosing CFS, 89.3% offered treatment strategies to the patient. None of the GP surgeries had trained nurses, occupational therapists or physiotherapists capable of offering support, advice or treatment to sufferers in the primary care setting. They also reported being unaware if any such services were presently available in their locality. The most common form of treatment offered by the GPs who responded to the survey was antidepressant therapy. 84% of GPs prescribed Selective Serotonin Re-uptake Inhibitors (SSRI) antidepressants, 28% preferring Serotonin/Noradrenalin Re-uptake Inhibitors (SNRI) and 24% prescribing the Tricyclic and related antidepressants. Only 14.8% of the surgeries surveyed carried information leaflets on CFS. Most of the literature, it was reported, was supplied by the ME Association. In terms of referrals to tertiary care, 56.7% of the GPs surveyed were aware that there was a consultant in the area who specialised in CFS. 16.7% referred patients to General Medical out-patients clinics, 6.7% to Rheumatology clinics, 6.7% to Neurology clinics and 6.7% referred patients to Psychological Medicine. 54.5% of the GP respondents were prepared to answer a more detailed questionnaire at a future date and 42.4% were willing to comment on the information leaflets referred to previously. Discussion This paper aims to describe the current thinking of GPs from a single health authority in Wales. The data was collected as part of an ongoing project which included, amongst others, the need highlight whether GPs were being made aware of up-to-date information on CFS centres of excellence, its diagnosis and management. If not, our aim was to rectify this by offering to provide GP surgeries with information compiled by healthcare professionals in the field of CFS research. It is acknowledged that the response rates, by both patients and GPs, for the current survey may appear to be low. However, a recent survey of members of local ME groups (supported by Action for ME and the ME Association) recorded patient response rates of 47% [12]. Furthermore, a ten-centre survey [5] reported GP response rates ranging from 35% to 55%. The latter is in sharp contrast to data presented recently by Bowen et al. [13] indicating a 77% GP response rate to their CFS survey. The data from this survey, however, was collected from GP surgeries served by medical laboratories within their region which may have acted as an incentive to respond. In addition to the differences in the method of sampling, data indicating initial response rates are not recorded; only those from the post-follow-up. With this in mind, the response rates of 48% for patients and 39% for the GPs in the present study seem more indicative of the types of group sampled. We can, therefore, put forward the view that the data reported here does represent an accurate portrayal of patient and GP opinions as long as it is discussed in relation to the situation within Wales and not to the UK as a whole. To further support this, data from the patient research panel group includes respondents who have recovered from Chronic Fatigue Syndrome (CFS). The current state of health measure also indicates that the health status of the group follows a similar profile to that of patients from previous studies (Thomas and Smith, in preparation). Likewise, the GP respondents are split approximately fifty-fifty between those who believe that the condition called CFS exists and those who do not. Therefore, no bias on the basis of patient 'wellness' or GP 'belief' in CFS is indicated here. Scepticism on the part of GPs in recognising that CFS actually exists remains a problem to this day. Only 56% of the GP responders believe that CFS is a recognised condition despite findings from reports by the joint Royal Colleges and the National Task Force being in the public domain. Of the 44% who did believe that the illness exists, none reported using the CDC or Oxford criteria for CFS definition. This is surprising as both case definitions are readily available to medical and research staff and patient groups alike. When questioned, only 57% of the GPs surveyed were aware that a CFS specialist was consulting within their local health authority region. The majority of those who were not aware of this referred patients to general medical outpatient clinics. This has been problematical in the past. Unless the patient is fortunate enough to be referred to a physician who, if not knowledgeable on the subject, is aware of specialist help, this will invariably result in the patient being told that there is 'nothing physically wrong with them'. The patient then returns to a GP who has two courses of action open to them: refer the patient to another outpatient department or try to manage the patient's condition themselves. This is bound to result in frustration on the part of the physician, who has the patient's best welfare at heart, as much as the patient. Comparisons between the patients who had received a diagnosis from their GP and the GPs, who reported diagnosing CFS, both indicate that the process took approximately 6 appointments. Interestingly, the range of 2 to 20 appointments to diagnose the condition is identical for both groups. It is important during the process of diagnosing CFS, that other illnesses presenting fatigue-like symptoms are ruled out. However, only two-thirds of the GP respondents reported conducting further investigations to exclude these conditions. It is encouraging to note that more GPs are currently offering treatment strategies compared to the past (89% and 60% respectively). However, antidepressants remain the preferred mode of treatment. Antidepressant therapy does have its role to play in treatment strategies in certain circumstances as described previously (Thomas and Smith, in preparation). But reports by CFS patients of heightened sensitivity to such medication have been widely documented and antidepressants should be prescribed with caution. In addition, findings from successful treatment trials of CBT and GET for the treatment of CFS do not seem to have filtered through to primary healthcare. The authors acknowledge that General Practitioners' time and resources are limited and that being able to keep up with advances in research is a luxury they can ill-afford. Following the report to the Chief Medical Officer, the Medical Research Council recently set aside a considerable sum of money to support CFS research projects and subsequent information dissemination within the UK. Unfortunately, none of the funding found its way to projects in Wales. This means that the Principality currently trails behind the rest of the country in terms of resources available for research in this area. Due to this short-fall in Welsh funding, it is not surprising that the GPs represented in our survey lack confidence when dealing with patients with CFS. On a positive note, almost half of the GPs surveyed would welcome helpful, practical advice written by healthcare professionals when dealing with patients whom they suspect may have CFS. However, the state of affairs with regard to the past experiences of the research panel patients and the current opinions of the GP respondents is all too familiar. Conclusion The proposed next step is to produce informative material for both GPs and patients. This material needs to be compiled in conjunction with CFS specialists and will include details of centres of excellence, diagnosis and management. Abbreviations CBT – Cognitive Behaviour Therapy GET – Graded Exercise Therapy CFS – Chronic Fatigue Syndrome CDC – Centre for Disease Control and Prevention GP – General Practitioner IBS – Irritable Bowel Syndrome OT – Occupational Therapy SNRI – Serotonin/Noradrenalin Re-uptake Inhibitor SSRI – Selective Serotonin Re-uptake Inhibitors Competing interests The author(s) declare that they have no competing interest. Authors' contributions MAT study design, recruitment of participants, collation of databases, data analysis and manuscript preparation. APS advice on study design and manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements This research was funded by the Gatsby Foundation. We would like to express our thanks to Dr J Watkins and staff at the Gwent Health Authority Head Office for their assistance. ==== Refs National Task Force on CFS/ME Bristol: Westcare NHS services for people with CFS/ME 1998 Deale A Wessely S Patients' perceptions of medical care in CFS Social Science and Medicine 2001 52 1859 1864 11352411 10.1016/S0277-9536(00)00302-6 Kisely S Treatments for CFS and the internet: a systematic survey of what your patients are reading Australia and New Zealand Journal of Psychiatry 2002 36 240 245 10.1046/j.1440-1614.2002.01017.x Åsbring P Närvänen A Ideal versus reality: physicians' perspectives on patients with CFS and Fibromyalgia Social Science and Medicine 2000 57 711 720 Steven S General Practitioners' beliefs, attitudes and reported actions towards CFS Australian Family Physician 2000 29 80 85 10721550 Campion P A report to the Linbury Trust National Survey of General Practitioners' beliefs about CFS/ME 2004 Raine R Carter S Sensky T Black N General Practitioners' perceptions of CFS and beliefs about its management, compared with irritable bowel syndrome: qualitative study British Medical Journal 2004 CFS/ME Joint Working Group London, CFS/ME Joint Working Group A report of a Joint Working Group between the Royal College of Physicians, the Royal College of General Practitioners and the Royal College of Psychiatrists 1996 Fukuda K Straus S Hickie I Sharpe MC Dobbins JG Komaroff A International Chronic Fatigue Syndrome Study Group The Chronic Fatigue Syndrome: a comprehensive approach to its definition and study Annals of International Medicine 1994 121 953 959 Smith A Pollock J Thomas M Llewelyn M Borysiewicz L The relationship between subjective ratings of sleep and mental functioning in healthy subjects and patients with chronic fatigue syndrome Human Psychopharmocology 1996 11 161 167 10.1002/(SICI)1099-1077(199605)11:3<161::AID-HUP784>3.0.CO;2-5 Sharpe M Archard L Banatvala J Borysiewicz L Clare A David A Edwards R Hawton K Lambert H Lane R McDonald E Mowbray J Pearson D Peto T Preedy V Smith A Smith D Taylor D Tyrrell D Wessely S White P CFS: guidelines for research Journal Royal Society of Medicine 1991 84 118 121 Cooper L A report in conjunction with Action for ME and ME Association Report on survey of local ME group members 2000 Bowen J Pheby D Charlett A McNulty C Chronic Fatigue Syndrome: a survey of GPs' attitudes and knowledge Family Practice 2005 389 393 15805128 10.1093/fampra/cmi019	16351714	PMC1325235	CC BY	2021-01-04 16:29:14	no	BMC Fam Pract. 2005 Dec 13; 6:49	utf-8	BMC Fam Pract	2,005	10.1186/1471-2296-6-49	oa_comm
==== Front BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-1091633664210.1186/1471-2334-5-109Case ReportCoxiella burnetii vascular graft infection Senn Laurence [email protected] Mario [email protected] Didier [email protected] Alexandre [email protected] Segesser Ludwig [email protected] Thierry [email protected] Gilbert [email protected] Infectious Diseases Service, Department of Internal Medicine, University Hospital, Lausanne, Switzerland2 Bellinzona Hospital, Switzerland3 Unité des Rickettsies, Université de la Méditerranée, Marseille, France4 University Institute of Pathology, Lausanne, Switzerland5 Department of Cardio-vascular Surgery, University Hospital, Lausanne, Switzerland6 Institute of Microbiology, University of Lausanne, Bugnon 48, 1011 Lausanne, Switzerland2005 7 12 2005 5 109 109 7 9 2005 7 12 2005 Copyright © 2005 Senn et al; licensee BioMed Central Ltd.2005Senn 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 Coxiella burnetii, the causative agent of Q fever, may cause culture-negative vascular graft infections. Very few cases of C. burnetii infection of a vascular graft have been reported. All were diagnosed by serology. Case presentation We report the first case of Coxiella burnetii vascular graft infection diagnosed by broad-range PCR and discuss the diagnostic approaches and treatment strategies of chronic C. burnetii infection. Conclusion C. burnetii should be considered as etiological agent in patients with a vascular graft and fever, abdominal pain, and laboratory signs of inflammation, with or without exposure history. Broad-range PCR should be performed on culture-negative surgical samples in patients with suspected infection of vascular graft. ==== Body Background Infection of synthetic abdominal aortic grafts occurs in ≤1% of patients, with a higher risk (1.5–2%) for grafts that extend to the femoral location. Vascular graft infection may result from intra-operative contamination, local extension from infected adjacent tissue or by hematogenous seeding. The most commonly involved pathogens are Staphylococcus aureus (30%), Enterobacteriaceae (25%), coagulase-negative Staphylococci (12%), Enterococci (9%), Pseudomonas aeruginosa (7%) and Streptococci (5%)[1]. Cultures remain negative in approximately 5% of cases [1]. C. burnetii account for some of these culture-negative vascular graft infections. Very few cases of C. burnetii infection of a vascular graft have been reported [2-5]. All previously reported cases were diagnosed by serology. The confirmation of the vascular localization of C. burnetii infection was obtained after the serological diagnosis of chronic Q fever by culture [3] and/or DNA amplification of C. burnetii from vascular graft samples [3-5]. Here, we report a case of C. burnetii vascular graft infection diagnosed by broad-range PCR from a surgical sample of a para-prosthetic abscess which was confirmed by serology. To our knowledge, ours is the first case where the diagnosis was made by broad-range PCR and suggests that broad-range PCR should be considered in all cases of culture-negative vascular graft infections. Case report A 63-year-old man presented to a regional hospital on September 8, 2003 with a 2-week history of diffuse abdominal pain and mild diarrhea, without fever. In 1988, he had received a Dacron aorto-bifemoral graft for an infra-renal aortic aneurysm. A computerized tomography (CT) of the abdomen revealed a para-prosthetic fluid collection. Blood cultures were sterile in the absence of any recent antibiotic therapy. Laboratory results showed a white blood cell count of 5.8 G/l, a CRP of 48 mg/l, no increase of liver enzymes and a normal serum creatinine level. Empirical ciprofloxacin and metronidazole therapy was initiated and abdominal pain improved. After two months of antibiotic therapy, the patient was admitted to the University Hospital in Lausanne for removal of the vascular prosthesis because of presumed persistent infection, despite two months of antibiotic treatment. On admission, the patient was afebrile. Clinical examination was normal except for mild periumbilical tenderness on deep palpation. Laboratory results showed a normal WBC count (4.9 G/l), a normal CRP (<2 mg/l), and normal renal and liver functions. At laparotomy, extensive adhesions and a right para-iliac purulent mass were found. The prosthetic graft was partially removed, and replaced by a homograft. Multiple intra-operative specimens did not grow any microorganisms in culture. Histopathology showed a chronic inflammatory infiltrate, ill-formed non-necrotizing granulomas, and degenerative changes such as calcifications and fibrosis (Figure 1A &1B). No microorganisms could be identified using Periodic acid-Schiff, Gram, Grocott methenamine silver and Giemsa stains. Figure 1 Histology of the aortic lesion: A. Chronic inflammatory infiltrate (yellow arrowhead), fibrosis (black arrowhead), and ill-formed granuloma (arrow). Hematoxylin-eosin, 100× magnification. B. Closer view of the ill-formed granuloma (arrow). Hematoxylin-eosin, 400× magnification. 16S rRNA PCR amplification plus sequencing performed on a fragment of the para-iliac mass was positive for Coxiella burnetii, using the BAK11w forward and the PC3mod reverse primers [6]. The diagnosis of C. burnetii chronic infection was confirmed by a positive serology performed at Unité des Rickettsies, Marseille, France: phase I antibody titer (IgG): 1600, phase II antibody titer (IgG): 3200. Antibiotic therapy with doxycycline (100 mg bid orally) and chloroquine (200 mg tid orally) was started. The dose of doxycycline was increased to 300 mg daily to reach a concentration of at least 5 μg/mL [7]. Eighteen months later (May 2005), the patient was asymptomatic and serology showed persistence of high levels of phase I IgG (1600) and phase II IgG (3200). C. burnetii is a strict intracellular bacterium. It is the causative agent of Q fever, a worldwide zoonosis mainly transmitted by inhalation of infected particles. There is a wide animal reservoir, and sheep and cattle are probably the main source of human infections. Our patient had no environmental exposure to C. burnetii. Chronic Q fever is mainly seen in patients with underlying risk factors such as valvulopathy, pregnancy, and immunosuppression. Endocarditis accounts for 73% of chronic Q fever cases, followed by vascular infection (8%), including infections of aneurysms and vascular grafts [4,8]. Given the significant morbidity and mortality of vascular infection, and given the importance of targeted and prolonged antibiotic therapy, the diagnosis of C. burnetii infection is crucial to a successful therapeutic outcome. We report here a rare case of C. burnetii vascular graft infection. This unexpected diagnosis was based on broad-range PCR. Q fever is probably underdiagnosed since the diagnosis will be missed if it is not systematically looked for in patients with vascular graft infections of unknown etiology. The diagnosis of Q fever is most frequently made by serology. C. burnetii presents a variation of phase (phases I and II). Antiphase I IgG at titers of ≥1:800 by microimmunofluorescence are indicative of chronic Q fever [9,10]. The diagnosis has also be made using molecular methods including 16S rRNA PCR amplification plus sequencing or a specific C. burnetii PCR. To shorten the diagnostic delay, Fenollar et al. developed a nested-PCR assay with serum as a template which showed a sensitivity of 64% and a specificity of 100% [11]. In the present case, 16s rRNA PCR amplification plus sequencing was critical for diagnosis, since Q fever had not been suspected clinically. Broad-range PCR was also recently shown to be useful for the diagnosis of blood culture-negative infectious endocarditis, identifying an etiological agent in 35.5% of cases (11/31), including 3 cases due to C. burnetii [12]. The histology of Q fever endocarditis is non-specific and is characterized by fibrosis and calcifications, mild inflammation, and small or absent vegetations. Brouqui et al. observed a granulomatous inflammation in one third of cases with Q fever endocarditis, but well formed granuloma were not identified. In these cases, degenerative changes such as valvular calcifications or foreign body reaction prevented the establishment of a specific association between C. burnetii infection and granulomatous inflammation [13]. The presence of foamy macrophages, also observed in our case, may suggest a Coxiella infection. Definite histological diagnosis relies upon immunohistology, that may demonstrate the presence of C. burnetii within the cytoplasm of macrophages [13]. However, this diagnostic method is less sensitive than PCR [14]. Thus, immunohistology is unlikely to be a useful tool for the diagnosis of C. burnetii vascular graft infection. Cell culture is not widely available as a diagnostic technique as it requires trained technicians and a biosafety level 3 laboratory. Tetracyclines are the antibiotics of choice for acute Q fever [10]. The bactericidal activity of doxycycline is maximal at pH 6.6. However, the Coxiella-containing vacuole is acidic. Therefore, addition of chloroquine that acts as an alkalinizing agent of the vacuole is essential to improve the efficacy of doxycycline therapy [15]. Combination of doxycycline and hydroxychloroquine for at least 18 months is the recommended therapy for Q fever endocarditis [15]. This combined treatment is probably also indicated in case of C. burnetii vascular graft infection. Since serum doxycycline concentration is correlated with decrease in levels of phase I antibodies, it is recommended to adjust the dosage of doxycycline [7]. Patients are considered cured when phase I IgG antibodies decrease below 1:800, and IgA and IgM antibodies below 1:50 [16]. In this case, after 18 months of therapy, antiphase I serology remains strongly positive suggesting persistent infection most likely due to the fact that the infected prosthesis could only be partially removed. It also shows how difficult it is to eradicate C. burnetii and emphasizes the need for prolonged antibiotic treatment course. Serological follow-up will guide our decision regarding the treatment duration. Conclusion C. burnetii infection should be considered in patients with a vascular graft and unexplained low-grade fever, abdominal discomfort, laboratory signs of inflammation, and/or a history of environmental exposure. Broad-range PCR should be performed on surgical samples in patients with suspected infection of aneurysm or vascular graft. C. burnetii serology and/or specific C. burnetii PCR should also systematically be performed in cases of culture-negative vascular graft infection. Competing interests The author(s) declare that they have no competing interests. Authors' contributions LS, MF, DR, LVS, TC and GG were involved in patient care. LS wrote the first draft of the manuscript. AM did the histology and provided the images. All authors improved the manuscript and approved its final version. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements We thank Dr Philip E. Tarr for critical reading of the manuscript. ==== Refs Goëau-Brissonnière OA Coggia M Arterial prosthetic infections Waldvogel FA, Bisno AL, eds Infections Associated with Indwelling medical Devices 3rd ed Washington, DC: ASM Press 2000 127 144 Ellis ME Smith CC Moffat MA Chronic or fatal Q-fever infection: a review of 16 patients seen in North-East Scotland (1967-80) Q J Med 1983 52 54 66 6878620 Fournier PE Casalta JP Piquet P Tournigand P Branchereau A Raoult D Coxiella burnetii infection of aneurysms or vascular grafts: report of seven cases and review Clin Infect Dis 1998 26 116 121 9455519 Raoult D Tissot-Dupont H Foucault C Gouvernet J Fournier PE Bernit E Stein A Nesri M Harle JR Weiller PJ Q fever 1985-1998. Clinical and epidemiologic features of 1,383 infections Medicine (Baltimore) 2000 79 109 123 10771709 10.1097/00005792-200003000-00005 Georghiou GP Hirsch R Vidne BA Raanani E Coxiella burnetii infection of an aortic graft: surgical view and a word of caution Interactive Cardiovascular and Thoracic Surgery 2004 3 333 335 17670253 10.1016/j.icvts.2004.01.013 Goldenberger D Kunzli A Vogt P Zbinden R Altwegg M Molecular diagnosis of bacterial endocarditis by broad-range PCR amplification and direct sequencing J Clin Microbiol 1997 35 2733 2739 9350723 Rolain JM Mallet MN Raoult D Correlation between serum doxycycline concentrations and serologic evolution in patients with Coxiella burnetii endocarditis J Infect Dis 2003 188 1322 1325 14593588 10.1086/379082 Sessa C Vokrri L Porcu P Maurin M Stahl JP Magne JL Abdominal aortic aneurysm and Coxiella burnetii infection: report of three cases and review of the literature J Vasc Surg 2005 42 153 158 16012465 10.1016/j.jvs.2005.03.022 Rolain JM Lecam C Raoult D Simplified serological diagnosis of endocarditis due to Coxiella burnetii and Bartonella Clin Diagn Lab Immunol 2003 10 1147 1148 14607881 10.1128/CDLI.10.6.1147-1148.2003 Maurin M Raoult D Q fever Clin Microbiol Rev 1999 12 518 553 10515901 Fenollar F Fournier PE Raoult D Molecular detection of Coxiella burnetii in the sera of patients with Q fever endocarditis or vascular infection J Clin Microbiol 2004 42 4919 4924 15528674 10.1128/JCM.42.11.4919-4924.2004 Greub G Lepidi H Rovery C Casalta JP Habib G Collard F Fournier PE Raoult D Diagnosis of infectious endocarditis in patients undergoing valve surgery Am J Med 2005 118 230 238 15745720 10.1016/j.amjmed.2004.12.014 Brouqui P Dumler JS Raoult D Immunohistologic demonstration of Coxiella burnetii in the valves of patients with Q fever endocarditis Am J Med 1994 97 451 458 7977434 10.1016/0002-9343(94)90325-5 Lepidi H Houpikian P Liang Z Raoult D Cardiac valves in patients with Q fever endocarditis: microbiological, molecular, and histologic studies J Infect Dis 2003 187 1097 1106 12660924 10.1086/368219 Raoult D Houpikian P Tissot DH Riss JM rditi-Djiane J Brouqui P Treatment of Q fever endocarditis: comparison of 2 regimens containing doxycycline and ofloxacin or hydroxychloroquine Arch Intern Med 1999 159 167 173 9927100 10.1001/archinte.159.2.167 Dupont HT Thirion X Raoult D Q fever serology: cutoff determination for microimmunofluorescence Clin Diagn Lab Immunol 1994 1 189 196 7496944	16336642	PMC1325236	CC BY	2021-01-04 16:28:17	no	BMC Infect Dis. 2005 Dec 7; 5:109	utf-8	BMC Infect Dis	2,005	10.1186/1471-2334-5-109	oa_comm
==== Front Int Semin Surg OncolInternational seminars in surgical oncology : ISSO1477-7800BioMed Central London 1477-7800-2-281632115110.1186/1477-7800-2-28Case ReportTattoo pigment in an axillary lymph node simulating metastatic malignant melanoma Jack CM [email protected] A [email protected] H [email protected] Breast Unit and Pathology Department Mayday University Hospital, London Road, Croydon, CR7 7YE, Surrey UK2005 1 12 2005 2 28 28 28 8 2005 1 12 2005 Copyright © 2005 Jack et al; licensee BioMed Central Ltd.2005Jack 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. We report a case of axillary lymphadenopathy thirty years after a decorative tattoo clinically mimicking metastatic melanoma. The importance of relying on histological confirmation of metastatic disease before altering extent of surgery is discussed. The importance of recording presence of decorative tattoos is stressed. Tattoo PigmentLymphadenopathyMalignant Melanoma ==== Body Background The presence of lymphadenopathy requires further investigation. Often its presence is explained by a simple viral illness or trauma. Rarer causes are often made apparent by thorough history taking and examination. The need for a biopsy is controversial. We report a case where the answer may have been staring us in the face if we knew where to look. The fact that a tattoo causes lymphadenopathy is well known in the acute phase. This is thought to be due to local inflammation from the initial insult. However, to our knowledge there have been no reports of a palpable node after time delay this long. Case report A 54 year old man presented with a lump in the right axilla of six months duration. The lump was non tender and had not changed in size. He complained of weight loss of 5 kg over the past two months. He denied foreign travel, night sweats, recent injury, cough, or the presence of any other lumps. His past medical history was unremarkable. There was no family history of breast or bowel cancer. The lump was clinically palpable and measured 3 cm. It was firm, non tender, not attached to the skin or deep tissues and was consistent with a clinical diagnosis of axillary lymphadenopathy. The left axilla and supraclavicular fossae were normal. There was no skin lesion in the drainage area of the axilla. Examination of the breasts, chest and abdomen were unremarkable. Haematology, Biochemistry and Chest X-rays were normal. Ultrasound confirmed a benign appearing lymph node with a fatty centre. In view of the size and longstanding nature of the lymph node, an excision biopsy was performed. At surgery the node was firm, suspicious and black in colour. On histology, the specimen of the lymph node with attached fatty tissue measured 3 × 2 × 0.8 cms. Black discolouration was present on the cut surface. Microscopic examination of the routine haematoxylin and eosin sections of the lymph node showed preservation of architecture with follicular hyperplasia. Black carbon like pigment was seen lying within the macrophages and dispersed outside them in the sinuses. There was associated fibrosis. Multiple sections did not reveal any evidence of metastatic malignant melanoma. Immunohistochemical staining for S 100 protein and histochemical stain (Masson's Fontana) was done to further exclude that possibility. Retrospectively we noted the 30-year old tattoo that the patient had on his right arm. Conclusion Lymphadenopathy refers to nodes that are abnormal in size, consistency or number [1]. There are various classifications of lymphadenopathy, but a simple and clinically useful system is to classify lymphadenopathy as "generalized" if lymph nodes are enlarged in two or more non-contiguous areas or "localized" if only one area is involved. Localised lymphadenopathy of the axilla is suggestive of infections, Cat-scratch disease, Lymphoma, Breast cancer, Silicone implants, Brucellosis and Melanoma. The presence or otherwise of a tattoo may not be noted in history taking for lymphadenopathy [2]. Little information exists to suggest that a specific diagnosis can be based on node size. However, in one series of 213 adults with unexplained lymphadenopathy, no patient with a lymph node smaller than 1 cm2 had cancer, while cancer was present in 8 percent of those with nodes from 1 cm2 to 2.25 cm2 in size, and in 38 percent of those with nodes larger than 2.25 cm2 [3]. In children, lymph nodes larger than 2 cm in diameter (along with an abnormal chest radiograph and the absence of ear, nose and throat symptoms) were predictive of granulomatous diseases (i.e. tuberculosis, cat-scratch disease or sarcoidosis) or cancer (predominantly lymphomas) [4]. The fact that a tattoo causes lymphadenopathy is well known in the acute phase due to local inflammation and probably resolves spontaneously. The natural history of tattoo is well documented. The tattoo ink particles may range from 2–400 nm and are most commonly 40 nm. They are initially found within large phagosomes in the cytoplasm of keratinocytes, phagocytic cells including fibroblasts, macrophages and mast cells. The skin layers initially appear homogenised but at one month, the basement membrane is reforming and aggregates are present within basal cells. At 2–3 months and at 40 years, ink particles are only found in dermal fibroblasts surrounded by a network of connective tissue that entraps and immobilises the cell. The tattoo may appear blurred with time due to ink movement into the deep dermis. Eventually the tattoo ink appears in the regional lymph nodes. This is thought to be due to local inflammation from the initial insult. However, to our knowledge there have been no reports of a palpable node after time delay this long. The dye used in skin tattooing is carbon based. The movement of dye through the lymph channels forms the basis of sentinel node biopsy. Complications of lymph node biopsy are reported as scaring, blood loss, infection and more rarely nerve damage and lymphoedema. The question remains whether it was necessary to biopsy this lymph node or was the presence of the tattoo enough to give reason for the enlarged node. In this instance the co factor of the weight loss meant that leaving the node would not be reasonable. Anderson [5] and Moehrle [6] reported that tattoo pigments in Lymph nodes can mimic metastatic malignant melanoma, but do not comment on age of the decorative tattoo. Such pigmentation in patients with malignant melanoma can look metastatic and may prompt the surgeon to proceed to complete nodal dissection. Nodal dissection should be delayed till conclusive histological diagnosis is made [7]. Migration of the carbon pigment through the lymphatics is usually seen in the hilar lymph nodes draining the lungs. The main differential diagnosis in our case would be metastatic malignant melanoma. This was excluded by the careful examination of the H&E sections for tumour cells (Figure 1, 2) and employing special stains. Immunohistochemical staining for S 100 protein is a very sensitive marker for melanoma cells and a Masson's Fontana stain helps to differentiate melanin pigment from carbon pigment. Figure 1 Lymph node with preserved architecture and the pigment. H&E × 100 Figure 2 The dark granular carbon pigment located in the sinuses. H&E × 400 Sentinel lymph node biopsy is becoming more common in Melanoma and Breast cancer. History taking and examination should include presence, site, age and colour of decorative tattoos especially in the drainage areas to the axilla. History of removal of tattoos is also important as nodes may persist for several years. Raising awareness of this problem among surgeons and pathologists treating malignant melanoma is important. Investigation of axillary Lymphadenopathy should include tattoos in the drainage areas as a probable cause. Acknowledgements The authors would like to thanks Mr SR Ebbs (Consultant Surgeon) for his support and guidance. ==== Refs Goroll AH May LA Mulley AG Jr Primary care medicine: office evaluation and management of the adult patient 1987 2 Philadelphia: Lippincott Ferrer R Lymphadenopathy: Differential Diagnosis and Evaluation American Family Physician October 15 1998 Pangalis GA Vassilakopoulos TP Boussiotis VA Fessas P Clinical approach to Lymphadenopathy Semin Oncol 1993 20 570 82 8296196 Slap GB Brooks JS Schwartz JS When to perform biopsies of enlarged peripheral lymph nodes in young patients JAMA 1984 252 1321 6 6471252 10.1001/jama.252.10.1321 Anderson LL Cardone JS McCollough ML Grabski WJ Tattoo pigment mimicking malignant melanoma Dermatological Surgery 1996 22 92 4 10.1016/1076-0512(95)00336-3 Moehrle M Blaheta HJ Ruck P Tattoo pigment mimics positive sentinel lymph node in melanoma Dermatology 2001 203 342 344 11752827 10.1159/000051787 Friedman T Westreich M Mozes SN Tattoo Pigment in Lymph Nodes Mimicking Metastatic Malignant Melanoma Plast Reconstructive Surg 2003 111 2120 2122 10.1097/01.PRS.0000057101.95872.A1	16321151	PMC1325237	CC BY	2021-01-04 16:38:36	no	Int Semin Surg Oncol. 2005 Dec 1; 2:28	utf-8	Int Semin Surg Oncol	2,005	10.1186/1477-7800-2-28	oa_comm
==== Front Malar JMalaria Journal1475-2875BioMed Central London 1475-2875-4-581633667110.1186/1475-2875-4-58ResearchEpidemiology of forest malaria in central Vietnam: a large scale cross-sectional survey Erhart Annette [email protected] Ngo Duc [email protected] Ky Phan [email protected] Ta Thi [email protected] Overmeir Chantal [email protected] Niko [email protected] Valerie [email protected] Le Xuan [email protected] Le Khanh [email protected] Marc [email protected]'alessandro Umberto [email protected] Prince Leopold Institute of Tropical Medicine, Nationalestraat 155, 2000 Antwerp, Belgium2 National Institute for Malariology, Parasitology and Entomology, Luong The Vinh street, BC 10200 Tu Liem district, Hanoi, Vietnam3 Provincial Centre for Malariology, Parasitology and Entomology, 156 Ngo Gia Tu Street, Phan Rang city, Ninh Thuan province, Vietnam2005 8 12 2005 4 58 58 17 8 2005 8 12 2005 Copyright © 2005 Erhart et al; licensee BioMed Central Ltd.2005Erhart 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. In Vietnam, a large proportion of all malaria cases and deaths occurs in the central mountainous and forested part of the country. Indeed, forest malaria, despite intensive control activities, is still a major problem which raises several questions about its dynamics. A large-scale malaria morbidity survey to measure malaria endemicity and identify important risk factors was carried out in 43 villages situated in a forested area of Ninh Thuan province, south central Vietnam. Four thousand three hundred and six randomly selected individuals, aged 10–60 years, participated in the survey. Rag Lays (86%), traditionally living in the forest and practising "slash and burn" cultivation represented the most common ethnic group. The overall parasite rate was 13.3% (range [0–42.3] while Plasmodium falciparum seroprevalence was 25.5% (range [2.1–75.6]). Mapping of these two variables showed a patchy distribution, suggesting that risk factors other than remoteness and forest proximity modulated the human-vector interactions. This was confirmed by the results of the multivariate-adjusted analysis, showing that forest work was a significant risk factor for malaria infection, further increased by staying in the forest overnight (OR= 2.86; 95%CI [1.62; 5.07]). Rag Lays had a higher risk of malaria infection, which inversely related to education level and socio-economic status. Women were less at risk than men (OR = 0.71; 95%CI [0.59; 0.86]), a possible consequence of different behaviour. This study confirms that malaria endemicity is still relatively high in this area and that the dynamics of transmission is constantly modulated by the behaviour of both humans and vectors. A well-targeted intervention reducing the "vector/forest worker" interaction, based on long-lasting insecticidal material, could be appropriate in this environment. ==== Body Introduction Controlling malaria in forested areas remains a challenge in many parts of Asia and South America [1-7]. In Vietnam, forest malaria occurs in 16 provinces (out of 64) situated in the central part of the country (11 in the Central area, 4 in the western highlands and 1 in the south-eastern region). According to the figures reported by the National Malaria Control Program (NMCP), about half of the total malaria cases, more than 90% of the severe cases and almost 95% of malaria deaths occur in these 16 forested provinces [8,9]. In a previous community-based study [5], regular forest activity was a strong risk-factor for malaria infection and its population-attributable fraction was estimated at 53%. Workers, when staying in the forest overnight, do not usually sleep under insecticide-treated bed nets (ITN) and are therefore exposed to infection. Moreover, due to the behaviour of the main vector Anopheles dirus (early biting, exophagy and exophily), neither ITNs nor indoor spraying seem to be suitable control measures [10]. New interventions targeted to forest workers are urgently needed and should be tested in field trials [11-13]. A cluster randomized trial to test the protective efficacy of Long Lasting Insecticidal Hammocks (LLIH) in controlling forest malaria was launched in 2004 in collaboration with the NMCP and it is still ongoing. According to the expected impact of the intervention, estimated on the basis of previously collected epidemiology data [5], 20 clusters of about 1,000 inhabitants each were identified in Ninh Thuan province, one of the poorest and more endemic provinces, on the basis of a preliminary screening survey carried out in 43 villages. This large scale-survey allowed the analysis of the spatial and temporal distribution of the malaria infections over a large forested area and the confirmation of a previous risk-factor analysis carried out in a similar but much more limited setting [5] Materials and methods Study area and population The survey was carried out from November to December 2003 in Ninh Thuan province, located on the southern coast of Central Vietnam (Figure 1). Forty three villages corresponding to the forested and mountainous part (north-west) of the province and with the highest annual malaria morbidity and mortality figures according to the Provincial Malaria Station were selected. The population is distributed over 12 communes located in 4 districts (Bac Ai, Ninh Son, Ninh Phuoc and Ninh Hai) and is mainly inhabited by the Rag Lays, traditionally nomadic but recently settled in permanent villages. Most people are farmers, cultivating maize in forest fields or and rice around the villages, and collecting forest products. The dry season lasts from December to April, the rainy season from May to November with one of the lowest mean annual rainfalls in the country (<800 mm/y). The monthly mean temperature is 25–30°C, with an annual mean of 27.3°C. The mean relative humidity over the past 3 years has been 74.7%, ranging from 70 to 80% throughout the year. Malaria transmission is perennial with 2 peaks (May-June and October-November), the monthly incidence of malaria cases following closely the monthly rainfalls (Figure 2). The two main malaria vector species are An. dirus A and Anopheles minimus A [14]. Figure 1 Land-cover map Ninh Thuan province with parasite rate and malaria seroprevalence for the 43 study villages-Nov.2003 (see details, Table 1). Figure 2 Evolution of rainfall and monthly malaria incidence in Ninh Thuan province: 2001–2004 (Data from Center of Malariology, Parasitology and Entomology, Ninh Thuan province). Data collection Households were numbered and a full census carried out. A sample size of 150 individuals per cluster of 1,000 inhabitants was calculated at 5% significance level assuming a 5% parasite rate with 3.5% precision. Households were randomly chosen and within each of them, an individual aged 10 to 60 years was randomly selected (the latter age range was estimated to represent the active population most at risk of malaria [5]). Socio-demographic data, risk factors for malaria infection as well as current prevention methods were collected on pre-coded standardized questionnaires. At clinical examination, axillary temperature and spleen size were measured. Blood samples for parasitaemia (thick film) and P. falciparum antibodies (filter paper) were collected. Suspected malaria cases were treated with a 7-day course of artesunate according to the national guidelines; other common diseases were treated accordingly. Village coordinates were recorded using a Geographic Position System or GPS (eTrex Summit, GARMIN Corporation) and later reported on the land-cover digital map for the year 2000 provided by the Institute of Geography – Hanoi (adapted from land-cover map compiled by the Department of Forestry – FIPI). The original map with 22 land-cover classes was simplified into the 5 following categories: Natural forests – including rich forest, mixed forest, medium forest, wood and bamboo mixed forests, bamboo forest and clear forest; Plantations – including different types of plantations; Dry forest deciduous; Other – including rice fields, land tenure, habitat with garden, swamps or wetlands, and different types of bare land; Water – including natural river & lake. Elevation varied from 10 m (Phuoc Chien commune) to 323 m above the sea-level in Phuoc Binh. The numbers presented on the map refer to the village unique identifier. Parasite rates and seroprevalences for each village (Table 1) were reported onto the map with different coloured symbols corresponding to the following categories: 0–10%; 10–30%; = 30% (Figure 1). Table 1 Parasite rate and malaria seroprevalence of the 43 study villages (see map Figure 1) Village Code Village Commune District Parasite rate (%) Sero-prevalence (%) 1 Lap La Lam Son Ninh Son 3,7 6,1 2A Tam Ngan1 Lam Son Ninh Son 10,4 15,4 2B Tam Ngan2 Lam Son Ninh Son 10,4 15,4 3 Thon Gon Lam Son Ninh Son 1,2 6,9 4 Thon Do Ma Noi Ninh Son 30,8 21,2 5 Ha Zai Ma Noi Ninh Son 24,7 49,4 6 Thon U Manoi Ninh Son 13 7,7 7 Ya Rot Ma Noi Ninh Son 15,4 10,3 8 Ta Noi Ma Noi Ninh Son 26,7 22,7 9 Ro On Phuoc Ha Ninh Phuoc 1,5 15,2 10 Gia Phuoc Ha Ninh Phuoc 8,7 38,5 11 La A Phuoc Ha Ninh Phuoc 5,5 42,7 12 Tra No Phuoc Ha Ninh Phuoc 6,2 17,1 13 Ma Trai Phuoc Chien Ninh Hai 3,2 23,2 14 Tap La Phuoc Chien Ninh Hai 14,8 58,3 15 Dau Suoi A-B Phuoc Chien Ninh Hai 21,1 36,1 16 Dong Thong Phuoc Chien Ninh Hai 32,1 20,2 17 Bac Ray 2 Phuoc Binh Bac Ai 22,7 18,7 18 Bac Ray 1 Phuoc Binh Bac Ai 9,4 26,9 19 Bo Lang Phuoc Binh Bac Ai 41,7 26,8 20 Gia E Phuoc Binh Bac Ai 34,1 31,8 21 Hanh Rac 2 Phuoc Binh Bac Ai 42,3 25 22 Hanh Rac 1 Phuoc Binh Bac Ai 39,5 21,9 23 Suoi Ro Phuoc Chinh Bac Ai 9,7 4,8 24 Suoi Kho Phuoc Chinh Bac Ai 5,9 20,8 25 Nui Ray Phuoc Chinh Bac Ai 0 2,1 26 Ta Lu (1,2,3) Phuoc Dai Bac Ai 2,8 10,3 27 Ma Hoa Phuoc Dai Bac Ai 6,9 25 28 Ta Lot Phuoc Hoa Bac Ai 23,3 29,3 29 Cha Panh Phuoc Hoa Bac Ai 12,2 13,1 30 Ma Lam Phuoc Tan Bac Ai 6,7 75,6 31 Ma Ty Phuoc Tan Bac Ai 23,6 37,3 32 Da Trang Phuoc Tan Bac Ai 4,4 22,5 33 Cha Dung1 Phuoc Thang Bac Ai 36,6 62,4 34 Ma Ty Phuoc Thang Bac Ai 11,8 49 35 Ha La Ha Phuoc Thang Bac Ai 5,3 12,3 36 Ma Oai Phuoc Thang Bac Ai 15,1 36,2 37 Ma Ro Phuoc Thanh Bac Ai 16,7 27,8 38 Da Ba Cai Phuoc Thanh Bac Ai 6,7 23,3 39 Suoi Lo Phuoc Thanh Bac Ai 14 38 40 Manai Phuoc Thanh Bac Ai 26,7 21,3 41 Mazu Phuoc Thanh Bac Ai 9,6 35,2 42 Da Ban Phuoc Tien Bac Ai 35 19 43 Suoi Rua Phuoc Tien Bac Ai 14,3 59,2 Laboratory methods Thick films were stained with a 5% Giemsa solution for 30 minutes. Parasite densities were computed based on the number of asexual forms and gametocytes per 200 white blood cells (WBCs), assuming a mean WBC count of 8,000/μl. A slide was classified as negative if no Plasmodium asexual form was found after counting 1,000 WBCs. Filter papers (Whatman N°3, Kent, United Kingdom) were stored at -20°C, and Indirect Fluorescent Antibody Tests (IFAT) were carried out to determine the total immunoglobulin titres against P. falciparum as described previously [15]. P. falciparum antigen was prepared from in vitro cultures of an isolate originating from a patient in Southern Vietnam. Negative control serum was obtained by pooling the sera of five malaria-free individuals; positive control serum, by pooling the sera of 5 patients who had already suffered several malaria episodes. The serum dilutions ranged from 1:80 to 1:640. Both slide and IFAT reading were blinded and carried out at NIMPE (National Institute of Malariology, Parasitology and Entomology), Hanoi. Quality control was performed at ITM (Institute of Tropical Medicine), Antwerp, on all positive and 10% negative samples and all discrepant results were re-read and confirmed by a third technician. Case definition - Suspected malaria: patient with typical malaria symptoms without microscopic diagnosis; - Malaria infection: positive slide with Plasmodium asexual forms, regardless of symptoms; - Clinical malaria: fever (BT≥37.5°C) and/or history of fever in the past 48 hours with a positive blood slide; - Positive IFAT: titre ≥ 1/80. The immune response to an active malaria infection is species and strain specific, and lasts 6–9 months without additional exposure [16,17]. Usually titres above 1:20 are considered positive. However, we chose a more specific screening titre of 1:80. - Current & recent malaria infections: to perform the multivariate adjusted risk-factor analysis for malaria infections (Table 4) all cases with either a positive slide and/or a positive IFAT were combined. Table 4 Risk factor analysis for malaria infections (positive slide and/or positive IFAT (1/80)): uni- and multi-variate adjusted analysis using survey logistic regression (n = 1,434) Risk factors Prevalence (%) Cases (n/N) Unadjusted Multivariate Adjusted OR 95%CI OR 95%CI Total 33.3 1,434/4,306 Sex: - Male 35.8 659/1841 1 1 - Female 34.1 775/2465 0.82 [0.71; 0.96]° 0.71 [0.59; 0.86]° Age (y): - 10–20 30.6 293/959 1 - 20–30 33.5 501/1494 1.15 [0.95; 1.39] - 30–40 35.6 346/972 1.26 [1.03; 1.53]° - 40–50 36.0 194/539 1.28 [1.09; 1.56]° - - - 50-max 29.2 100/342 0.94 [0.67; 1.32] Ethnic group: - Kinh 13.2 30/227 0.26 [0.17; 0.43]°° 0.67 [0.29; 1.52] - Rag Lay 36.6 1355/3705 1 1 - Others 13.1 49/374 0.26 [0.12; 0.58]° 0.38 [0.22; 0.66]° Education: - No 40.6 798/1968 1 1 - Primary 30.1 556/1849 0.63 [0.50; 0.80]°° 0.68 [0.52; 0.88]° - Secondary + 16.4 80/489 0.29 [0.22; 0.38]°° 0.46 [0.34; 0.64]°° Bed-net availability: - No net 44.6 41/92 1.83 [1.05; 3.21]° - Untreated nets 40.9 9/22 1.58 [0.70; 3.54] - - > 2pers./net 34.3 948/2761 1.19 [0.98; 1.45] - 1–2 pers./net 30.5 436/1431 1 House structure: - Bamboo 40.5 883/2183 1 1 - Wooden 27.6 338/1224 0.56 [0.42; 0.76]°° 0.85 [0.68; 1.08] - Dried mud 29.7 70/236 0.62 [0.40; 0.97]° 0.62 [0.41; 0.94]° - Bricks 21.6 143/663 0.40 [0.30; 0.56]°° 0.81 [0.62; 1.07] Socio-economic status: - No radio, TV, moto 37.8 1037/2742 1 1 - Only radio 34.6 155/448 0.87 [0.65; 1.17] 0.97 [0.74; 1.27] - TV only 27.6 82/297 0.63 [0.38; 1.03]° 0.73 [0.48; 1.09] - TV + radio 20.7 34/164 0.43 [0.26; 0.71]° 0.60 [0.39; 0.93]° - Moto (+/-radio +/-TV) 19.2 126/655 0.39 [0.28; 0.55]°° 0.54 [0.37; 0.78]° Profession: - Other work 14.7 65/441 1 - Forest work 36.5 1245/3412 3.32 [1.39; 3.42]°° - - No (children, students, old) 27.4 124/453 2.18 [2.36; 4.68]° Forest work: - Other work 14.7 65/441 1 1 - Occasional (+child/retired) 25.9 258/993 2.03 [1.14; 3.18]° 1.48 [1.01; 2.17]° - Regular, no sleep 35.7 681/1906 3.22 [2.09; 4.94]°° 1.76 [1.05; 2.94]° - Regular + sleep 48.8 327/670 5.51 [3.42; 8.90] °° 2.86 [1.62; 5.07]° - Missing data 34.8 103/296 - °p < 0.05 °°p < 0.001 Statistical analysis Data were entered in Epi Info 6.04 (CDC, Atlanta; WHO, Geneva 1996) and analysed in STATA 8 software (Stata Corp.2003, College Station, TX). Malaria parasite rate and sero-prevalence were computed by village and a survey chi-square test ("svytab" command in STATA) was used to test for significant differences for these two variables between villages of each commune and between the communes themselves. A survey logistic regression ("svylogit" command in STATA) was carried out taking into account the survey characteristics, with villages as primary sampling units and communes as strata. Uni- and multivariate adjusted analysis for the risk of malaria infection during the rainy season was carried out. Therefore, current (positive slide) and recent (IFAT ≥ 1:80) malaria infections were combined in order to include most malaria infections during the previous six months up to the time of the survey. The primary exposure was forest work and this variable was coded as follows: 1) not working in forest (other job); 2) occasionally (less than weekly); 3) regularly (at least weekly), but not staying overnight; 4) regularly working in forest and sleeping there at various frequency. The category – no profession, i.e. children, students and retired people, was included in the category occasionally working in forest since they do so with their family. The Ethical Committee of ITM, Antwerp and that of NIMPE, Hanoi approved the study (Ethical clearance registration n°: OG 018). Informed consent was received from all village leaders and People Committees after explaining the study objectives and methodology. Selected individuals were informed and invited to be part of the survey some days in advance but were free to refuse. Other people not chosen but being sick at the time of the survey were examined and treated accordingly but were not included in the survey database. Results The study population (Table 2) Table 2 Baseline characteristics of the study population and forest workers Study population, n = 4,306 N (%) 95%CI Sex: - Male 1,841 (42.8) [40.7; 44.8] - Female 2,465 (57.2) [55.2; 59.3] Age categories: - 10–19 y 959 (22.3) [21.0; 23.6] - 20–29 y 1,494 (34.7) [33.4; 36.0] - 30–39 y 972 (22.6) [20.9; 24.3] - 40–49 y 539 (12.5) [11.4; 13.7] - >50 y 342 (7.9) [6.5; 9.4] Ethnic groups: - Rag Lay 3,705 (86.0) [83.1; 89.0] - Kinh 227 (5.3) [3.0; 7.5] - Ko'ho 349 (8.1) [3.9; 12.3] - Others (Cham, Ede, Chu,) 25 (0.6) [0.3; 0.9] Education level: - None 1,968 (45.7) [42.4; 49.0] - Primary school 1,849 (42.9) [40.0; 45.9] - Secondary school or university 489 (11.6) [9.4; 13.3] Bed-net coverage, median [range] 2.5 pers/net [0.2; 11] Bed-net categories: - No net 92 (2.1) [1.1; 3.2] - Untreated nets 22 (0.5) [0; 1.1] - >2 pers./net (treated) 2,761(64.1) [60.3; 68.0] - 1–2 pers./net ( " ) 1,431(33.2) [29.4; 37.1] House structure: - Thatched bamboo 2,183 (50.7) [45.3; 56.1] - Wooden boards 1,224 (28.4) [23.4; 33.4] - Dried mud 236 (5.5) [2.9; 8.0] - Bricks 663 (15.4) [12.7; 18.1] Socio economic level: - No radio, no TV, no motorbike 2,742 (63.7) [58.8; 68.5] - Only radio (no TV, no moto) 448 (10.4) [8.7; 12.1] - TV (+/-radio) 461 (10.7) [8.3; 13.1] - Motorbike (+/- radio or TV) 655 (15.2) [12.0; 18.4] Profession: - Forest work (farming & other) 3,412 (79.2) [76.1; 82.3] - Other (rice farmer, teacher, health staff...) 441 (10.2) [7.4; 13.1] - None (children, students, retired people) 453 (10.5) [9.1; 12.0] Forest workers, n = 3,412 Type of forest activity: - Farming 3,395 (99.5) [99.2; 99.8] - Exploitation forest products (fishing, hunting, etc...) 15 (0.4) [0.1; 0.7] - Other (cow breeding, guardian) 2 (0.06) [0; 0.1] Median number of days/month spent in forest: 27 range [0;30] Sleeping in the forest: - Yes 1,003 (29.4) [24.0; 34.8] - No 2409 (70.6) [65.2; 76.0] Median number of nights/month spent in forest(n = 1003): 12 range [1; 30] Sleeping place in the forest: - Outside 112 (11.2) [5.1; 17.2] - Plot huts 891 (88.8) [82.8; 94.9] Using hammocks when sleeping in forest: - Yes (hammock nets = 27) 85 (8.5) [4.8; 12.2] - No 918 (91.5) [87.8; 95.2] Using preventive an/or standby treatment when going to forest: - Yes 80 (2.3) [0.8; 3.9] - No 3332 (97.7) [96.1; 99.2] 4,306 (92%) of the 4,679 selected individuals participated to the survey with a sex ratio of 1.3 in favour of females. The Rag Lay ethnic group represented the vast majority of the study population (86%), and educational level was generally low since almost half of the population had never attended school. However, communication was easy since most people (91.7%) could speak Vietnamese. Bed-nets (usually double nets) were widely used (98% of households) with a median of 2.5 persons/net. Most of the nets had been treated with insecticide the previous year, but no indoor spraying had been done. The socio-economic status was very low since half of the population was living in houses made of thatched bamboo and more than 60% owned none of the three items asked for (radio, television, or motorbike). About 80% of the study population was defined as forest workers, i.e. their main source of income was from forest activities (farming, hunting, logging, etc...). Forest work characteristics (Table 2) Farming represented the main forest activity (99.5%), mostly maize and manioc, rice being generally cultivated around the villages. During the month prior to the survey, forest workers spent a median of 27 days in the forest. Most of them usually went to the forest with other family members, including children and elderly people (categorized in our study as "no profession"). About 30% of the forest workers stayed sometimes overnight in the forest (or forest fields), with a median of 12 nights during the month prior to survey. Important seasonal variations exist: April (preparation of fields before rains), July-August (first harvest) and for some villages November (second harvest) are the periods of highest forest activity. Few forest workers reported using hammocks (because they could not afford them) and hammock-nets were even rarer. Other preventive measures, such as standby and/or preventive malaria treatment, were rarely used (2.3%). Malariometric indices (Table 3) Table 3 Malariometric indices Indicators (N = 4,306) n (%) [95%CI] - Prevalence of fever cases 668 (15.5) [12.5; 18.5] - Fever cases attributable to malaria 125 (18.7) [13.7; 23.7] - Spleen rate (n = 301) 301 (7.0) [4.3; 9.7] - Median age if enlarged spleen [range] 30 y [11; 60] - Mean parasite rate 571 (13.3) [10.6; 15.9] - Species distribution (n = 571): P. falciparum 276 (48.3) [41.3; 55.3] P. vivax 261 (45.7) [39.0; 52.4] P. malariae 3 (0.5) [0; 1.1] Mixed infections 31 (5.4) [3.2; 7.6] - Parasite density/μl (geometric mean): P. falciparum 178.3 [141.2; 225.2] P. vivax 51.5 [44.0; 60.3] - Proportion of asymptomatic infections 423 (74.1) [67.8; 80.4] - P. falciparum sero-prevalence (N = 4,253) 1,085 (25.5) [20.7; 30.3] Six hundred sixty eight (15.5%) individuals had fever at the time of the survey, 125 (18.7%) of them being attributable to malaria. The spleen rate was 7% (all but two individuals had a Hackett Class 1) and the median age among people with enlarged spleen was 30 years (range [11–60]). The mean parasite rate (PR) was 13.3%, ranging from 0 to 42.3% across the 43 villages, with a large majority of asymptomatic infections (74.1%). P. falciparum and Plasmodium vivax were equally represented, though the mean parasite density was significantly lower for the latter, and only three patients were found positive with Plasmodium malariae. About 5% (31) of the infections were mixed, mostly P. falciparum with P.vivax (28), one patient had the three species, two had Pf+Pm and one P.v +P.m. P. falciparum seroprevalence (SPV) was generally high, 25.5%, ranging from 2.1 up to 75.6% across villages. Mapping of PR and SPV (Figure 1 & Table 1) All 43 villages were located in places classified as 'dry forests deciduous' or 'others', situated at various distance from the natural forest. Within each commune the respective distances to the different types of vegetations are similar between the villages, since the latter are usually close to each other, i.e.12 km (except in Phuoc Binh commune where the 6 villages stretch over almost 10 km and in Phuoc Tan, over 5 km). The mapping of PR and SPV shows an extremely patchy distribution without geographic clustering or trends per commune, per direction or elevation. Significant differences for both PR and SPV were found between the 12 communes, and between villages in almost all (10/12) communes (F (design based) p < 0.05)). Four villages (n°1-3-23-25) had both low PR and SPV, below 10%. These villages are situated in the plain, near the main district road leading to Dalat or to the district town in Phuoc Dai commune. Nui Ray was the only village among the 43 with a PR at zero. At the other extreme, four villages had a SPV above 50% (n°14-30-33-43): all of them were remote with difficult access and near to the forest. In these villages, the PR was relatively low (ranging from 6.7% in Ma Lam to 36.6% in Cha Dung), reflecting higher transmission rates during the past six months, at the time of higher forest activity. Similarly, in Phuoc Ha and in Phuoc Dai communes (villages n° 9–12 and 26, 27), all six villages had a PR<10% but a SPV>10% (from 15.2 to 42.7%). However, remoteness was not the only determining factor for high PR and SPV since some remote villages such as Tanoi (n°8) in Manoi commune and Bac Ray1 (n°18) in Phuoc Binh commune, did not have higher PR or SPV than the other villages within the same commune. Overall, more than half of the villages (24) had either a PR>10% and/or a sero-prevalence>25%. Usually the SPV was higher than the PR, though some villages had a SPV lower than the PR. This could be explained either by the higher current exposure or by the higher prevalence of P. vivax (as in Bo Lang or Da Ban (data not shown)). Risk factors for malaria infection (Table 4) All current and recent malaria infections were combined to perform a risk-factor analysis. A total of 1,434 infections (33.3%) were included. Among risk factors identified by uni-variate analysis, Rag Lay ethnicity, education, socio-economic status and regular forest work had the strongest effect on the odds of malaria infection. After adjusting for the effect of potential confounders, regularly working and sleeping in the forest was a strong risk-factor with the odds almost three times higher than that of people not working in the forest. Even occasional forest work was a significant risk-factor for malaria (Adjusted OR = 1.48; 95%CI [1.01; 2.17]). The malaria risk was significantly lower in people living in houses made of dried mud, while wooden or brick houses were no more protective, after adjusting for socio-economic status. There was a decreasing trend for the odds of malaria infection with increasing socio-economic status, the highest category had about 50% lower malaria risk compared to the poorest one. Discussion This large-scale cross-sectional survey confirms that Ninh Thuan province remains one of the highest endemic malaria provinces in Vietnam [8,9], especially in its north-western forested and hilly area where the study was carried out. The mean PR and SPV were high, respectively 13.3 and 25.5%, with upper limits at 42.3% and 75.1% respectively. Forest malaria is extensively described in different Asian [2,4,6,18] and South American countries [3,7,19], where it remains a real challenge for control programs [1]. A previous study carried out in a village located in a forested area of the neighbouring province Binh Thuan, also reported high malaria incidence (11/100 person-years) and SPV (20–25%)[5]. However, unlike the previous study, the SPV was high also in people not working in the forest, indicating that village transmission might be important. Indeed, many surveyed villages were surrounded by the forest or not far from it so that the man/vector contact could be high for all their inhabitants. The mapping of the PR and SPV by village showed great heterogeneity between villages located in a comparable ecological setting. This suggests that factors other than environmental ones intervene in modulating the human/vector interaction. The population in this study was mainly Rag Lays, an ethnic minority that used to live and work in the forest practising "slash and burn" cultivation, previously reported as a risk-factor for malaria [6]. Moreover, Rag Lays used to be nomads in forested mountains. Although attempts to settle them down in permanent villages are ongoing, the forest where they use to collect various products (bamboo, nuts, berries, game animals and birds) remains their natural environment. This way of life exposes them to a higher risk of malaria infection than other ethnic groups. Even after adjusting for the effect of forest work, ethnic group, age, and education, women were still significantly less at risk of malaria, confirming the results of a previous study [5]. Compared to men, women usually remain well-covered, particularly when working outside, thus reducing the risk of exposure to mosquito bites. Women go to sleep earlier (with the children under a net, if available) while men like to sit outside around a fire. Considering that An. dirus bites early and is a highly anthropophylic and exophagic vector, men are obviously more exposed than women. Education was an important protective factor and there was a decreasing trend of malaria risk with increasing level of education, even after adjusting for socio-economic status and forest work. This confirms its importance in preventing diseases in general and malaria in particular. Hence, the Vietnamese national programme for poverty alleviation [13,20] which, among other strategies, provides full subsidies to ethnic minorities for their children education (primary and secondary school), might significantly contribute to reduce not only poverty but also diseases like malaria. Independently of the socio-economic status, education is a crucial factor for adherence to malaria prevention measures, and even more crucial is the education of mothers to give prompt and effective malaria treatment to small children [21-23]. Numerous studies from SEA and Africa have extensively reported poor housing conditions as a significant risk factor for malaria infections due to greater human-vector contacts, and this is especially true for endophagic vectors like Anopheles gambiae and Anopheles culicifacies [24-27]. In this study, even if the main vector is exophagic and exophilic, the house structure might still have an impact on man-vector contact since bamboo houses are usually on stilts with wide openings in the floor as well as in the walls between bamboo canes, decreasing the outdoor/indoor difference and allowing An dirus to bite indoors [10]. Even if wooden and bricks houses showed a trend for a protective effect, only dried mud houses had a significant and strong protective effect and this might be due to their architecture as windows are extremely small and might reduce the human/vector contact. One advantage of this study is the dissociation of housing effect from the socio-economic status, an important point in the setting, where even rich people still prefer to live in the Rag Lay traditional stilt bamboo houses. Even though the socio-economic classification used in this study was not the result of an exhaustive in-depth analysis of family resources and levels of income, it gives nevertheless a rough idea of their purchase power, independently of the housing conditions. Thus, even if residual confounding might have occurred, it is unlikely that it would greatly change the results, since there is a clear trend for decreasing odds of malaria infections with increasing level of socio-economic status. Malaria is and remains a poverty-related disease [28-30]. The number of current malaria infections identified during the survey (571) contrasts sharply with those collected by the provincial health information system (499 cases in the whole province during the 2 months prior to the survey). The fact that 75% of the infections detected during the survey were asymptomatic raises the question of the development of immunity in this ethnic group regularly exposed to malaria infections, and/or under-reporting of cases in the HIS due to either self-treatment or treatment by traditional healers or private practitioners. Conclusion This study confirms the high malaria morbidity in this area of Ninh Thuan province and the reality represented by the complex problem of "Forest malaria". The PR and SPV mapping with its patchy distribution suggested a highly variable dynamic of transmission in time and space, since various combinations of values for both variables could be observed between and within communes. This variability is driven by the constant interplay of human behavioural factors, as showed by the risk factor analysis, and vector behaviour. Forest malaria is a complex phenomenon which results in an epidemiological mosaic [1] and more in-depth entomological and anthropological studies are necessary to understand the specific mechanisms of forest malaria transmission in Central Vietnam. In the framework of the LLIH trial entomological and anthropological studies are ongoing and will contribute to broaden our knowledge on how malaria is maintained and subsequently could be controlled in these settings. Conflict of interest statement The authors hereby certify that no conflict of interest of any kind occurred in the framework of this study. Financial support The present study was supported by the UBS-Optimus Foundation, Zurich (Switzerland), and the Belgium Cooperation, in the framework of a four-year project investigating the efficacy of Long Lasting Insecticidal Hammocks in preventing forest malaria. This project is the continuation of an over 10-year long bilateral collaboration between the National Institute for Malariology, Parasitology and Entomology, Hanoi and the Prince Leopold Institute of Tropical Medicine, in Antwerp. Authors' contributions Dr. Thang (Assistant Project Coordinator in NIMPE) and Dr. Ky (Director of the Malaria Center in Ninh Thuan province) organized the survey, coordinated and supervised the field work (3 study teams worked simultaneously for 1 month) and the data entry. Valérie Obsomer (agronomist expert on GIS) elaborated the Ninh Thuan map, and Niko Speybroeck, (agronomist and statistician) contributed to the data analysis. Dr. Hung (Vietnamese Project Co-ordinator) helped in the study design and choice of study villages to be included in the LLIH trial, and reviewed the manuscript. Pr. Thuan (Director of NIMPE) contributed to the choice of the study sites, and contributed to review the manuscript. Pr. U.D'Alessandro (Principal investigator of the LLIH trial) contributed to all steps, from the elaboration of the project up to the final review of this article. Pr. Marc Coosemans, entomologist, reviewed the manuscript. Dr. Ta Thi Tinh and Mrs Chantal Van Overmeir processed all the filter papers with the IFAT and performed the quality controls of all blood samples results. Dr A. Erhart (Assistant Project Co-ordinator in ITM) contributed to the study design, implementation, supervision, the data analysis and wrote the manuscript. Acknowledgements We would like to specifically thank all the staff of the Center for Malariology, Parasitology & Entomology (CMPE) of Ninh Thuan province who was involved in the preparation and implementation of the survey, as well as all the health staff from each Districts and Communes Health Centres. A special acknowledgement should be addressed here to the provincial administrative and health authorities who provided a full support to our study as well as future investigations on the LLIH efficacy. Finally, all study participants are warmly acknowledged for their motivation towards our project and for their warm and unforgettable hospitality during the whole survey. We would also like to thank the UBS-Optimus Foundation, Zurich (Switzerland) for its financial support and more particularly Dr Susanna Hausman-Muela for her encouragement and support. ==== Refs Sharma VP and Kondrashin AV Forest malaria in Southeast Asia: Proceedings of an informal consultative meeting WHO/MRC 1991 New Delhi Chaveepojnkamjorn W Pichainarong N Malaria infection among the migrant population along the Thai-Myanmar border area Southeast Asian J Trop Med Public Health 2005 35 48 52 15272744 Danis-Lozano R Rodriguez MH Gonzalez-Ceron L Hernandez-Avila M Risk factors for Plasmodium vivax infection in the Lacandon forest, southern Mexico Epidemiol Infect 1999 122 461 469 10459651 10.1017/S0950268899002319 Das NG Talukdar PK S.C. D Epidemiological and entomological aspects of malaria in forest-fringed villages of Sonitpur district, Assam J Vect Borne Dis 2004 41 5 9 Erhart A Thang ND Toi LV Hung LX Tuy TQ Cong LD Speybroeck N Coosemans M D'Alessandro U Forest malaria in Vietnam: a challenge for control Am J Trop Med Hyg 2004 70 110 118 14993619 Klinkenberg E van der Hoek W Amerasinghe FP A malaria risk analysis in an irrigated area in Sri Lanka Acta Tropica 2004 89 215 225 14732243 10.1016/j.actatropica.2003.08.007 Povoa MM Conn JE Schlichting CD Amaral JCOF Segura MNO Da Silva ANM Dos Santos CCB Lacerda RNL De souza RTL Galiza D Santa Rosa EP Wirtz RA Malaria vectors, epidemiology, and the re-emergence of Anopheles darlingi in Belem, Para, Brazil J Med Entomol 2003 40 379 386 14680100 NIMPE NIMPE Annual Report of the National Malaria Control Program in Vietnam: 2002 2003 Hanoi, Vietnam NIMPE NIMPE Annual Report of the National Malaria Control Program in Vietnam: 2003 2004 Hanoi,Vietnam Trung HD Van-Bortel W Sochanta T Keokenchanh K Briët OJT Coosemans M Behavioural heterogeneity of Anopheles species in ecologically different localities in Southeast Asia: a challenge for vector control. Trop Med Int Health 2005 10 1 12 15655007 10.1111/j.1365-3156.2004.01378.x TDR TDR fifteenth programme report progress 1999-2000 2000 WHO WHO Expert committee on malaria. Twentieth report 2002 WHO Combatting communicable diseases. Focus 2: Malaria, other vectorborne and parasitic diseases 2003 Trung HD Van Bortel W Sochanta T Keokenchanh K Quang NT Cong LD Coosemans M Malaria transmission and major malaria vectors in different geographical areas of Southeast Asia Trop Med Int Health 2004 9 230 237 15040560 10.1046/j.1365-3156.2003.01179.x Demedts P Wery M Indirect immunofluorescence using differentiating antigens in detection of imported malaria Ann Soc Belg Med Trop 1985 65 Suppl 2 89 96 3907545 Cavanagh DR Elhassan IM Roper C Robinson VJ Giha H Holder AA Hviid L Theander TG Arnot DE McBride JS A longitudinal study of type-specific antibody responses to Plasmodium falciparum merozoite surface protein-1 in an area of unstable malaria in Sudan Journal of Immunology 1998 161 347 359 Marsh K Warrell DA and Gilles HM Immunology of malaria Essential Malariology 2002 11 4th New York, Arnold 252 267 Lansang MA Belizario VY Bustos MD Saul A Aguirre A Risk factors for infection with malaria in a low endemic community in Bataan, the Philippines Acta Trop 1997 63 257 265 9088438 10.1016/S0001-706X(96)00625-0 Gadelha P From "forest malaria" to "bromeliad malaria": a case-study of scientific controversy and malaria control Parassitologia 1994 36 175 195 7898955 GTZ Vietnamese program for poverty alleviation 2004 Amuge B Wabwire-Mangen F Puta C Pariyo GW Bakyaita N Staedke S Kamya M Okui O Health-seeking behavior for malaria among child and adult headed households in Rakai district, Uganda. Afr Health Sci 2004 4 119 124 15477191 Safeukui-Noubissi I Ranque S Poudiougou B Keita M Traore A Diakite M Cisse MB Keita MM Dessein A Doumbo OK Risk factors for severe malaria in Bamako, Mali: a matched case-control study. Microbes Infect 2004 6 572 578 15158191 10.1016/j.micinf.2004.02.007 Cropley L The effect of health education interventions on child malaria treatment-seeking practices among mothers in rural refugee villages in Belize, Central America. Health Promotion International 2004 19 445 452 15520038 10.1093/heapro/dah406 Gamagemendis AC Carter R Mendis C Dezoysa APK Herath PRJ Mendis KN Clustering of Malaria Infections Within An Endemic Population - Risk of Malaria Associated with the Type of Housing Construction Am J Trop Med Hyg 1991 45 77 85 1867350 Gunawardena DM Wickremasinghe AR Muthuwatta L Weerasingha S Rajakaruna J Senanayaka T Kotta PK Attanayake N Carter R Mendis KN : Malaria risk factors in an endemic region of Sri Lanka, and the impact and cost implications of risk factor-based interventions : Am J Trop Med Hyg 1998 58 533 42 9598437 Konradsen F Amerasinghe P van der Hoek W Amerasinghe F Perera D Piyaratne M Strong association between house characteristics and malaria vectors in Sri Lanka Am J Trop Med Hyg 2003 68 177 181 12641408 Lindsay SW Jawara M Paine K Pinder M Walraven GEL Emerson PM Changes in house design reduce exposure to malaria mosquitoes Trop Med Int Health 2003 8 512 517 12791056 10.1046/j.1365-3156.2003.01059.x Barat LM Palmer N Basu S Worrall E Hanson K Mills A Do malaria control interventions reach the poor? A view through the equity lens Am J Trop Med Hyg 2004 71 174 178 15331835 Malaney P Spielman A Sachs J The malaria gap Am J Trop Med Hyg 2004 71 141 146 15331830 Sachs J Malaney P The economic and social burden of malaria Nature 2002 415 680 685 11832956 10.1038/415680a	16336671	PMC1325238	CC BY	2021-01-04 16:24:14	no	Malar J. 2005 Dec 8; 4:58	utf-8	Malar J	2,005	10.1186/1475-2875-4-58	oa_comm
==== Front Aust New Zealand Health PolicyAustralia and New Zealand Health Policy1743-8462BioMed Central London 1743-8462-2-321635429610.1186/1743-8462-2-32ResearchImplementation failures in the use of two New Zealand laws to control the tobacco industry: 1989–2005 Thomson George [email protected] Nick [email protected] Department of Public Health, Wellington School of Medicine and Health Sciences, University of Otago, Box 7343 Wellington South, New Zealand2 Department of Public Health, Wellington School of Medicine and Health Sciences, University of Otago, Box 7343 Wellington South, New Zealand2005 14 12 2005 2 32 32 3 10 2005 14 12 2005 Copyright © 2005 Thomson and Wilson; licensee BioMed Central Ltd.2005Thomson and Wilson; 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 reviewed the implementation of New Zealand laws in relation to the activities of the tobacco industry and their allies. Material for two brief case studies was obtained from correspondence with official agencies, official information requests, internet searches (tobacco industry documents and official government sites), and interviews with 12 key informants. Results The first case study identified four occasions over a period of 14 years where New Zealand Government agencies appeared to fail to enforce consumer protection law, although apparent breaches by the tobacco industry and their allies had occurred in relation to statements on the relative safety of secondhand smoke. The second case study examined responses to a legal requirement for the tobacco industry to provide information on tobacco additives. There was failure to enforce the law, and a failure of the political process for at least 13 years to clarify and strengthen the law. Relevant factors in both these cases of 'policy slippage' appear to have been financial and opportunity costs of taking legal action, political difficulties and the fragmented nature of government structures. Conclusion Considered together, these case studies suggest the need for governments to: (i) make better use of national consumer laws (with proper monitoring and enforcement) in relation to tobacco; and (ii) to strengthen international law and resources around tobacco-related consumer protection. A number of options for achieving these aims are available to governments. ==== Body Introduction A number of authors argue that specific attention by governments and advocates to the policy implementation process is needed in tobacco control, to avoid the risk of 'policy slippage' [1-3]. Policy slippage is the gap between the intent of policy in legislation or government directives, and the implementation of such policy. This paper investigates the way two national laws relevant to the activities of the tobacco industry were implemented. One law concerns deceptive statements to potential consumers and the public, the second, the disclosure of tobacco additives. The setting is New Zealand, which has some components of a comprehensive tobacco control programme [4,5]. Background The New Zealand tobacco control programme has over the last 15 years introduced tobacco promotion controls, smoking bans in nearly all indoor work and public places, a free Quitline cessation service, and access to heavily subsidised nicotine replacement treatment. There have also been active non-government tobacco control groups since the early 1980s [5,6]. Although per capita consumption has declined over the last decade, until 2004 smoking prevalence has stayed at around 25–26% [[7] p.9]. This may be due to high uptake of smoking by youth and the lack of fiscal incentives to support quitting (ie, there has been only two tax rises on cigarettes beyond the rate of inflation since 1991 [8]). The programme operates in the context of the activity in New Zealand of three major international tobacco companies – British American Tobacco, Philip Morris, and Imperial Tobacco. The activities of such companies have been part of the impetus for the recent formation (and ratification by most countries) of the Framework Convention on Tobacco Control, the first treaty under the World Health Organization's treaty making powers. Methods Brief case studies were used to review the implementation of the two laws. The first case, around the Fair Trading Act 1986, involved the collection of material from: correspondence with official agencies, requests to the New Zealand Government (using the Official Information Act process), the opportunistic receipt of an official memorandum sent to a Member of Parliament, a search of the New Zealand Commerce Commission's website for media releases issued from 1994 to September 2002 [9], and a search of the Factiva media website for 'New Zealand major newspapers' for the period September 2001 – August 2002, using the phase 'Commerce Commission' and the word 'court'. The period covered for the case study was from 1989 to 2004. The second case, concerning the provision for tobacco additives disclosure in the Smoke-free Environments Act 1990, involved website searches for both tobacco industry and New Zealand Government correspondence. The tobacco industry document websites tobaccodocuments.org, Legacy Tobacco Documents Library, and British American Tobacco Documents Archive were searched for material on the disclosure of tobacco additives in New Zealand, using the searchwords 'zealand' 'additives' 'ingredients' 'Wellington' and 'Department/Ministry of Health' and then by following names revealed by the search. The period covered was from 1990 to 2004. For both cases, material from interviews with four government officials and ex-officials, two academics, four politicians and two lawyers was used to assess the validity of the documentary evidence, to provide leads for inquiries, and to provide policy and political context. Results Context Two aspects of the context for the case studies were: a lack of priority for tobacco control by the dominant party in government during 1991–1999 (the conservative National Party) [10], and the wish by all governments during the period covered to be seen as 'business friendly' [[11] p.40; [12-14]]. Government officials and ex-officials who were interviewed stated that during the 1990s they were given very clear signals that tobacco was not a priority area for the government. These signals in turn resulted in internal Ministry of Health decisions on the allocation of resources. Secondhand smoke (SHS) is estimated to cause over 300 deaths per year in New Zealand [15] making it the most important environmental health hazard. Accurate knowledge about the extent of SHS harm is important, as it affects smoking behaviour [16,17]. In addition, the tobacco industry has demonstrated the importance of obscuring public knowledge of SHS harms by their investment in deception about it [18-22]. In 2000, the Minister of Health wrote to the Hospitality Association of New Zealand (HANZ), noting that 'It is extremely important to treat what is said about SHS ... with due caution. ... Continuing statements by the tobacco industry, in particular, that the science is not in on SHS ... have [been] shown to be untrue ...' (King A. [Letter to Bruce Robertson of the Hospitality Association of New Zealand from the Minister of Health]. Wellington: 25th October 2000). Tobacco additives are important because of their role in tobacco product design to increase sales, and because of their potential role in addiction [23-27]. Public knowledge about the additives is important to help ensure informed consumers [28-30]. Case study 1: The tobacco industry and the Fair Trading Act 1986 The Fair Trading Act states that: 'No person shall, in trade, engage in conduct that is liable to mislead the public as to the nature, ... characteristics, [or] suitability for a purpose, of goods' [[31] s.10]. The Commerce Commission has stated that its job is to enforce 'legislation that ... prohibits misleading and deceptive conduct by traders' [32]. The Commission has suggested that, on the basis of court decisions, businesses must meet a fairly high standard to comply with the Fair Trading Act. The standard of business behaviour expected under the Act includes protecting: 'the gullible, [those] of less than average intelligence or poorly educated.' Generally, the question of intent is not relevant, 'rather the issue is whether their actions did or could deceive or mislead'. The Commission has suggested that the Act also applied to 'conduct likely to mislead or deceive', (emphasis added) [33]. Over the period 1989–2003, we found four episodes (in 1989, 2000, 2001 and 2002) where the Government had an opportunity to apply this law to the statements of tobacco companies and their allies about SHS. In 1989, the advocacy group ASH NZ wrote to the Commerce Commission about advertisements by the Tobacco Institute of New Zealand, which had stated that 'science has not established that other people's cigarette smoking causes diseases in non-smokers'. The Commission replied that it was 'extremely difficult to accurately gauge the effect of the Institute's advertising campaign and therefore whether the campaign has or is likely to have been misleading and deceptive.' The Commission also noted that 'the cost of any legal action would be considerable' [34]. In 2000, a new Minister of Consumer Affairs asked her officials about the use of the Fair Trading Act for legal action against the tobacco industry. She then reported that such 'action may not be possible under this Act' [35]. A request to the Ministry of Consumer Affairs under the Official Information Act for the advice given to the Minister was declined (Manch K. [Letter from Ministry of Consumer Affairs to George Thomson]. Wellington: August 23, 2001). In July 2001, we sent information to the Commerce Commission about the public statements of British American Tobacco (BAT) New Zealand. The material included comments from 1998 and 1999 by BAT officials about the IARC (International Agency for Research on Cancer) SHS study, that: 'The study confirms a view that the industry had long held that while smoke in the air may annoy some non-smokers, passive smoke is not a lung cancer risk' [36]. and more generally that: 'The overwhelming majority of [independent] studies found no overall meaningful increase in risk for those married to a smoker' [37]. These statements misrepresented the IARC findings [38]. Statements to the media in 2001 by HANZ were also given to the Commission, on the basis that the statements could be misleading and deceptive about the dangers of using HANZ members' premises where there was SHS. The statements included: ' A seven year study by the World Health Organisation found no links between passive smoking and health risks' [39] and '... science has not established a link between passive smoking and cancer' [40]. The statements by BAT and HANZ about SHS were forwarded because they appeared to meet four criteria that could help the Commission make a legal case – a pattern of activity, clear deception, sufficiently recent date, and large potential consequences to public health. The statements appeared to deny or obscure the harm caused by SHS, in a manner that could be described as misleading about tobacco products sold, or the safety of services provided by HANZ members. The Commission replied that although the material 'appears to be a breach of the Fair Trading Act, we will not be investigating it in more depth at this stage' and that the Commission targeted 'issues and trading practices that have the greatest potential detriment to consumers' (Gibson D. [Letter to George Thomson from the Commerce Commission (FTWN 49513)]. Wellington: Commerce Commission; August 30, 2001). A spokesperson was reported as saying that the Commission 'had decided to refer the complaint to the Ministry of Health and would take no further action'. This was because 'it is dealt with better under their legislation and they have the staff with the knowledge and expertise' [41]. In reply to the information about statements about SHS that appeared to contravene the Fair Trading Act, the Ministry of Health pointed out that the: 'Public Health Directorate lacks the staff and financial resources that would be needed to investigate the examples you presented in detail' but 'in general the Ministry supports the interpretations you present' (ie, that BAT and HANZ appeared to have contravened the Act's sections 9 and 13) (Matheson D. Letter to George Thomson from the Deputy Director General, Public Health Directorate (PP70-15-2). Wellington: New Zealand Ministry of Health; July 30, 2002). However, because the Ministry considered that the statements 'were not made in a trading context' and because of resource reasons, the Ministry decided that 'it would not be profitable' to pursue the matter under the Fair Trading Act (ibid). In May 2002, a politician from a minor party outside the government provided information to the Commerce Commission about deceptive practices by the industry. The information included two reported statements from the media during 2001. One, by a BAT official was that 'we have gone through the international evidence on secondhand smoke and there is a pattern of research [indicating] that it is not a serious health issue. Nothing is risk-free in this world. There is a small risk to young people and babies ...' [41]. The Director of the Commission's Fair Trading section wrote of this statement 'reasonable members of the public are unlikely to be misled into believing that this statement suggests that the death of adults is not a serious health risk, or that it denies that adults die from second hand smoke .... The statement represents an opinion and is unlikely to breach the Act' (Battell D. Memorandum to Fair Trading Committee: Tobacco companies – referral from Sue Kedgley MP. Wellington: Commerce Commission; September 11, 2002). The second reported statement sent to the Commission was that the science on SHS showed that 'if there was a risk, and there may be a risk, it's not a large one'. Commenting on this in September 2002, the Director of the Commission's Fair Trading section wrote that it was 'more an expression of opinion than fact. .... The people buying cigarettes are unlikely to be misled by comments on the harm or otherwise of passive smoking' [42]. To put the actions and statements of the Commission into context, we examined the legal actions it did initiate under the Fair Trading Act during the year to August 2002. In that time the Commission prepared or undertook at least twenty legal actions under the Act about deceptive statements. The issues were all relatively minor in terms of health risks to consumers, and included the claim of a fish and chip shop that it used cholesterol-free oil, eg. [43-46]. The Commission took at least eight other cases to court on competition matters, including two Court of Appeal and one Privy Council case [47-49]. The Commission also decided that it was justified in investing considerable resources in the control of the competitive aspects of the tobacco industry, by lengthily contesting a tobacco company merger [50]. Over the period 1994–2004, the most serious breach of the Fair Trading Act (in terms of mortality) where the Commission has responded, was on children's cots. These were implicated in the deaths of 22 children between 1985 and 1994 [51] (ie, around two deaths per year). In contrast, the total national death toll from SHS in year 2000 alone was estimated at 325 [15]. In September 2005, the BAT New Zealand website continued to cast doubt on the conclusions of the IARC SHS study, and stated: '... we don't believe that [SHS] has been shown to cause chronic disease, such as lung cancer, cardiovascular disease or chronic obstructive pulmonary disease, in adult non-smokers. ....' [52]. Case study 2: The non-disclosure of tobacco additives by brand In August 1990, the New Zealand Smoke-free Environments (SFE) Act was passed by Parliament. Section 35 of the Act required tobacco companies to submit an annual return showing the weight of all additives used in each tobacco product. In early October 1990 the SFE Regulations were passed by an Order in Council, just before an election changed the government. These regulations, amongst other things, set out the way that Section 35 was to be carried out. The regulation stipulated that companies give the weight of every additive for every product and then the quantity of 'each brand and each brand variant sold'. Thus, while 'product' could be considered to be a cigarette brand or brand variant, the regulations could also give the impression that a 'product' was a generic type, such as cigarettes, cigars, or roll-your-own tobacco. From then to the present, tobacco companies operating in New Zealand have endeavoured to continue to both keep the additives used in New Zealand secret from the public, and to not disclose to government the additives for particular brands. In late 1990, a Philip Morris official in Australia reported to the USA on some tactics he might use to get the additives disclosure law changed, and concluded 'short of sinking N.Z. if you have any other ideas, could you let me know' [53]. In 1991 the Government decided to insist on the disclosure of additives by 'each brand and brand variants' [54,55]. Negotiations dragged on, and by December 1992, the Department of Health (DoH) was accepting a 'temporary' compromise position where they accepted information only for all brands, rather than for particular brands [56]. This position was valued by the tobacco companies, and Philip Morris wrote to the Tobacco Institute of New Zealand in 1993 to emphasise that they thought the DoH: 'under the current regulations, could impose brand-by-brand disclosure .... The DoH could become irritated with the industry and simply impose a brand-by-brand disclosure' [57]. A 1993 note by Tony Andrade of the tobacco industry law firm of Shook, Hardy & Bacon also stated that the: 'ingredients law in New Zealand would require complete disclosure of individual ingredients and amounts by brand if strictly enforced' [58]. A Rothmans (NZ) official warned a RJ Reynolds official that for the 1994 return: 'there is a real possibility that the Department of Health will insist on a return in the correct form, that is, identifying the additives actually present rather than listing over 2000 possible additives. They may decide ... that we will be prosecuted if we file an incorrect return again. We will dispute this, but we would much prefer that it does not come to this. ... We have argued with them for over three years and their patience together with industry credibility must be close to breaking point' [59]. The government had three options in relation to implementing the SFE regulations: to compromise in some way; to insist on information by each brand (and prosecute if necessary); or to make clearer regulations that put the industry's obligations beyond dispute. There was little will by the political party in power until 1996 to confront the tobacco industry [10] and the compromise position was continued. In March 1994, the New South Wales Cancer Council in Australia released the list of additives given to the New Zealand Government by the tobacco companies. This list had been obtained by a request by ASH New Zealand, under the New Zealand Official Information Act [60]. This release, along with the ongoing risk of further New Zealand Government action on disclosures, appears to have resulted in at least two international level meetings of the tobacco companies involved [61,62]. After the election in late 1996, the political balance was slightly altered. The new Associate Minister with responsibility for tobacco control, Neil Kirton, was a member of the new minority party (New Zealand First) in the ruling coalition. After the four-year period of the compromise which did not insist on disclosure by brand, in August 1997 the Ministry of Health wrote to the tobacco companies, requiring information on additives for each brand [63]. However, that month Kirton was dismissed as Associate Minister, and successive Ministers responsible for tobacco control did not pursue the matter as a priority. An official described the problem from the bureaucracy's point of view: 'It's basically an issue of resourcing. ... Any regulation of the tobacco industry is very confrontational. It requires a lot of servicing in terms of the kind of official information that is requested and the kind of expectations around consultation.' Finally in 2003, the new amendments to the Smoke-free Environments Act required the testing and disclosure of each separate brands and brand variant, 'as the regulations may require'. In September 2005, these regulations were apparently being developed but their contents were unknown. Discussion Main findings and interpretation The two cases illustrate the sub-optimal implementation of existing laws that could have aided the control of a product that is a major cause of premature death and health inequalities in New Zealand. Possible reasons for this level of implementation of legislation relevant to tobacco companies include: the financial and opportunity costs of implementation, the political difficulties involved in confronting an aggressive industry, and the fragmented nature of government. We considered that the Ministry of Health's statement that the statements on SHS 'were not made in a trading context' as untenable, as the context for the sale of tobacco include the health claims made for the products. The two cases illustrate the importance of the costs of implementation when governments are faced with very large transnational companies, known to be tenacious litigators willing to use every aspect of the legal process. They also illustrate the need for substantial investment in an aggressive comprehensive tobacco control programme, to ensure that progress is made in spite of difficulties or hesitancies in the enforcement of relevant laws. When the Fair Trading Act issues arose, the relevant government agencies appeared to first think about their budgets and staff resources. In 1989 and 2001, the costs of legal proceedings and the staff resources needed were factors explicitly mentioned by the Commerce Commission. In contrast, the Commission prosecuted a number of other (often small) businesses for minor matters, and were willing to take tobacco and other large companies to court on competition matters. This may indicate a greater focus within the Commission on market regulation compared to the health of consumers. The Ministry of Health stated explicitly that it lacked 'the staff and financial resources that would be needed to investigate the examples you presented in detail' (Matheson D. Letter to George Thomson from the Deputy Director General, Public Health Directorate (PP70-15-2). Wellington: New Zealand Ministry of Health; July 30, 2002) let alone to prosecute tobacco companies on this issue. With the additives issue, officials knew that any attempt to enforce the law would take scarce staff time and budgets just for consultation and Official Information Act request processing, even before any legal expenses. The fear of high legal costs is justified by the experience of governments in Canada [64], the USA [65,66], the European Union [67] and elsewhere, where tobacco companies use all procedural and appeal avenues in court actions. The intended effect is partly to discourage future litigation against them. Internationally, government agencies generally find that effective investment in long-term population health issues can be squeezed out by short-term issues on political agendas [68-70]. While the New Zealand Ministry of Health has a large operational budget, any legal costs in an action such as this would come from the funds of the Public Health Directorate. This Directorate controls less than 3% of the Ministry budget, much of which has been devoted to politically sensitive issues in recent years (Matheson D. Letter to George Thomson from the Deputy Director General, Public Health Directorate. Wellington: New Zealand Ministry of Health; 8 July, 2002). In comparison with resources for immunisation and cancer screening, the use of health resources for legal expenses was politically difficult. The fragmented focus of government also contributed to unwillingness to take advantage of the legislative opportunities. While the Commerce Commission was willing to take tobacco companies to court on monopoly issues, it considered that aspects of the Fair Trading Act with health implications were 'dealt with better' by the Ministry of Health [41]. The statement of the Commission staff, that they focused on 'issues and trading practices that have the greatest potential detriment to consumers', may indicate that that they had not realised (or chose to ignore) the health risks from secondhand smoke, and the negative effects on health of continued statements that downplayed those risks. The Ministry, while it had some focus on smokers, was less able to focus on the tobacco industry as a major upstream threat to public health. In New Zealand and elsewhere, there has generally been pressure to invest the health sector budget in the treatment of disease, and action on individual risk behaviours [71], rather than invest in 'non-health' avenues such as legal action against tobacco companies. The cases indicate that for both deceptive statements about SHS and for the non-disclosure of additives in New Zealand, tobacco companies have been able to assume that the risk of legal action by government was low. Much of the usefulness of law is diminished if those businesses that governments are trying to control are not convinced that government legal action will occur or will be effective [72-74]. The New Zealand Government failed to convince tobacco industry investors and managers that government would consistently use the law to seriously challenge industry activities that affected public health in a substantial way. The investment in legal resources for tobacco control can be contrasted with the actions of the New Zealand Government in another arena. During the first years of its life from 1993, the government agency Pharmac consistently and successfully undertook multi-million dollar legal actions against the pharmaceutical industry. Some of the credit for the success could be ascribed to the clear focus of the agency in improving the efficiency of pharmaceutical purchasing on behalf of the State, and to its relatively independent decision making structure [75,76]. However, the impetus for the structure and policies of Pharmac was the rapidly increasing cost of pharmaceuticals to the publicly funded health system. By contrast, the net adverse consequences from the dominance by transnational tobacco companies of the local market were seen by government to generally impact on smokers, rather than on government. The advice from Treasury (disputed by some health advocates) has been that the tax revenue from tobacco sales heavily outweighed the direct costs to government from tobacco use. Therefore, governments may have been under the impression that there was little cost incentive to take a 'Pharmac' approach to the tobacco industry. The situation in New Zealand on misleading statements by the tobacco industry and its allies on SHS, despite the explicit statute law available to address the problem, has been in contrast to the Australian experience. There, a consumer organisation (largely funded by government) in the early 1990s took the local Tobacco Institute to court on the issue of false information about SHS issues. The court required that the industry could not describe SHS as not being shown to be unsafe [77]. The comparative level of action appears to argue for government agencies that are sufficiently funded, and are mandated to act for consumers (including smokers). In 2005, the Australian Competition and Consumer Commission 'obtained court enforceable undertakings from Imperial Tobacco Australia Limited' to remove misleading descriptors from its products [78]. The much greater practice in Australia compared to New Zealand, of making the tobacco industry accountable in court for their statements and actions, eg, [79-82], may be part of the explanation for the much greater decline in smoking prevalence there [83]. This could be due to the effects of publicity on smoking rates [84]. We have presented the case elsewhere that there may also be a relationship between a government focus on the tobacco industry, and more effective tobacco control [85]. The issue of additives disclosure illustrates some of the consequences of the contrasting national and international perspectives of governments and tobacco companies. The New Zealand Government saw the disclosure as a relatively minor matter, not requiring a high priority. In contrast, the companies are able and willing to focus international resources for long periods in New Zealand [86] and in other countries such as Australia [24], Thailand [87] and the Netherlands [88]. Limitations of these case studies These case studies were not exhaustive, but rather intended to illuminate aspects of implementation of tobacco control policy. Interviews with other key informants, and access to further documentation could potentially have provided more in-depth evidence and context. Nevertheless, we consider it likely that the evidence gathered around these case studies covers the major dimensions of the issues involved. Policy implications To more adequately address the type of problems identified in this article, and to reduce the risk of policy slippage, the following options could be considered by many governments: 1) Increasing central government capacity to work together across national boundaries to take effective and coordinated action in relation to tobacco companies [89]. This includes continuing the development of the Framework Convention on Tobacco Control, so as to strengthen international consumer protection law against tobacco-related marketing and misleading claims. 2) Provision of support for an international consumer protection law resource centre in a United Nations organisation (eg, one for health related consumer protection in the World Health Organization). 3) Ensuring explicit responsibility for the implementation of national consumer laws by particular government agencies, and provision of funding to ensure that the laws are adequately monitored and enforced. 4) Explicitly acknowledging the role of the tobacco industry as a factor in tobacco-related harm, and acting to minimise their influence. Options 2 and 3 are likely to be particularly cost-effective, as they can also assist in dealing with other major and extremely costly threats to public health (eg, the control of obesogenic foods and alcohol) [90,91]. Non-governmental organisations can advocate for all these responses, and use the media to highlight deficiencies in existing consumer protection laws. The cases suggest immediate actions. In New Zealand, court action is needed to clearly establish the relevance of the Fair Trading Act to misleading statements that affect the perceptions of at least some of the public about SHS. The writing of the new regulations, on the disclosure of tobacco product constituents and design, needs to allow for considerable powers by government to address any attempts by the tobacco industry to find loopholes in legislation designed to protect the public interest. Conclusion Existing New Zealand laws on misleading statements about secondhand smoke, and about the provision to government on tobacco additives, appear to have been insufficiently enforced. The health impact of secondhand smoke, and the importance of information for consumers about dangerous products suggest that such enforcement is necessary. Competing interests Both authors have undertaken contract work for tobacco control related non-government agencies and NW has undertaken contract work in tobacco control for the New Zealand Ministry of Health. Authors' contributions GT gathered the great majority of the data and conceived the article. Both authors designed the study, analysed the data and wrote and approved the article. Acknowledgements The authors thank those interviewees and others who provided information for these case studies. Two anonymous referees provided very helpful comments. This work was assisted by the National Institutes of Health (USA) grant #1 R01 CA87110-01A1 to the University of Sydney. Part of the work was funded by ASH New Zealand. ==== Refs Bialous S Glantz S Arizona's tobacco control initiative illustrates the need for continuing oversight by tobacco control advocates Tobacco Control 1999 8 141 51 10478397 Balbach ED Traynor MP Glantz SA The implementation of California's tobacco tax initiative: the critical role of outsider strategies in protecting Proposition 99 J Health Polit Policy Law 2000 25 689 715 10979517 10.1215/03616878-25-4-689 Jacobson P Wasserman J The implementation and enforcement of tobacco control laws: policy implications for activists and the industry J Health Polit Policy Law 1999 24 567 98 10386327 Laugesen M Swinburn B New Zealand's tobacco control programme 1985–1998 Tob Control 2000 9 155 62 10841851 10.1136/tc.9.2.155 Thomson G Wilson N Davis P, Ashton T Tobacco control policy Health and public policy in New Zealand 2001 Melbourne: Oxford University Press Studlar DT The Political Dynamics of Tobacco Control in Australia and New Zealand: Explaining Policy Problems, Instruments, and Patterns of Adoption Australian Journal of Political Science 2005 40 255 274 Ministry of Health Tobacco facts 2003: Public Health Intelligence Occasional Report No20 2003 Wellington: Ministry of Health Wilson N Thomson G Tobacco tax as a health protecting policy: a brief review of the New Zealand evidence N Z Med J 2005 118 U1403 15843832 Commerce Commission Media Releases Wellington: Commerce Commission Thomson G Wilson N Lost in the smoke: tobacco control in New Zealand during the 1990s NZ Med J 2000 113 122 3 Menz G A Model Strategy for Small States to Cope and Survive in a Globalised World Economy? An Analysis of the "New Zealand Way" European Consortium for Political Research (ECPR): Workshop 26: National Models and Transnational Structures – Globalisation and Public Policy: 26 – 31 March 1999 Mannheim, Germany. Accessed February 7, 2005 McCully M Forest Owners Association Speech at International Forest Industry Safety Conference Auckland: Scoop News Agency; October 5,1999 Accessed February 7, 2005 Shipley J Address To National Party Conference Scoop Press Agency; 21 August 2000 Accessed February 7, 2005 Asia Pacific Bulletin New Zealand's Election Keeps the Door Open to Business 69 Asia Pacific Bulletin. August 2, 2002 Accessed February 7, 2005 Woodward A Laugesen M How many deaths are caused by second hand smoke? Tobacco Control 2001 10 385 388 10.1136/tc.10.4.383 Blackburn C Spencer N Bonas S Effect of strategies to reduce exposure of infants to environmental tobacco smoke in the home: cross sectional survey Br Med J 2003 327 257 61 12896936 Gilpin E White M Farkas A Home smoking restrictions: which smokers have them and how they are associated with smoking behavior Nicotine Tob Res 1999 1 153 62 11072396 10.1080/14622299050011261 Diethelm PA Rielle J-C McKee M The whole truth and nothing but the truth? The research that Philip Morris did not want you to see Lancet 2005 365 86 92 15993237 10.1016/S0140-6736(05)66474-4 Trotter L Chapman S "Conclusions about exposure to ETS and health that will be unhelpful to us": how the tobacco industry attempted to delay and discredit the 1997 Australian National Health and Medical Research Council report on passive smoking Tob Control 2003 12 iii102 6 14645955 Samet J Burke T Turning science into junk: the tobacco industry and passive smoking Am J Public Health 2001 91 1742 4 11684591 Barnes D Bero L Why review articles on the health effects of passive smoking reach different conclusions JAMA 1998 279 1566 1570 9605902 10.1001/jama.279.19.1566 Barnes D Hanauer P Slade J Environmental tobaccosmoke. The Brown and Williamson documents JAMA 1995 274 248 53 7609234 10.1001/jama.274.3.248 Tuffs A German scientists demand a ban on tobacco additives Br Med J 2005 330 690 15790628 Chapman S "Keep a low profile": pesticide residue, additives, and freon use in Australian tobacco manufacturing Tob Control 2003 12 iii45 53 14645948 Wayne GF Connolly GN How cigarette design can affect youth initiation into smoking: Camel cigarettes 1983–93 Tob Control 2002 11 I32 9 11893812 Connolly G Wayne G Lymperis D How cigarette additives are used to mask environmental tobacco smoke Tobacco Control 2000 9 283 91 10982572 10.1136/tc.9.3.283 Vagg R Chapman S Nicotine analogues: a review of tobacco industry research interests Addiction 2005 100 701 12 15847628 10.1111/j.1360-0443.2005.01014.x Bates C McNeill A Jarvis M The future of tobaccoproduct regulation and labelling in Europe: implications for the forthcoming European Union directive Tobacco Control 1999 8 225 35 10478414 Chapman S Liberman J Ensuring smokers are adequately informed: reflections on consumer rights, manufacturer responsibilities, and policy implications Tob Control 2005 14 ii8 13 16046703 10.1136/tc.2005.012591 May H Wigand J The right to choose: Why governments should compel the tobacco industry to disclose their ingredients Essays in Philosophy 2005 6 Fair Trading Act New Zealand Government 1986 Commerce Commission What we do 2005 Wellington: Commerce Commission Accessed February 1, 2005 Commerce Commission The Fair Trading Act Wellington: Commerce Commission Accessed August 8, 2001 ASH NZ Commerce Commission back down Ashcan 1989 6 Bunkle P Address to ASH annual meeting Auckland: New Zealand Government speech archive; August 28, 2000 Accessed February 7, 2005 Graham C Now no reason for smoke bans in public places Tobacco Times 1 Graham C Double standards cloud the 'passive smoking' argument Tobacco Times 1998 3 Ong E Glantz S Tobacco industry efforts subverting International Agency for Research on Cancer's second-hand smoke study Lancet 2000 355 1253 59 10770318 10.1016/S0140-6736(00)02098-5 Evening Post Bar-goes the target in smoking studies Evening Post Wellington 4 April 30, 2001 Paltridge A Passive smoking victims may sue Evening Post Wellington 2 May 5th, 2001 Paltridge A Smoking claims a health issue, commission says Evening Post Wellington 27 August 1, 2001 Battell D Memorandum to Fair Trading Committee: Tobacco companies – referral from Sue Kedgley MP Wellington: Commerce Commission September 11, 2002 Commerce Commission Commerce Commission considers prosecution (media release 2002/2) Wellington: Commerce Commission; January 10, 2002 Accessed February 7, 2005 Commerce Commission Plumbing World fined $10,000 for misleading advertising (Media Release 2002/26) Wellington: Commerce Commission; August 22, 2002 Accessed February 7, 2005 Commerce Commission Auckland business fined $5,000 for Asian-made KIWI MADE product (Media Release 2002/16) Wellington: Commerce Commission; July 29, 2002 Accessed February 7, 2005 New Zealand Press Association Fish and chip man pleads ignorance Dominion Wellington 7 May 21, 2002 Dominion Commission loses appeal to Privy Council Dominion Wellington 10 20 February 2002 National Business Review Court of Appeal begins hearing claim by Commerce Commission National Business Review Auckland 30 November 2001 Richardson P Gault J Blanchard J Carter Holt Harvey Products Group Ltd – vs – The Commerce Commission Wellington: New Zealand Court of Appeal 5 November 2001 Chapple I Tobacco case needs work: High Court New Zealand Herald Auckland C5 February 16, 2002 Commerce Commission Commerce Commission will visit stores to check cots (Media release 2001/47) Wellington: Commerce Commission; May 9, 2001 Accessed February 7, 2005 British American Tobacco New Zealand Environmental tobacco smoke Auckland: British American Tobacco New Zealand Accessed September 10, 2005 Francis P New Zealand [Letter to Mary Pottorff, Philip Morris International] Philip Morris (Australia); November 19, 1990 Bates No. 2023246579. Accessed August 11, 2004 Upton S [Letter from the Minister of Health to Michael Thompson of the Tobacco Institute of New Zealand] Wellington: New Zealand government; February 12, 1991 Bates Nos: 2023894281-4282. Accessed January 7, 2005 Upton S [Letter from the Minister of Health to MJ Thompson of the Tobacco Institute of New Zealand] Wellington: New Zealand government; June 21, 1991 Bates Nos: 2023246459-6467. Accessed January 7, 2005 Talbot P [Letter to Trevor Smith, Managing Director of Rothmans New Zealand] Wellington: Department of Health; December 10, 1992 Bates Nos: 2023894279-4280. Accessed September 7, 2005 McGrath M New Zealand additives returns [Letter to Michael Thompson of the Tobacco Institute of New Zealand] Melbourne: Philip Morris Australia; February 5, 1993 Bates Nos: 2064815424-5425. Accessed January 7, 2005 Andrade A Update on active jurisdictions concerning ingredients issues [Memo to Matthew Winokur and Anne Oknoniewski] London: Shook, Hardy & Bacon; February 10, 1993 Bates Nos: 2051827255-7258. Accessed January 7, 2005 Lorrigan P [Fax to D Cooper Rees of RJ Reynolds Tobacco Company] Auckland: Rothmans New Zealand; November 11, 1993 Bates No: 515747823. Accessed January 10, 2005 Associated Press Cigarette additives Associated Press; March 30, 1994 Bates No. 2505444392. Accessed January 10, 2005 Lorrigan P Minutes of New Zealand Additives meeting held at Denham Place 11 July 1994 Rothmans New Zealand; July 20, 1994 Bates Nos: 2076248467-8472. Accessed January 10, 2005 Lorrigan P Ingredients meeting [letter to Dr JD Heck of Lorillard Tobacco Company] Auckland: Rothmans New Zealand; October 13, 1994 Bates No: 88026433. Accessed January 10, 2005 Poutasi K Smoke-free Environments Act 1990 – Section 35: Tobacco returns [letter to Samantha Kraftt of Chapman Trip Sheffield Young] Wellington: Ministry of Health; August 18, 1997 Bates Nos: 517667119-7121. Accessed January 10, 2005 Spurgeon D Canadian antismoking law is watered down BMJ 2005 331 474 16141145 10.1136/bmj.331.7515.474-c Civil Division U.S Department of Justice Litigation Against Tobacco Companies Washington DC: U.S. Government; October 4, 2005 Accessed November 29, 2005 Nixon M Mahmoud L Glantz S Tobacco industry litigation to deter local public health ordinances: the industry usually loses in court Tob Control 2004 13 65 73 14985600 10.1136/tc.2003.004176 Watson R European Court adviser rules against British American Tobacco and Imperial Bmj 2002 325 616 12269308 10.1136/bmj.325.7365.616/c Newacheck PW Benjamin AE Intergenerational equity and public spending: The United States should embrace a new doctrine of fairness to ensure that vulnerable populations are not forced to compete for resources Health Aff (Millwood) 2004 23 142 6 15371378 10.1377/hlthaff.23.5.142 Gloyd S Suarez Torres J Mercer MA Immunization campaigns and political agendas: retrospective from Ecuador and El Salvador Int J Health Serv 2003 33 113 28 12641267 10.2190/25BV-JW2P-22CN-WNRR Jan S A perspective on the analysis of credible commitment and myopia in health sector decision making Health Policy 2003 63 269 78 12595126 10.1016/S0168-8510(02)00119-7 Woodward A Looking upstream: A user's guide to disease prevention New Zealand Family Practice 2004 31 212 3 Anderson FR Environmental liability dangers Corporate Board 1993 14 6 10 Gruner RS Brown LM Organizational justice: Recognizing and rewarding the good citizen corporation Journal of Corporation Law 1996 21 731 765 Kaltenheuser S Schmiergeld Across the Board 1998 35 36 42 PHARMAC Briefing paper for the Minister of Health 1999 Wellington: PHARMAC Accessed January 20, 2003 Braae R McNee W Moore D Managing pharmaceutical expenditure while increasing access. The pharmaceutical management agency (PHARMAC) experience Pharmacoeconomics 1999 6 649 60 10724792 Chapman S Woodward S Australian court decision on passive smoking upheld on appeal Br Med J 1993 306 120 2 8435609 Australian Competition and Consumer Commission ACCC resolves 'light' and 'mild' cigarette issue with BAT and Philip Morris Canberra: Australian Competition and Consumer Commission; 12th May 2005 Accessed September 20, 2005 Liberman J Loff B Australian court rules against tobacco company in lung-cancer case Lancet 2002 359 1586 12047979 10.1016/S0140-6736(02)08538-0 Liberman J The shredding of BAT's defence: McCabe v British American Tobacco Australia Tob Control 2002 11 271 4 12198281 10.1136/tc.11.3.271 Stewart BW Semmler PC Sharp v Port Kembla RSL Club: establishing causation of laryngeal cancer by environmental tobacco smoke Med J Aust 2002 176 113 6 11936307 Woodward SD Winstanley MH Lung cancer and passive smoking at work: the Carroll case Med J Aust 1990 153 682 4 2174094 Durkin S Germain D Letcher T Smoking prevalence and consumption in Victoria: key findings from the 1998–2003 population surveys 2005 Melbourne: Cancer Control Research Institute: The Cancer Council Victoria. CBRC Research Paper Series No. 12 Laugesen M Meads C Advertising, price, income and publicity effects on weekly cigarette sales in New Zealand supermarkets British Journal of Addiction 1991 86 83 9 2009401 Thomson G Wilson N Directly eroding tobacco industry power as a tobacco control strategy NZ Med J 2005 118 Additives Reporting – New Zealand [Internal industry memo to tobacco companies operating in New Zealand] October 29, 1997 Bates Nos: 521588370-8371. Accessed January 10, 2005 MacKenzie R Collin J Sriwongcharoen K "If we canjust stall' new unfriendly legislations, the scoreboard is already in our favour": transnational tobacco companies and ingredients disclosure in Thailand Tob Control 2004 13 ii79 87 15564225 10.1136/tc.2004.009233 Associated Press Philip Morris Sues Dutch Government August 28, 2003 Accessed September 7, 2005 Wipfli H Stillman F Tamplin S Achieving the Framework Convention on Tobacco Control's potential by investing in national capacity Tob Control 2004 13 433 7 15564631 10.1136/tc.2003.006106 Swinburn BA The obesity epidemic in Australia: can public health interventions work? Asia Pac J Clin Nutr 2003 12 S7 15023592 Toomey TL Wagenaar AC Policy options for prevention: the case of alcohol J Public Health Policy 1999 20 192 213 10408171	16354296	PMC1325239	CC BY	2021-01-04 16:25:23	no	Aust New Zealand Health Policy. 2005 Dec 14; 2:32	utf-8	Aust New Zealand Health Policy	2,005	10.1186/1743-8462-2-32	oa_comm
==== Front AIDS Res TherAIDS Research and Therapy1742-6405BioMed Central London 1742-6405-2-111632422110.1186/1742-6405-2-11Short ReportHIV in semen: Still more to be learned Vernazza Pietro L [email protected] Division of Infectious Diseases and Hospital Epidemiology, Deparmtent of Medicine, Cantonal Hospital, St. Gallen Switzerland2005 3 12 2005 2 11 11 7 11 2005 3 12 2005 Copyright © 2005 Vernazza; licensee BioMed Central Ltd.2005Vernazza; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ==== Body In 1983, during the earliest days of AIDS research, Deborah Anderson and her colleagues in Boston, Massachusetts hypothesized that AIDS was transmitted by virally-infected "Trojan horse leukocytes" in semen [1]. This prediction has been supported by numerous studies over the past two decades, although many questions remain concerning HIV infection of the male genital tract. In this issue of AIDS Research and Therapy, the Anderson group presents an important research tool to help address some of the critical unanswered questions in this area [2]. A series of salient studies have shaped current concepts on HIV-1 in semen. In 1984, Ho et al. described retroviral particles and infected cells in the semen of a homosexual man with AIDS [3]. Shortly thereafter Stewart et al. reported infection of four out of eight women following artificial insemination with semen from one seroconverting individual [4], leading to a mandatory semen quarantine requirement and HIV testing of semen donors in Assisted Reproduction clinics. At that time, the detection of HIV, which was still termed HTLV-III, was not routinely feasible from blood, let alone from semen. Since then, technological advances have enabled the detection and quantitation of HIV-1 RNA and proviral DNA and greatly improved our understanding of the dynamics of HIV-1 in semen and sexual transmission risks. HIV-infected white blood cells have been detected throughout the male genital tract, and in preejaculatory fluid and semen from HIV+men [5,8]; the weight of evidence suggests that sperm are not infectious [9], leading to the successful development of sperm wash procedures to reduce the risk of HIV transmission from HIV-infected men to uninfected partners through assisted reproduction techniques [10]. A combination of epidemiological and clinical research studies have determined a relationship between HIV-1 RNA viral load in semen and the risk of sexual transmission. The most important factors associated with increased HIV viral loads in semen and risk of sexual transmission are: HIV-viremia and coinfections with other sexually transmitted pathogens [11,12]. HAART dramatically suppresses HIV-1 RNA viral loads in blood and semen, but HIV-1 proviral DNA can persist in semen WBCs for months after the initiation of HAART [13]. Data from other studies showing discordantly higher levels of HIV in semen than blood in some individuals support this finding. In addition, molecular sequencing studies indicate that the male genital tract is a compartment, like the central nervous system, in which HIV-1 replication and divergent evolution can occur under the influence of local factors [14]. Several clinically important questions remain: 1) Is HIV-1 primarily sexually transmitted by infected cells, cell-free virus or both? 2) What is the origin of cell-free and cell-associated HIV-1 in semen? 3) Are men on HAART with undetectable peripheral viral loads capable of sexually transmitting drug-resistant HIV-1? Episomal HIV-1 c-DNA, a by-product of HIV-1 infection, is currently used in clinical trials as a marker of residual viral replication and potential evolution of drug resistance mutations in viral reservoir sites in individuals on HAART [15]. Such a marker would be useful for identifying sites of HIV-1 replication in the male genital tract, and for monitoring cryptic HIV-1 infection in the genital tract of men on antiretroviral therapy. The only reported study that measured episomal HIV-1 c-DNA in blood and semen of men before and after initiation of HAART failed to detect HIV episomal 2-LTR cDNA in semen [16]. The method used to recover HIV-infected cells from semen in this study – separation of seminal WBC on Ficoll gradients – likely decreased the sensitivity of HIV episomal c-DNA detection because infected macrophages and a proportion of infected T-cells are lost through this approach. The paper by Xu et al. used a direct lysis technique optimizing recovery of DNA from HIV-infected cells in semen. Using this approach, combined with quantitative PCR and DNA sequencing, the investigators show that episomal 2-LTR cDNA is detectable in semen from a subset of men with other evidence of seminal HIV-1 infection. The marker was not detected in semen from 22 men at 1- and 6-months after peripheral viral suppression due to addition of indinavir to their ART regimen. This study is important because it provides a new tool for studying HIV infection of the male genital tract, and provides preliminary evidence that cryptic HIV-1 infection may not occur in the genital tract of men on HAART. Further studies will surely follow to confirm and extend these observations. ==== Refs Anderson DJ Yunis EJ "Trojan Horse" leukocytes in AIDS N Engl J Med 1983 309 984 985 6621629 Xu C Politch JA Mayer KH Anderson DJ Detection of Human Immunodeficiency Virus type-1 episomal cDNA in semen AIDS Res and Therapy 2005 000 000 Ho DD Schooley RT Rota TR Kaplan JC Flynn T Salahuddin SZ Gonda MA Hirsch MS HTLV-III in the semen and blood of a healthy homosexual man. Science 1984 226 451 453 6208608 Stewart GJ Tyler JP Cunningham AL Barr JA Driscoll GL Gold J Lamont BJ Transmission of human T-cell lymphotropic virus type III (HTLV- III) by artificial insemination by donor Lancet 1985 2 581 585 2863597 10.1016/S0140-6736(85)90585-9 Van Voorhis BJ Martinez A Mayer K Anderson DJ Detection of human immunodeficiency virus type 1 in semen from seropositive men using culture and polymerase chain reaction deoxyribonucleic acid amplification techniques Fertil Steril 1991 55 588 594 2001759 Pudney J Anderson D Orchitis and human immunodeficiency virus type 1 infected cells in reproductive tissues from men with the acquired immune deficiency syndrome Am J Pathol 1991 139 149 160 1853930 Pudney J Oneta M Mayer K Seage G Anderson DJ Pre-ejaculatory fluid as potential vector for sexual transmission of HIV-1 Lancet 1992 340 1470 1470 1360584 10.1016/0140-6736(92)92659-4 Xu C Politch JA Tucker L Mayer KH Seage GR Anderson DJ Factors associated with increased levels of human immunodeficiency virus type 1 DNA in semen J Infect Dis 1997 176 941 947 9333152 Quayle AJ Xu C Mayer KH Anderson DJ T lymphocytes and macrophages, but not motile spermatozoa, are a signi ficant source of human immunodeficiency virus in semen J Infect Dis 1997 176 960 968 9333154 Semprini AE Vucetich A Hollander L Sperm washing, use of HAART and role of elective Caesarean section Curr Opin Obstet Gynecol 2004 16 465 470 15534441 10.1097/00001703-200412000-00005 Vernazza PL Gilliam BL Dyer JR Fiscus SA Eron JJ Frank AC Cohen MS Quantitation of HIV in semen: Correlation with antiviral treatment and immune status AIDS 1997 11 987 993 9223732 10.1097/00002030-199708000-00006 Cohen MS Hoffman IF Royce RA Kazembe P Dyer JR Daly CC Zimba D Vernazza PL Maida M Fiscus SA Eron JJJ Reduction of concentration of HIV-1 in semen after treatment of urethritis: implications for prevention of sexual transmission of HIV-1. Lancet 1997 349 1868 1873 9217758 10.1016/S0140-6736(97)02190-9 Vernazza PL Troiani L Flepp MJ Cone RW Schock J Roth F Boggian K Cohen MS Fiscus SA Eron JJ Potent antiretroviral treatment of HIV-infection results in suppression of the seminal shedding of HIV. The Swiss HIV Cohort Study AIDS 2000 14 117 121 10708281 10.1097/00002030-200001280-00006 Eron JJ Vernazza PL Johnston DM Seillier-Moiseiwitsch F Alcorn TM Fiscus SA Cohen MS Resistance to HIV-1 to antiretroviral agents in blood and seminal plasma: Implications for Transmission AIDS 1998 12 F189 Dornadula G Nunnari G Vanella M Roman J Babinchak T DeSimone J Stern J Braffman M Zhang H Pomerantz RJ Human immunodeficiency virus type 1-infected persons with residual disease and virus reservoirs on suppressive highly active antiretroviral therapy can be stratified into relevant virologic and immunologic subgroups J Infect Dis 2001 183 1682 1687 11343220 10.1086/320715 Nunnari G Otero M Dornadula G Vanella M Zhang H Frank I Pomerantz RJ Residual HIV-1 disease in seminal cells of HIV-1-infected men on suppressive HAART: latency without on-going cellular infections AIDS 2002 16 39 45 11741161 10.1097/00002030-200201040-00006	16324221	PMC1325240	CC BY	2021-01-04 16:25:14	no	AIDS Res Ther. 2005 Dec 3; 2:11	utf-8	AIDS Res Ther	2,005	10.1186/1742-6405-2-11	oa_comm
==== Front Microb Cell FactMicrobial Cell Factories1475-2859BioMed Central London 1475-2859-4-351635430010.1186/1475-2859-4-35ResearchEncapsulated Escherichia coli in alginate beads capable of secreting a heterologous pectin lyase Papi Rigini M [email protected] Sotiria A [email protected] Fotini A [email protected] Dimitrios A [email protected] Laboratory of Biochemistry, Department of Chemistry, Aristotle University of Thessaloniki, Thessaloniki, 54124, Greece2 National Hellenic Research Foundation, 48 Vasileos Konstantinou Ave., 11635 Athens, Greece2005 14 12 2005 4 35 35 19 10 2005 14 12 2005 Copyright © 2005 Papi et al; licensee BioMed Central Ltd.2005Papi 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 Production of heterologous proteins in the E. coli periplasm, or into the extracellular fluid has many advantages; therefore naturally occurring signal peptides are selected for proteins translocation. The aim of this study was the production in high yields of a recombinant pectin lyase that is efficiently secreted and the encapsulation of transformed E. coli cells for pectin degradation in a biotechnological process. Results The nucleotide sequence of Bacillus subtilis α-amylase's signal peptide was fused to the N-terminal of an heterologously expressed pectin lyase in E. coli BL21 [DE3]. Thus pectin lyase secretion was achieved into the extracellular growth medium. E. coli cells harboring the recombinant plasmid heterologously express pectin lyase to around 22% of the total cellular proteins, as it was estimated by SDS-PAGE and image analysis. IPTG induces the heterologously expressed enzyme, which is initially distributed extracellularly (7 hour) and later on at the periplasmic (9 hours) or cytosolic fraction (20 hours). No pectin lyase activity was found in the membranes fraction and in the inclusion bodies. Encapsulation of the recombinant strains of E. coli in alginate or alginate/silica beads 1:5 showed that pectin lyase could degrade effectively its substrate, for at least ten operational cycles. Conclusion Secretion of an heterologously overexpressed pectin lyase in E. coli BL21 [DE3] was achieved in this study. For this purpose the signal peptide of α-amylase from B. subtilis was fused to the N-terminal domain of pectin lyase. Encapsulated E. coli BL21 [DE3] cells harboring pET29c/exPNL were used successfully for pectin degradation up to ten operational cycles indicating that under special conditions this might have biotechnological implementations. ==== Body Background Pectin lyase (PNL, EC 4.2.2.10) belongs to polysaccharide lyase family 1 and cleaves α-1,4 bonds between galacturonic acid residues of pectin. The mode of PNL action is trans-elimination [1] with end-product oligosaccharides of 4-deoxy-6-methyl-alpha-D-galact-4-enuronosyl groups. Pectins are present in the primary cell wall and intracellular space of higher plants. The enzymic degradation of pectin is of great interest in food industry for fruit juice clarification, reduction of turbidity and increase of the amount of juice extracted from fruit pulp [2,3]. The enzymic degradation is also used in the textile industry in the processing of natural fibers [4]. Among the pectinolytic enzymes, pectin lyase is the only one known to degrade highly esterified pectin without the previous action of other enzymes [5]. Heterologous expression of pectin lyase gene (pnl) from Pseudomonas marginalis N6301 in E. coli BL21 [DE3] has previously been reported [6]. The enzyme was overproduced in the cytoplasm, bearing a 6-His-tag at the C-terminal, and it was purified close to homogeneity by affinity chromatography on a Ni++-NTA agarose column. Even though the heterologous expression of PNL was achieved in high yields there was a number of limitations compared to PNL production from its native source, the bacterium P. marginalis. It was necessary to disrupt bacterial cells to recover the enzyme with the proper protein folding. Since secretion of recombinant proteins is preferred, many signal sequences derived from naturally occurring secretory proteins could be selected to support the efficient translocation of this heterologous polypeptide across the inner membrane, when fused to its amino termini [7]. Signal peptides usually consist of a positively charged amino-terminus (n-region, 1–5 residues in length), a central hydrophobic domain (h-region, 7–15 residues) and a neutral but polar carboxy-terminus domain (c-region, 3–7 residues). The c-region specifies the signal peptide cleavage site for specific signal peptidases whereas the other two domains are required for efficient translocation [8]. The length of signal peptides ranges from 22–32 amino acids. The majority of E. coli secreted proteins is either localized in the periplasm, or is associated with the inner or outer membranes [9]. Although recombinant proteins targeted to the periplasm remain to this cellular compartment, their release into the extracellular fluid may occur through non-specific leakage, or due to cell lysis [10,11]. The signal peptide of α-amylase from the gram-positive bacterium Bacillus subtilis cloned and excreted in E. coli, a Gram- bacterium, maps among the signal peptides of E. coli excreted proteins [12]. This Bacillus signal peptide can also achieve secretion in E. coli of the periplasmic space protein β-lactamase to the extracellular surroundings [13]. In this report, we studied the fusion of the 33-aminoacid-signal sequence of B. subtilis α-amylase to the amino-terminal domain of pectin lyase and the secretion of the heterologous expressed pectin lyase. The transformed bacteria were further encapsulated in alginate beads and were tested successfully for pectin degradation. Results Construction of a himeric PNL carrying the signal peptide of B. subtilis α-amylase Isolated genomic DNA from B. subtilis was used as template for the amplification of the signal sequence of α-amylase gene. Since the desired signal sequence of α-amylase was very small, various primers were designed to anneal upstream and downstream of the promoter and the coding region of α-amylase. After electrophoresis of PCR products on 1.5% w/v agarose gel the amplification fragment of the appropriate size was used as DNA template in a second PCR with primers only for the signal sequence of α-amylase. Amplification of α-amylase signal sequence was confirmed by agarose gel electrophoresis. Fragments were extracted from agarose gel with QIAEXII Gel Extraction Kit (Qiagen) and both signal sequence and pET29c/pnl were digested with NdeI and BamHI. After ligation of the digested plasmid and the insert, plasmid pET29c/expnl was generated (Fig. 1). The existence of α-amylase's signal sequence in pET29c/expnl was confirmed by DNA sequencing. Figure 1 Plasmid diagram of pET29c/expnl. Genes and major features of the plasmid are indicated by thick boxes. The restriction sites used in the construction are indicated as well. Overexpression of exPNL Cultures of E. coli BL21 [DE3] harboring pET29c/expnl were cultivated in LB medium and expnl gene was overexpressed after induction with IPTG. The total expressed proteins are shown at Fig. 2A. Overexpression of exPNL appears at lanes 3–5 of Fig. 2A. The presence of the recombinant exPNL was also confirmed by immunostaining of the same gel by anti-6-His (1:500 dilution), as shown at Fig. 2B lanes 3–5. The overproduction of exPNL was around 22% of the total amount of proteins, as estimated by the Gel-Pro Analyzer (version 3, Media Cybemetics, 1993–97) with the expected molecular mass of 39 kDa. Figure 2 Overexpression of expnl gene. A: SDS-PAGE of E. coli exPNL total cells, lane 1: pre-stained protein markers, lane 2: before the induction, lane 3: 1 h after the induction, lane 4: 2 h after the induction, lane 5: 3 h after the induction. Gel was stained with Coomassie Brilliant Blue R-250 B: transfer of proteins electrophoresed as in A and immunostained with anti-His (1:500) Subcellular distribution of exPNL activity The appearance of pectin lyase in the extracellular fluid as well as at different subcellular fractions was studied. The presence of exPNL was estimated both by following its activity (Fig 3) and by SDS-PAGE (Fig. 4). In the periplasmic fraction exPNL activity was induced right after the addition of IPTG, with a peak at the first hour of induction. One hour later exPNL activity dramatically declined and then another peak of activity appeared 4–8 hours after IPTG addition. An increment of exPNL activity in the extracellular fluid was observed with a small delay comparing to that in the periplasm. Figure 3 The appearance of exPNL activity in different fractions of E. coli, transformed with pET29c/expnl, during bacterial growth. Figure 4 SDS-PAGE of the subcellular fractions of E. coli transformed with pET29c/expnl, at different times of cell growth in presence of IPTG. A: extracellular medium, B: periplasmic space, C: cytoplasm, D: membranes and inclusion bodies – Lane 1: protein markers, lane 2: before the induction with IPTG, lane 3: 20 min after the induction, lane 4: 40 min after the induction, lane 5: 1 h after the induction, lane 6: 1 h & 40 min after the induction, lane 7: 3 h after the induction, lane 8: 6 h after the induction. Gels were stained with silver nitrate. In the cytosolic fraction exPNL activity was increased 4 hours after induction with IPTG, whereas activity was not detectable in other fractions. Total protein (10 μg) of each subcellular fraction and the extracellular medium, were electrophoresed and the proteins were stained with silver nitrate (Fig. 4). A large amount (12%) of inactive exPNL was present in the 100,000 × g pellet, as inclusion bodies, as it was judged by measuring the stained protein by the Gel-Pro Analyzer (Fig. 4D). Solubilization of inclusion bodies with 8 M urea or 6 M guanidine hydrochloride and renaturation by dialysis as described in Material and Methods resulted in an inactive enzyme. The extracellularly secreted exPNL was also immunostained with antiHis (1:500 dilution) and antiPNL (1:500 dilution) (Fig. 5). These results indicate that exPNL is overproduced by IPTG induction and migrates to the periplasmic fraction and is finally excreted to the extracellular medium. Figure 5 Total proteins electrophoresed as described at figure 5A, transferred and immunostained with anti-six-His (A) and anti-PNL (B). Cell encapsulation in alginate beads Sodium alginate is soluble in water and becomes a hydrogel in the presence of multivalent cations (Ca2+, Ba2+, Sr2+). The resultant gel shows good permeability for small molecules [14]. In addition, the gelation of alginate can be achieved under mild conditions for immobilizing living cells. Thus, a variety of living cell phenotypes has been enclosed in such matrices [15-17]. To use more successfully the induced exPNL, transformed bacteria were encapsulated in alginate beads and used to degrade pectin for at least 10 operational cycles. Figure 6 shows the liberated exPNL activity in the extracellular fluid of the encapsulated bacteria at different operational cycle. After each operational cycle the encapsulated cells were allowed to rest for 1 hour in LB medium at 37°C. The pattern of liberated exPNL activity in the extracellular fluid is very similar for the repeated batches, reaching maximal exPNL activity at 20 minutes of incubation. Even though the activity of exPNL during the first cycle was very small, in all other cycles it was very sufficient. As figure 6 shows a repetitive motif of increase and decrease of enzymic activity is observed that may reflect the cell growth inside the beads. Furthermore, the addition of toluene at concentrations of 0.2 mg/ml and 0.4 mg/ml in the reaction mixture resulted, in both cases, in the increase of cell membrane permeability and finally in the 55% increase of liberated exPNL by the encapsulated cells. Figure 6 Enzymic degradation of pectin by encapsulated in alginate beads E. coli cells transformed with pET29c/expnl. Each column represents pectin lyase activity measured at 20 min of incubation. Discussion There is a large number of reports on the secretory expression of recombinant proteins from E. coli [9,18]. Secreted polypeptides are synthesized as preproteins containing an amino-terminal signal sequence that is cleaved during the translocation process. This signal sequence is recognized and then cleaved by inner-membrane-associated signal peptidases. The secretion efficiency of these proteins is strongly dependent on the signal sequence used, the overall structure of the precursor protein and the processing of the signal peptide. Periplasmic production of recombinant proteins has several advantages with respect to cytoplasmic production, but frequently overproduced foreign proteins are not efficiently secreted to the periplasm in E. coli. According to Sjöström et al. [12] the signal peptide of B. subtilis α-amylase maps among the signal peptides of E. coli excreted proteins and when it was added to β-lactamase the protein was secreted to the periplasm [13]. It is also stated that the secretion mechanism of B. subtilis is similar to that of E. coli with minor differences with respect to substrate specificities of their signal peptidases, or in other proteins related to secretion. In particular, the signal peptide of α-amylase has a sequence of Ala-Ala-Ser-Ala recognized by type I signal peptidases, which are common in B. subtilis and E. coli [19]. These data lead us to the construction of plasmid pET29c/expnl from pET29c/pnl carrying the B. subtilis α-amylase's signal sequence, introduced by NdeI and BamHI restriction sites (Fig. 1). The recombinant PNL was secreted in the extracellular medium at around 30% of the total activity at the early stage of induction and later on the activity accumulated in the cytoplasm. Those results suggested that either the bacterial secretory mechanism was not working effectively, or that the degradation of the protein was very fast. Although the transformed bacteria were cultivated at low temperature, the overproduced exPNL was partially segregated as inclusion bodies. Recent reports presented that co-overproduction of SecB, DnaK-DnaJ and GroEL-GroES has met variable success and improved secretion for a number of heterologous proteins in E. coli [20-23]. Since the signal sequence influences secondary and tertiary structure formation of the secretory proteins, which in turn affect chaperone recognition, we would like to try several signal sequences and/or overproduce different chaperones to optimize the translocation of the overproduced exPNL in the near future. Even though exPNL was not secreted in high levels and the rate of hydrolysis was low, in our experiments we showed that transformed E. coli with pET29/expnl encapsulated in alginate beads or alginate/silica gel 1:5 can be used for pectin degradation effectively, for at least ten operational cycles. So far the commercial pectolytic preparations that are used in food processing industries, are crude multi-enzyme preparations from Aspergillus niger and contain hemicellulase, cellulase, xylanase, pectin esterase, polygalacturonase and pectin lyase [24]. Although pectolytic enzymes are widely used for industrial purposes, little research has been carried out on pectolytic enzymes immobilized in solid matrices and even less on cell encapsulation. Previously pectin lyase has been immobilized on DEAE-cellulose, on porous glass [25] and recently on nylon, with no loss of enzymic activity after 12 cycles of operation [26]. Our results of using encapsulated bacterial cells capable of secreting pectin lyase appear to be of potential interest for an industrial application. Conclusion The aim of this study was the secretion of an heterologously overexpressed pectin lyase in E. coli BL21 [DE3]. For this purpose, the signal peptide of α-amylase from B. subtilis was fused to the N-terminal domain of pectin lyase that had been previously [6] cloned and overexpressed in E. coli BL21 [DE3]. Secretion into the extracellular medium was observed right after IPTG induction but then the fusion protein was accumulated in the periplasm and cytoplasm. Encapsulated E. coli BL21 cells harboring pET29c/exPNL in alginate beads, capable of secreting pectin lyase were used successfully for pectin degradation up to ten operational cycles, providing expenditure for future biotechnological applications. Materials and methods Material All chemical compounds were purchased from SIGMA, Fluka, Baker and BDH Chemicals. Restriction enzymes, Taq polymerase and other reagents needed for molecular biology experiments were purchased from New England Biolabs. Isolation of DNA Genomic DNA from B. subtilis was isolated according to Saito and Miura as modified by Gay [27,28]. Plasmid DNA was extracted by alkaline lysis [29]. Construction of pET29c/expnl containing α-amylase's signal sequence We have recently reported that pnl gene of P. marginalis was amplified by PCR and cloned into plasmid pET29c using BamHI and HindIII restriction sites [6]. In this paper plasmid pET29c/expnl is constructed from pET29c/pnl and B. subtilis α-amylase's signal sequence [30] to secrete pectin lyase. Larger fragments containing α-amylase's signal sequence were amplified by PCR using B. subtilis genomic DNA as the template and proper oligonucleotides. The reaction was performed as follows: 50 μl reaction volume was prepared to contain 2 units of Taq polymerase, 5 μl of polymerase buffer, 200 μM of each dNTP, 0.13 μg/μl of each primer and 1 ng of template DNA. The reaction mixture was cycled 30 times using the following conditions: 92°C for 1 min, 54°C for 30 sec, 72°C for 50 sec and another 5 min at 72°C at the end of the cycles. The amplification products were electrophoresed in 1.5% w/v agarose gel and products of right size were used as the template DNA in a second PCR, where primers specific for α-amylase's signal sequence and with NdeI and BamHI restriction sites were used (forward: 5'-GGAATTCCATATGTTTGCAAAACGA-3', reverse: 5'-CGCGGATCCTACTCGCAGCC-3'). The reaction mixture and the conditions for PCR were the same as described above. The PCR product was electrophoresed in 1.5% w/v agarose gel and then was cut and extracted from the gel using the Qiagen gel extraction kit. The 99 bp NdeI-BamHI PCR fragment was introduced into the NdeI-BamHI fragment of pET29c/pnl resulting in pET29c/expnl plasmid. E. coli BL21 [DE3] competent cells were prepared according to Sambrook et al. [29] and transformed with pET29c/expnl. Microorganisms and growth conditions The recombinant strains of E. coli BL21 [DE3] with the constructed plasmid pET29c/expnl were grown at 37°C, in LB medium containing kanamycin (50 μg/ml) with vigorous agitation. When cells reached an optical density of 0.6 at 600 nm, IPTG (1 mM) was added and the cultures were incubated at 24°C, for 3 hours. B. subtilis was grown at 37°C, in LB medium with vigorous agitation. Preparation of cell extracts and subcellular fractionation Recombinant strains of E. coli BL21 [DE3] harboring pET29c/expnl, grown as described above, were harvested by centrifugation at 3,000 × g for 15 min at 4°C. The cell pellet was washed twice with PBS and suspended in 5 × ml of 0.5 M sucrose, 250 mM EDTA and 2 mg/ml lysozyme. The suspension was incubated on ice for 30 min and centrifuged at 20,000 × g for 15 min. The supernatant was kept/saved as the periplasmic fraction. The remaining pellet was suspended in 50 mM Tris-HCI (pH 7.5), 2.5 mM EDTA and then sonicated (5 times × 30 sec, 0.5 cycle, 50% amplitude – UP200 s dr. hielscher, GmbH). After addition of MgSO4 to a final concentration 40 mM, the mixture was centrifuged at 100,000 × g for 40 min. The 100,000 × g supernatant was kept as the cytoplasmic fraction and the 100,000 × g pellet as the membrane fraction with the inclusion bodies. The pellet 100,000 × g was suspended in 1 ml of 50 mM Tris-HCl buffer pH 7.5 [31]. Enzyme assay PNL activity was determined spectrophotometrically by monitoring the increase in absorbance at 235 nm of a solution containing 0.1% w/v pectin (Sigma, P-9135) in 50 mM citrate/phosphate buffer pH 6.5 at 37°C. The increase in absorbance was measured after 20 minutes of incubation. One unit of PNL activity was defined as the nmoles of unsaturated oligogalacturonides produced per minute, under the above-specified condition [32]. Specific activity of PNL was defined as the ratio of PNL units per mg of protein extract. The absorption coefficient of unsaturated oligogalacturonides is 5500 M-1cm-1. Electrophoresis and immunoblotting SDS electrophoresis was performed in 10% w/v polyacrylamide gels containing 0.1% w/v SDS as described by Laemmli [33]. Protein content was determined by the method of Bradford [34]. Gels were stained either with Coomassie Brilliant Blue R250 [35] or with silver nitrate [36]. Proteins were transferred to nitrocellulose membranes following the method of Towbin et al. [37] and immunostained as described [38]. Solubilization and renaturation of inclusion bodies The 100,000 × g pellet containing the inclusion bodies was either dissolved in 10 fold volume of 50 mM Tris-HCl pH 8.0, 5 mM NaCl, 5 mM MgCl2, 2.5 mM β-MSH, 0.1% w/v Tween 20 (DR, Denaturation – Renaturation buffer) containing 8 M urea, or in 50 mM Tris-HCl pH 8.0, 6 M guanidine hydrochloride, 1 mM DTT to a final concentration of 1 mg/ml. Solubilization of inclusion bodies was proceed at room temperature for 1 hour. In the first case the mixture was centrifuged at 100,000 × g for 20 min at 4°C and then the 100,000 × g supernate was dialyzed gradually against DR buffer containing lower concentrations of urea (6, 4, 2, 1 mM urea) and finally overnight just in DR buffer. In the second case the mixture was diluted tenfold with the same buffer and dialyzed overnight against a 100 fold volume of 50 mM Tris-HCl pH 8.0, 150 mM NaCl, 1 mM DTT and 5% glycerol. Dialysis was followed by centrifugation at 50,000 × g for 30 min at 4°C and enzyme activity was determined. In all steps 1 mM PMSF was added. Formation of alginate beads After 2 hours of induction with IPTG, cells were harvested at 3000 × g for 20 min and mixed with sterile water (300 mg cells/ml). Sodium alginate solution (2% v/v) was mixed with the above cell suspension at a volumetric ratio of 9:1 and the mixture was dropped with an insulin syringe into a gently stirred 1% w/v solution of BaCl2 at a volumetric ratio of 1:50. A spontaneous cross-linking reaction produced spherical hydrogel beads of barium alginate, which remained for 30 minutes at 4°C in order to stabilize the formed network. After stabilization, the beads were washed with 0.9% v/v NaCl (twice) and used immediately. All solutions that were used are sterile. Pectin hydrolysis by the transformed bacteria encapsulated in alginate beads Pectin hydrolysis was performed in conical flasks at 37°C under gentle stirring and each operational cycle lasts for 1 h. The substrate used was the same as for enzyme assay with the addition of LB solution (25:1 final concentration) and BaCl2 at a final concentration of 0.005% v/v in order to preserve cell viability and network stability. At the end of each cycle, beads were washed with 0.9% v/v NaCl (twice) and placed with fresh LB solution at room temperature under gentle shaking for 1 h. This procedure was repeated after each operational cycle. Abbreviations PNL, pectin lyase; exPNL, extracellular pectin lyase; IPTG, isopropylthiogalactoside Authors' contributions All authors contributed equally in this paper. ==== Refs Linhardt RJ Galliher PM Cooney CL Polysaccharide lyases Appl Biochem Biotechnol 1986 12 135 176 3521491 Gupta MN Guoqiang D Mattiasson B Purification of endo-polygalacturonase by affinity precipitation using alginate Biotechnol Appl Biochem 1993 18 321 327 8297509 Pilnik W Rombouts FM Polysaccharides and food processing Carbohydr Res 1985 142 93 105 3000594 10.1016/S0008-6215(00)90736-5 Sharma HSS Enzymatic degradation of residual non cellulosic polysaccharides present on dew-retted flax fibres Appl Microbiol Biotechnol 1987 26 358 362 10.1007/BF00256669 Alana A Alkorta I Dominguez JB Llama MJ Serra JL Pectin lyase activity in a Penicillium italicum strain Appl Environ Microbiol 1990 56 3755 3759 16348377 Papi RM Kyriakidis DA Overexpression of the pectin lyase of Pseudomonas marginalis in Escherichia coli and purification of the active enzyme Biotechnol Appl Biochem 2003 37 187 194 12630908 10.1042/BA20020055 Baneyx F Recombinant protein expression in Escherichia coli Curr Opin Biotechnol 1999 10 411 421 10508629 10.1016/S0958-1669(99)00003-8 Von Heijne G How signal sequences maintain cleavage specificity J Mol Biol 1984 173 243 251 6423828 10.1016/0022-2836(84)90192-X Georgiou G Segatori L Preparative expression of secreted proteins in bacteria: status report and future prospects Curr Opin Biotechnol 2005 16 1 8 10.1016/j.copbio.2005.07.008 Lazzaroni JC Portalier RC Genetic and biochemical characterization of periplasmic leaky mutants of Escherichia coli K-12 J Bacteriol 1981 145 1351 1358 7009581 Hewinson RG Russell WP Processing and secretion by Escherichia coli of a recombinant form of the immunogenic protein MPB70 of Mycobacterium bovis J Gen Microbiol 1993 139 1253 1259 8360619 Sjöström M Wold S Wieslander A Rilfors L Signal peptide amino acid sequences in Escherichia coli contain information related to final protein localization. A multivariate data analysis EMBO J 1987 6 823 831 3556168 Nakazawa K Takano T Sohma A Yamane K Secretion activities of Bacillus subtilis alpha-amylase signal peptides of different lengths in Escherichia coli cells Biochem Biophys Res Commun 1986 134 624 631 3080993 10.1016/S0006-291X(86)80465-X Li RH Altreuter DH Gentile FT Transport characterization of hydrogen matrices for cell-encapsulation Biotechnol Bioeng 1996 50 365 373 10.1002/(SICI)1097-0290(19960520)50:4<365::AID-BIT3>3.0.CO;2-J Willaert RG Baron GV Gel entrapment and microencapsulation: methods, applications and engineering principles Rev Chem Eng 1996 12 85 115 Smidsrod O Skjak-Braek G Alginate as immobilization matrix for cells Trends Biotechnol 1990 8 71 78 1366500 10.1016/0167-7799(90)90139-O Sakai S Ono T Ijima H Kawakami K Behavior of enclosed sol- and gel-alginates in vivo Biochem Engin J 2004 22 19 24 10.1016/j.bej.2004.07.010 Wan EW-M Baneyx F TolAIII co-overexpression facilitates the recovery of periplasmic recombinant proteins into the growth medium of Escherichia coli Protein Expres Purif 1998 14 13 22 10.1006/prep.1998.0941 Tuteja R Type I signal peptidase: An overview Arch Biochem & Biophys 2005 441 107 111 16126156 10.1016/j.abb.2005.07.013 Bergés H Joseph-Liauzun E Fayet O Combined effects of the signal sequence and the major chaperone proteins on the export of human cytokines in Escherichia coli Appl Environ Microbiol 1996 62 55 60 8572712 Thomas JG Ayling A Baneyx F Molecular chaperones, folding catalysts and the recovery of active recombinant proteins from Escherichia coli: to fold or to refold Appl Biochem Biotechnol 1997 66 197 238 9276922 Pérez-Pérez J Martinez-Caja C Barbero JL Gutiérrez J DnaK/DnaJ supplementation improves the periplasmic production of human granulocyte-colony stimulating factor in Escherichia coli Biochem Biophys Res Commun 1995 210 524 529 7538761 10.1006/bbrc.1995.1691 Hayhurst A Harris WJ Escherichia coli Skp shaperone coexpression improves solubility and phage display of single chain antibody fragments Protein Expr Purif 1999 15 336 343 10092493 10.1006/prep.1999.1035 Naidu GSN Panda T Production of pectolytic enzymes – a review Bioprocess Engineering 1998 19 355 361 10.1007/s004490050532 Hanisch WH Rickard PAD Nyo S Poly(methoxygalacturonide) lyase immobilized via titanium onto solid supports Biotechnol Bioeng 1978 20 95 106 10.1002/bit.260200108 Alkorta I Garbisu C Llama ML Serra JL Immobilization of pectin lyase from Penicillium italicum by covalent binding to nylon Enzyme Microb Technol 1996 18 141 146 10.1016/0141-0229(95)00094-1 Saito H Miura K Preparation of transforming deoxyribonucleic acid by phenol treatment Biochim Biophys Acta 1963 72 619 629 14071565 10.1016/0006-3002(63)90288-9 Gay NJ Construction and characterization of an Escherichia coli strain with a uncl mutation J Bacteriol 1984 158 820 825 6327640 Sambrook J Fritsch EF Maniatis T Molecular cloning. A laboratory manual 1989 Cold Spring Harbor Laboratory Press: New York Yang M Galizzi A Henner D Nucleotide sequence of the amylase gene from Bacillus subtilis Nucleic Acids Res 1983 11 237 249 6186986 Yamane K Mizushima S Introduction of basic amino acid residues after the signal peptide inhibits protein translocation across the cytoplasmic membrane of Escherichia coli J Biol Chem 1988 263 19690 19696 3058705 Albersheim P Killias U Studies relating to the purification and properties of pectin transeliminase Arch Biochem Biophys 1962 97 107 115 13860094 10.1016/0003-9861(62)90050-4 Laemmli UK Cleavage of structural proteins during the assembly of the head of bacteriophage T4 Nature 1970 227 680 685 5432063 10.1038/227680a0 Bradford MM A rapid and sensitive method for the quantitation of microgram quantities of proteins utilizing the principle of protein-dye binding Anal Biochem 1976 72 248 254 942051 10.1016/0003-2697(76)90527-3 Hames BD Rickwood D Gel Electrophoresis. A practical approach 1985 IRL Press: Oxford-Washington DC Blum H Beir H Gross HJ Improved silver staining of plant proteins, RNA and DNA in polyacrylamide gels Electrophoresis 1987 8 93 98 10.1002/elps.1150080203 Towbin H Staehelin T Gordon J Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications Proc Natl Acad Sci (USA) 1979 76 4350 4354 388439 Blake MS Johnston KH Russels-Jones GJ Gotschlich EC A rapid, sensitive method for the detection of alkaline phosphatase-conjugated anti-antibody on western blots Anal Biochem 1984 136 175 179 6424501 10.1016/0003-2697(84)90320-8	16354300	PMC1325241	CC BY	2021-01-04 16:24:35	no	Microb Cell Fact. 2005 Dec 14; 4:35	utf-8	Microb Cell Fact	2,005	10.1186/1475-2859-4-35	oa_comm
==== Front BMC CancerBMC Cancer1471-2407BioMed Central London 1471-2407-5-1551633226010.1186/1471-2407-5-155Research ArticleIncreased staining for phospho-Akt, p65/RELA and cIAP-2 in pre-neoplastic human bronchial biopsies Tichelaar Jay W [email protected] Yu [email protected] Jean C [email protected] Paul W [email protected] Stephen [email protected] Marshall W [email protected] Department of Environmental Health, University of Cincinnati College of Medicine, Cincinnati, OH, 45267 USA2 Department of Genome Science, University of Cincinnati College of Medicine, Cincinnati, OH, 45237 USA3 Cancer Imaging Department, British Columbia Cancer Agency, Vancouver, BC, Canada V5Z4E6. USA4 Department of Pathology and Laboratory Medicine, University of Cincinnati College of Medicine, Cincinnati, OH, 45267. USA2005 6 12 2005 5 155 155 15 8 2005 6 12 2005 Copyright © 2005 Tichelaar et al; licensee BioMed Central Ltd.2005Tichelaar 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 development of non-small cell lung carcinoma proceeds through a series of well-defined pathological steps before the appearance of invasive lung carcinoma. The molecular changes that correspond with pathology changes are not well defined and identification of the molecular events may provide clues on the progression of intraepithelial neoplasia in the lung, as well as suggest potential targets for chemoprevention. The acquisition of anti-apoptotic signals is critical for the survival of cancer cells but the pathways involved are incompletely characterized in developing intra-epithelial neoplasia (IEN). Methods We used immunohistochemistry to determine the presence, relative levels, and localization of proteins that mediate anti-apoptotic pathways in developing human bronchial neoplasia. Results Bronchial epithelial protein levels of the phosphorylated (active) form of AKT kinase and the caspase inhibitor cIAP-2 were increased in more advanced grades of bronchial IEN lesions than in normal bronchial epithelium. Additionally, the percentage of biopsies with nuclear localization of p65/RELA in epithelial cells increased with advancing pathology grade, suggesting that NF-κB transcriptional activity was induced more frequently in advanced IEN lesions. Conclusion Our results indicate that anti-apoptotic pathways are elevated in bronchial IEN lesions prior to the onset of invasive carcinoma and that targeting these pathways therapeutically may offer promise in prevention of non-small cell lung carcinoma. ==== Body Background Lung cancer is the leading cause of cancer mortality in both men and women in the United States [1]. Non-small cell lung carcinomas arise from the respiratory epithelium and progress through well-defined pathological stages prior to becoming invasive and metastatic tumors. While many studies have identified lung tumor markers of clinical or prognostic significance, survival rates for this deadly disease have remained essentially unchanged for the past 30 years. The slow advance in treating lung cancer is due in part to continued gaps in our understanding of the molecular mechanisms of lung tumorigenesis. Thus, studies that aid in our understanding of molecular mechanisms of lung tumorigenesis are important steps towards developing better detection, prevention and treatment of this disease. Evasion of apoptosis by tumor cells is a critical step during tumorigenesis. The serine/threonine kinase Akt is a critical mediator of anti-apoptotic signaling in eukaryotic cells and is activated in a signaling cascade downstream of Ras activation and phosphoinositide-3-kinase (PI3K) [2]. Amplification of PI3K is common in many tumor types, including lung cancer [3-5] and in lung cancer is correlated with increased phosphorylation of Akt [4]. Activation of Akt, as measured by phosphorylation of the protein, is also increased in multiple tumor types including lung cancer [6-10]. Increased phosphorylation of Akt kinase has also been reported in developing bronchial hyperplasias and dysplasias [7,11,12] and pre-neoplastic atypical alveolar hyperplasia [13], indicating that activation of this pro-survival pathway may be a relatively early event in lung tumorigenesis. The NF-κB transcription factor family can stimulate both pro- and anti-apoptotic signals. Many studies have described a critical role for NF-κB activity in promoting cell survival. Increased staining for NF-κB subunits has been detected in breast [14] and cervical carcinoma [15]. Inhibition of NF-κB activity either pharmacologically or genetically can sensitize tumor cells to pro-apoptotic agents [16-19] or to tumor necrosis factor-α (TNF-α) induced apoptosis [20]. Proteasome inhibition, which blocks the degradation of inhibitor of κB (IκB) protein, thus blocking NF-κB nuclear translocation and activation, also sensitized NSCLC cells to apoptosis [21,22]. Similarly, expression of a super-repressor form of IκBα sensitized lung cancer cell lines to apoptosis-inducing drugs [16,23]. The super-repressor form of IκBα, as well as a dominant negative form of IKK, also blocked Ras-mediated transformation of cells [24,25] and expression of the IκB super-repressor inhibited anchorage independent growth and metastatic spread of human lung cancer cell lines in a tumor xenograft model [26]. Furthermore, Akt can activate the transcriptional potential of the p65/RELA subunit of NF-κB [27,28], providing a potential link between Akt kinase activity and NF-κB activation. Western blot analysis has demonstrated over expression of the p50 subunit of NF-κB in lung cancer [29], but localization of NF-κB family members has not been described in lung tumors. Nevertheless, abundant evidence links NF-κB transcriptional activation with lung tumorigenesis. Several NF-κB-regulated genes that function in control of apoptosis have been described including cIAP-1, cIAP-2, A1/Bfl1, Traf1, Traf2 and Bcl-XL [30-33]. cIAP-1 and -2 are members of the baculoviral IAP repeat-containing (BIRC) gene family. cIAP-2/BIRC3 is expressed in lung adenocarcinoma cell lines [34] and can be induced by TNF-α [35]. Elevated expression of cIAP-2/BIRC3 has been reported in human NSCLC [36-38] and elevated expression of the related proteins XIAP/BIRC4 and survivin/BIRC5 are also seen in NSCLC [36,39], implicating the BIRC family of proteins as important mediators of lung tumorigenesis. While the expression of pro-survival genes has been examined in some detail in lung tumors, relatively few studies have examined the expression of these proteins in pre-neoplastic lesions. To determine the presence and abundance of pro-survival proteins in developing lung neoplasia, we examined human bronchial biopsies of various grades (normal, hyperplasia, mild, moderate and severe dysplasia, carcinoma in situ and carcinoma) for the presence of phospho-Akt, p65/RELA and cIAP-2/BIRC3 and determined their localization within cells. The results obtained provide new insight into the distribution of pro-survival genes in human lung neoplasia and pre-neoplasia. This information may be useful in designing studies to test the importance of these pathways experimentally and ultimately may lead to improved diagnostic or therapeutic strategies for human lung cancer. Methods Description of samples Bronchial biopsies were obtained during autofluorescence [40] bronchoscopic examination, formalin fixed and embedded in paraffin. Five-micron sections were stained with H&E, examined and scored for pathology grade by an experienced lung pathologist (JCL) using the WHO classification [41]. A separate group of surgically resected NSCLC samples (adenocarcinoma and squamous cell carcinoma) was obtained from Surgical Pathology at the University of Cincinnati. All human samples were obtained under approved IRB protocols at the respective institutions. Table 1 describes the general properties of the biopsies stained including age and gender. Age and gender distribution among the groups was not statistically significant (one way ANOVA, p = 0.558 for age; Kruskal-Wallis one way ANOVA on ranks, p = 0.158 for gender). Of the 22 carcinomas where staging data was available, 36% were stage IA, 23% stage IB, 9% stage IIB, 23% stage III or IV and 9% unclassified. Tumor stages were assigned based on pTNM classifications according to UICC guidelines [42]. Table 1 Pathology Grade No. patients No. biopsies Median age Gender (% female) Normal 10 12 62.7 70 Hyperplasia 11 11 64.3 45 Mild/Moderate Dysplasia 17 23 64.2 53 Severe Dysplasia/CIS 7 9 67.9 29 Carcinoma 22 24 60.5 77 Antibodies and Immunohistochemistry Rabbit polyclonal antibodies for phosphorylated (Ser473, catalog # 9277) and total Akt (catalog # 9272) were purchased from Cell Signaling Technologies (Beverley, MA) and were both used at 1:100 dilution. The p65/RELA antibody (catalog # ab7970) was purchased from Abcam America (Cambridge, MA) and used at a 1:4000 dilution. The cIAP-2/BIRC3 antibody (catalog # AF817) was from R&D Systems (Minneapolis, MN) and used at a 1:500 dilution. Biotinylated goat anti-rabbit secondary antibody and normal goat serum was from Vector Laboratories (Burlingame, CA). Paraffin embedded sections were deparaffinized through xylene and a graded series of ethanols. Endogenous peroxidase activity was quenched by incubation in 2% hydrogen peroxide in methanol for 15 minutes then cleared in PBS for 5 minutes. High temperature antigen unmasking in citrate buffer was used for all antibodies and carried out as described previously [43]. Non-specific binding was blocked by incubation with 5% normal goat serum in PBS + 0.2% Triton X-100 (blocking serum) for 2 hours at room temperature. Slides were then incubated overnight at 4°C with primary antibody at the appropriate dilution in blocking serum. The next day slides were washed 5× for 5 minutes each in PBS + 0.2% Triton X-100 before addition of secondary antibody. Secondary antibody was added at a 1:200 dilution in PBS+Triton and incubated for 30 minutes at room temperature with shaking. Slides were then washed 5× for 5 minutes each with PBS+Triton. A Vectastain ABC kit (Vector Laboratories) was used to prepare avidin-biotin complexes for detection of secondary antibody. Antigen localization was enhanced with Ni-DAB and Tris-cobalt [44] followed by counterstaining with Nuclear Fast Red. Evaluation of immunohistochemical staining and statistical analysis Biopsies received a numerical score of 0 for negative staining, 1 for predominantly faint staining, 2 for predominantly moderate staining or 3 for predominantly strong staining. For the purpose of this study, the bronchial biopsies were divided into 5 categories for phosphorylated Akt and cIAP-2/BIRC3 analysis: normal, hyperplasia, mild dysplasia or moderate dysplasia, severe dysplasia or carcinoma in situ and carcinoma. For analysis of p65/RELA nuclear translocation, the mild and moderate dysplasia groups were analyzed separately as there was a large difference in the percentage of nuclear staining between these two groups. No significant difference in staining intensity between mild and moderate dysplasia was seen for phosphorylated Akt or cIAP-2/BIRC3. Statistical analysis was performed using SigmaStat software (SyStat Software, Inc.). Analysis of variance for each stain was calculated using a Kruskal-Wallis chi-squared test. Pairwise comparisons were calculated using a non-parametric Mann-Whitney rank-sum test. A p-value < 0.05 was considered significant. Results Immunohistochemistry for Akt protein kinase The serine-threonine kinase Akt is a central regulator of cell survival and apoptosis that is itself activated by phosphorylation. Antibodies that specifically recognize the phosphorylated form (Ser 473) were used to identify the phosphorylated, and thus activated, form of Akt. Intensity of staining was scored on a scale of 0 for negative, 1 for predominantly faint staining, 2 for predominantly moderate staining and 3 for predominantly strong staining. In normal bronchial epithelium, phospho-Akt staining typically consisted of faint cytoplasmic stain with occasional cells staining with a more intense nuclear pattern (Figure 1A). In some biopsies, there was a concentration of staining at the apical surface of the epithelium. The underlying interstitial cells were predominantly negative. With increasing pathological grade, there was an overall increase in the intensity of phospho-Akt staining (Figure 1) and an increase in the IHC score (Figure 2A). As pathology grade increased to mild, moderate and severe dysplasia, there were also an increased number of cells with nuclear staining for phospho-Akt (Figure 1B,1C and 1D) in addition to an increase in cytoplasmic staining intensity (analysis of variance analysis using the Kruskal-Wallis Chi-square test, p = 0.018). In addition, pairwise comparisons using the non-parametric Mann-Whitney rank-sum test indicated that staining intensity of the mild/moderate dysplasia and severe dysplasia/CIS groups were significantly increased compared to normal (p = 0.02 and 0.004, respectively). Similar to what has been previously reported [12], invasive carcinoma had a lesser degree of staining than advanced pre-neoplastic lesions (Figure 2A). In squamous cell carcinomas, several tumors had intense plasma membrane associated staining for phospho-Akt (Figure 3A), while an antibody that recognized both phosphorylated and unphosphorylated Akt showed equal staining throughout the cytoplasm and nucleus (Figure 3B). Figure 1 Immunohistochemical localization of phospho-Akt in human bronchial biopsies. Sections from human bronchial biopsies were incubated with antibodies specific for the phosphorylated form (ser473) of Akt, color developed with nickel-DAB (black) and counterstained with nuclear fast red. Representative stains of normal (A), mild dysplasia (B), moderate dysplasia (C) and severe dysplasia (D). Figure 2 Frequency plots of predominant IHC score for phospho-Akt and cIAP-2/BIRC3 in human pre-neoplastic bronchial biopsies. Pre-neoplastic biopsies were scored based on the predominant staining intensity observed using a scale of 0 for negative, 1 for faint, 2 for moderate and 3 for strong. Each symbol represents a separate biopsy and the horizontal line represents the mean score for each category: normal, hyperplasia, mild or moderate dysplasia, severe dysplasia or carcinoma in situ and carcinoma. (A), Scores for predominant phospho-Akt staining intensity. Mean scores were: normal, 1.0; hyperplasia, 1.3; mild or moderate dysplasia, 2.0; severe dysplasia or carcinoma in situ, 2.6; carcinoma, 1.6. (B), Scores for predominant cIAP-2/BIRC3 staining intensity. Mean scores were: normal, 1.1; hyperplasia, 0.89; mild or moderate dysplasia, 0.86; severe dysplasia or carcinoma in situ, 1.8; carcinoma, 1.8. Figure 3 Immunohistochemical localization of phospho-Akt and total Akt in human squamous cell carcinoma. Serial section from human squamous cell carcinoma was stained with antibodies specific for phospho-Akt (ser473) (A), or an antibody that recognizes both phosphorylated and non-phosphorylated forms of Akt (B). Localization of phospho-Akt to the plasma membrane of a subset of tumor cells is apparent. Incubation without primary antibody (C) was used as a negative control. Immunohistochemistry for NF-κB subunit p65/relA The transcription factor NF-κB is activated following Akt phosphorylation via several mechanisms and is in itself known to play a role in cell survival and protection from apoptosis. We thus characterized the immunohistochemical staining pattern of the NF-κB subunit p65/RELA in pre-neoplastic bronchial biopsies. NF-κB subunits are normally held in the cytoplasm, thus preventing them from activating transcription, and are translocated to the nucleus following various activation signals such as oxidative stress or growth factor/cytokine stimulation. We therefore determined the staining pattern of p65/RELA in the cytoplasmic or nuclear compartment of airway epithelial cells. Faint to moderate cytoplasmic staining for p65 was present in the cytoplasm of epithelial cells in normal biopsies with little staining in the underlying stroma (Figure 4A and 4B). However, with the exception of one isolated cluster of cells in a single biopsy, p65/RELA staining was not present in the nucleus. With increasing pathology score, nuclear p65/RELA staining in the bronchial epithelium was seen in an increasing percentage of biopsies (Table 2). There was also a trend towards increased intensity of cytoplasmic staining with increasing pathology grade (Figure 4). Analysis of variance gave a significant p-value of 0.010 for differences in percent nuclear localization of p65 among the groups. Pairwise comparisons indicated that moderate dysplasia and severe dysplasia/CIS were significantly different from normal (p = 0.022 and 0.048, respectively). In normal bronchial epithelium, staining of the submucosa was restricted to a few isolated cells, however, many biopsies with moderate to severe dysplasia showed intense staining in submucosal cells in addition to intense epithelial staining (Figure 4C and 4D). In NSCLC, nuclear localization of p65/RELA was seen more frequently in squamous cell carcinomas (Figure 4E and 4F) then adenocarcinomas. In all cases, nuclear p65/RELA staining was focal within a biopsy. Figure 4 Nuclear translocation of p65/RELA in progressing lung neoplasia. Photomicrographs of normal bronchial epithelium (A, B), moderate dysplasia (C, D), or squamous cell carcinoma (E, F) at low (A, C and E) or high (B, D and F) magnification. Staining is diffusely cytoplasmic in normal bronchial epithelium. Nuclear translocation of p65/RELA, rarely detected in normal bronchial epithelium, is present in an increased percentage of moderate to severe dysplasias and squamous cell carcinoma (arrows, D and F). Table 2 Percent of biopsies with nuclear p65/RELA staining. Pathology Grade % nuclear p65/RELA stain Normal 11 (n = 9) Hyperplasia 20 (n = 10) Mild Dysplasia 10 (n = 10) Moderate Dysplasia 73 (n = 11) Severe dysplasia/CIS 71 (n = 7) Carcinoma 41 (n = 17) Immunohistochemistry for cIAP-2 The inhibitor of apoptosis protein (IAP) family function to block apoptosis by caspase-dependent and caspase-independent means [45]. Several members of this family, including cIAP-2, are known transcriptional targets of NF-κB [31]. In normal bronchial epithelium, cIAP-2 staining was typically present in a faint cytoplasmic pattern (Figure 5A). In some biopsies, a perinuclear pattern was observed in epithelial basal cells (Figure 5A and 5B) and there were scattered cells with positive nuclear stain. The average intensity and localization of staining did not differ significantly from normal in hyperplasias or in mild to moderate dysplasias (Figure 2B). Staining intensity was increased in the severe dysplasia/CIS category and in carcinomas (Figure 5C). Strong perinuclear and nuclear staining was common and more widespread in the higher pathology grades compared to normal. Staining of cells in the submucosa was also seen more frequently in the higher pathology grades (severe dysplasia/CIS and carcinoma) compared to normal (Figure 5C). Analysis of variance on cIAP-2 staining intensity indicated a significant difference among the groups (p = 0.048). Figure 5 Immunohistochemical localization of cIAP-2/BIRC3. Photomicrographs of normal bronchial epithelium (A), mild dysplasia (B), severe dysplasia (C) and a negative control processed without primary antibody (D). In normal and mild dysplasia, perinuclear staining was observed, predominantly in cells adjacent to the basement membrane (for example, arrow in panel B). Increased staining intensity and increased frequency of nuclear staining was observed in severe dysplasias (panel C). Correlation between cIAP-2 and phospho-Akt staining intensity We compared staining intensity of phospho-Akt and cIAP-2 in biopsies where both stains were performed on serial sections (n = 53). There was a significant correlation between phospho-Akt and cIAP-2 staining intensities in these biopsies (Spearman correlation coefficient, r = 0.484, p = 0.0002). This correlation was seen at all pathology grades. The cIAP-2 gene is a transcriptional target for NFκB activation [31]. To examine whether cIAP-2 expression was increased specifically in areas of nuclear p65/RELA staining, we examined serial sections of biopsies stained with both antibodies. AN example of serial sections of a moderate dysplasia (Figure 6A–C) demonstrates nuclear p65/RELA staining accompanied by increased staining intensity of phospho-Akt and cIAP-2. An adenocarcinoma with strong cytoplasmic staining for p65/RELA with moderate staining for phospho-Akt and cIAP-2 is also shown (Figure 6D–F). Co-localization of nuclear p65/RELA staining and increased cIAP-2 staining was not a consistent observation, and elevated cIAP-2 staining was seen in bronchial IEN lesions without detectable nuclear staining for p65/RELA. Additionally, even in biopsies with nuclear localization of p65/RELA, the number of cells with nuclear p65/RELA was fewer than the number of cells with increased cIAP-2 staining intensity. Thus, there was not a clear correlation between nuclear p65/RELA staining and increased staining intensity for cIAP-2 in human bronchial IEN lesions or human NSCLC. Figure 6 Staining for p65/RELA, phospho-Akt and cIAP-2 on serial biopsy sections. Serial sections from a moderate dysplasia (A–C) or adenocarcinoma (D–F) underwent immunohistochemical staining with antibodies directed against p65/RELA (A and D), phospho-Akt (B and E) or cIAP-2 (C and F). Cells with nuclear localization of p65/RELA (arrows in A) correlated with regions of increased staining for phospho-Akt and cIAP-2 in this moderate dysplasia but the correlation was not consistently observed. High levels of cytoplasmic p65/RELA staining in the adenocarcinoma corresponded with faint to moderate staining for phospho-Akt and cIAP-2 (D–F). There was a statistically significant correlation between cIAP-2 and phospho-Akt staining intensity at all pathology grades. We also examined if there was a correlation between tumor stage and staining intensity or nuclear localization of p65/RELA in carcinomas. No correlation was seen between tumor stage and any of the stains examined. Discussion The pathways utilized by lung tumors to evade apoptosis, allowing the tumors continued survival and growth, are incompletely characterized. Equally important are the pathways that become activated in IEN lesions before they become invasive cancer. In this study, we examined the presence and cellular localization of the phosphorylated form of Akt kinase, the transcription factor p65/RELA and the cellular inhibitor of apoptosis protein cIAP-2/BIRC3 in human bronchial IEN lesions. In normal human bronchial epithelium, staining for phosphorylated Akt was present as a faint cytoplasmic stain. Staining intensity increased with increasing pathology grade and was present in a nuclear, perinuclear or plasma membrane pattern. The intensity of p65/RELA staining also increased with increasing pathology grade. Furthermore, more biopsies had nuclear localization of p65/RELA with advancing pathology grade, indicating nuclear translocation of the protein and the potential for transcriptional activation. The apoptosis inhibitor cIAP-2/BIRC3 also had increased staining intensity with increasing pathology grade. These results indicate that Akt, p65/RELA and cIAP-2/BIRC3, components of an anti-apoptotic pathway, are elevated in pre-neoplastic lung lesions. The serine/threonine kinase Akt is a central mediator of anti-apoptotic pathways in eukaryotic cells [2]. Activation of this kinase occurs when it is itself activated by PI3K-dependent protein kinase 1 or 2 (PDK1 or 2). Previous studies have noted an increase in the amount of phosphorylated Akt in a number of different human malignancies including cancer of the lung [4,7,12], head and neck [46], prostate [9], and in multiple myeloma [8,47]. In addition, loss of the tumor suppressor phosphatase and tensin homologue (PTEN), a lipid phosphatase that inhibits PI3K activity, is found in a subset of lung tumors [48-50]. Increased staining for phosphorylated Akt was also frequently observed in human bronchial pre-neoplastic lesions [4,7,12]. In bronchial IEN, increased staining for phospho-Akt correlated with increased staining for phospho-FKHR, a transcription factor that directs expression of anti-apoptotic genes in response to Akt signaling [7]. Our present results agree with previous studies demonstrating an increase in phospho-Akt staining in pre-neoplastic human bronchial lesions. We observed increased staining for the phosphorylated form of Akt beginning as early as mild dysplasia lesions in human bronchial biopsies. The staining for phospho-Akt observed in the biopsies in our study was diverse, consisting of cytoplasmic, nuclear or plasma membrane associated patterns. As PDK must associate with PI3K at the plasma membrane to be activated, localization of Akt has often also been localized to the cytoplasm or plasma membrane [8,9]. However, several targets of Akt are nuclear transcription factors and nuclear localization of Akt has been described in several other human tumor types [7,12,47]. The differences in subcellular localization observed for phospho-Akt in different biopsies may reflect the large number of pathways regulated by this kinase and the different mechanisms by which Akt can mediate its anti-apoptotic effects. The functional significance of the variable localization of phospho-Akt in human lung tumors remains to be determined. The NF-κB transcription factor family is known to direct both apoptotic and anti-apoptotic signaling in eukaryotic cells [51]. In unstimulated cells, NF-κB transcription complexes are held in the cytoplasm by IκB proteins, thus preventing their function as transcriptional activators in the nucleus. Cytokine or growth factor stimulation initiates a signaling cascade that leads to activation of the IκB kinase (IKK) complex that in turn phosphorylates IκB and targets it for ubiquitination and ultimately degradation by the proteasome. Degradation of IκB allows NF-κB to be translocated to the nucleus where it can activate transcription of target genes. Inhibition of NF-κB blocked Ras-mediated transformation of cell lines [24] indicating NF-κB activity is critical for Ras-mediated transformation. The Akt kinase has been reported to activate NF-κB by phosphorylation of IKK [52], causing degradation of IκB and nuclear translocation of NF-κB complexes, and by directly phosphorylating the p65/RELA subunit of NF-κB and increasing its transcriptional activity [27,28]. It should also be noted that Akt can be activated by NF-κB [53], suggesting the regulatory pathways utilized by these molecules is complex. While ample evidence has implicated NF-κB activation as playing an important role in cell transformation, particularly in vitro, there is little data on the presence or localization of NF-κB components in human lung tumors. Chemotherapeutic agents induced NF-κB activity in NSCLC cell lines, increasing the cells resistance to these agents [17] suggesting that in addition to a role in tumor cell evasion of apoptosis, NF-κB may render lung tumor cells more resistant to chemotherapeutic agents. The NF-κB subunit p50 was increased in NSCLC as detected by immunoblotting but no information on localization of this protein within tumors was obtained [29]. To our knowledge, no other study has examined the localization of the p65 subunit of NF-κB in lung IEN lesions or NSCLC. We therefore decided to examine human IEN lesions and lung carcinomas for the presence and localization of the p65/RELA subunit of NF-κB. In our study, nuclear translocation of the p65/RELA subunit of NF-κB was significantly increased in moderate dysplasias compared to lower grade lesions. In the carcinomas, there was a reduction of p65/RELA nuclear positivity similar to the decreased intensity of phospho-Akt staining observed in carcinomas compared to moderate and severe dysplasias. The percentage of cells with positive nuclear staining in carcinomas was also very low compared to moderate and severe dysplasias generally being restricted to isolated foci. In addition to the increased number of biopsies with nuclear staining at higher pathology grades, the overall intensity of cytoplasmic tended to be increased in moderate dysplasias in both the epithelium and stroma compared to normal epithelium (although this difference was not statistically significant in our analysis). In a number of carcinomas, there was intense cytoplasmic staining for p65/RELA with an apparent paucity of staining in the nucleus. While these biopsies were scored negative for nuclear staining of p65/RELA, it was impossible to discount the possibility that nuclear p65/RELA was present in these cells but at levels lower than those observed in the cytoplasm. In the lung adenocarcinoma cell line NCI-H441 we have also observed higher levels of p65 immunoreactivity in the cytoplasm than the nucleus although these cells have very high levels of NF-κB activity as measured by reporter gene assays (JWT and MWA, unpublished data). It should also be noted that activation of NF-κB has been reported in lung cancer cells without an increase in nuclear localization [24,54]. In all lesions with positive nuclear staining, a majority of tumor cells did not contain nuclear p65/RELA staining. The focal nature of nuclear positivity may reflect the induction of this pathway in a subset of cells that has acquired new genetic or epigenetic changes that activate NF-κB and thus may activate downstream anti-apoptotic pathways. Tumor cells acquire the ability to escape apoptosis that is normally initiated in damaged cells. One mechanism for inhibiting apoptosis is to block the activity of the effector caspases that initiate apoptosis by degrading specific cellular targets [55]. The cellular inhibitor of apoptosis (cIAP) family of proteins, also known as the baculoviral IAP repeat containing (BIRC) family, can inhibit apoptosis by binding effector caspases and blocking their proteolytic activity. Several family members have been described including cIAP-1/BIRC2, cIAP-2/BIRC3, XIAP/BIRC4, survivin/BIRC5 and NAIP/BIRC1, and are thought to be important in tumorigenesis [45]. We chose to examine the levels of cIAP-2/BIRC3 in developing lung neoplasia as it is expressed in lung tissue [38] and is a target for NF-κB transcriptional activation [31]. While increased staining for cIAP-2/BIRC3 in human lung adenocarcinomas has been reported [38], levels of cIAP-2/BIRC3 in IEN lesions in lung epithelium have not been reported. Staining for cIAP-2/BIRC3 was increased in the severe dysplasia/CIS category and in carcinomas compared to less advanced pathology grades. The increase in cIAP-2/BIRC3 staining occurred at a higher pathology grade than the grade at which phospho-Akt staining increased (mild/moderate dysplasias) or p65/RELA nuclear localization was increased (moderate dysplasia). This may indicate that inhibition of apoptosis via this protein is not required until later in lung tumorigenesis. While high levels of cIAP-2/BIRC3 expression correlated with nuclear localization of p65/RELA in some biopsies, this was not a consistent observation. The lack of consistent p65/RELA and cIAP-2/BIRC co-expression, in combination with the observation that the overall increase in cIAP-2/BIRC staining occurred at a later pathological stage than p65/RELA nuclear translocation, may indicate that cIAP-2/BIRC3 is not a direct target of p65/RELA in bronchial pre-neoplasia. The possibility remains that p65/RELA transcriptional activation initiates a program that may indirectly lead to cIAP-2/BIRC3 induction later in the progression of pre-neoplastic bronchial lesions. While nuclear staining localization of p65/RELA and increased staining for cIAP-2 did not correlate, we observed a positive correlation between staining intensity for cIAP-2 and phospho-Akt. While it is impossible to state whether this indicates that cIAP-2 is regulated by activation of Akt phosphorylation, this is a possibility. It should be noted, however, that staining intensity for phospho-Akt occurred one pathology grade prior to the observed increase in cIAP-2 staining intensity. Conclusion Increased staining for phospho-Akt and cIAP-2/BIRC3 is present in pre-neoplastic human bronchial biopsy lesions compared to normal human bronchial epithelium. Additionally, the percentage of biopsies containing nuclear p65/RELA staining is increased in pre-neoplastic lesions compared to normal bronchial epithelium. The activation of genes involved in protecting cells from apoptosis in pre-neoplastic lesions suggests that the ability of lung epithelial cells to evade apoptosis is acquired prior to the cells becoming fully transformed. As pre-neoplastic lesions have not acquired all the genetic and epigenetic changes present in lung carcinoma, they may be more amenable to treatment with chemopreventive agents. Thus, identification of the pathways activated in human bronchial pre-neoplastic lesions is important in increasing our understanding of how these lesions develop into lung cancer with the eventual goal of targeting these pathways therapeutically. Competing interests The author(s) declare that they have no competing interests. Authors' contributions JWT reviewed histological staining, participated in design of the study and drafted the manuscript. YZ carried out the immunohistochemical analyses. JCL determined pathology grades of bronchial IEN lesions and reviewed histological staining. PWB determined pathology grades of lung carcinomas and helped draft the manuscript. SL participated in the design of the study, obtained IEN lesions and helped draft the manuscript. MWA conceived of the study, participated in study design and helped draft the manuscript. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements The authors wish to thank C. Ralph Buncher and Padmini Sekar in the Division of Epidemiology and Biostatistics, University of Cincinnati College of Medicine for assistance with statistical analyses. The work was supported by NIH grants CA096964 (JWT, YZ, JCL, SL, MWA) and CA96109 (JCL, SL). ==== Refs Jemal A Tiwari RC Murray T Ghafoor A Samuels A Ward E Feuer EJ Thun MJ Cancer statistics, 2004 CA Cancer J Clin 2004 54 8 29 14974761 Nicholson KM Anderson NG The protein kinase B/Akt signalling pathway in human malignancy Cell Signal 2002 14 381 395 11882383 10.1016/S0898-6568(01)00271-6 Bellacosa A de Feo D Godwin AK Bell DW Cheng JQ Altomare DA Wan M Dubeau L Scambia G Masciullo V Molecular alterations of the AKT2 oncogene in ovarian and breast carcinomas Int J Cancer 1995 64 280 285 7657393 Massion PP Taflan PM Shyr Y Rahman SM Yildiz P Shakthour B Edgerton ME Ninan M Andersen JJ Gonzalez AL Early Involvement of the Phosphatidylinositol 3-kinase/Akt Pathway in Lung Cancer Progression Am J Respir Crit Care Med 2004 170 1088 1094 15317667 10.1164/rccm.200404-487OC Cheng JQ Ruggeri B Klein WM Sonoda G Altomare DA Watson DK Testa JR Amplification of AKT2 in human pancreatic cells and inhibition of AKT2 expression and tumorigenicity by antisense RNA Proc Natl Acad Sci U S A 1996 93 3636 3641 8622988 10.1073/pnas.93.8.3636 Ikebe T Nakayama H Shinohara M Shirasuna K NF-kappaB involvement in tumor-stroma interaction of squamous cell carcinoma Oral Oncol 2004 40 1048 1056 15509497 10.1016/j.oraloncology.2004.05.007 Balsara BR Pei J Mitsuuchi Y Page R Klein-Szanto A Wang H Unger M Testa JR Frequent activation of AKT in non-small cell lung carcinomas and preneoplastic bronchial lesions Carcinogenesis 2004 25 2053 2059 15240509 10.1093/carcin/bgh226 Hsu J Shi Y Krajewski S Renner S Fisher M Reed JC Franke TF Lichtenstein A The AKT kinase is activated in multiple myeloma tumor cells Blood 2001 98 2853 2855 11675360 10.1182/blood.V98.9.2853 Malik SN Brattain M Ghosh PM Troyer DA Prihoda T Bedolla R Kreisberg JI Immunohistochemical demonstration of phospho-Akt in high Gleason grade prostate cancer Clin Cancer Res 2002 8 1168 1171 11948129 West KA Brognard J Clark AS Linnoila IR Yang X Swain SM Harris C Belinsky S Dennis PA Rapid Akt activation by nicotine and a tobacco carcinogen modulates the phenotype of normal human airway epithelial cells J Clin Invest 2003 111 81 90 12511591 10.1172/JCI200316147 Okudela K Hayashi H Ito T Yazawa T Suzuki T Nakane Y Sato H Ishi H KeQin X Masuda A Takahashi T Kitamura H K-ras gene mutation enhances motility of immortalized airway cells and lung adenocarcinoma cells via Akt activation: possible contribution to non-invasive expansion of lung adenocarcinoma Am J Pathol 2004 164 91 100 14695323 Tsao AS McDonnell T Lam S Putnam JB Bekele N Hong WK Kurie JM Increased phospho-AKT (Ser(473)) expression in bronchial dysplasia: implications for lung cancer prevention studies Cancer Epidemiol Biomarkers Prev 2003 12 660 664 12869408 West KA Linnoila IR Belinsky SA Harris CC Dennis PA Tobacco carcinogen-induced cellular transformation increases activation of the phosphatidylinositol 3'-kinase/Akt pathway in vitro and in vivo Cancer Res 2004 64 446 451 14744754 10.1158/0008-5472.CAN-03-3241 Cogswell PC Guttridge DC Funkhouser WK Baldwin ASJ Selective activation of NF-kappa B subunits in human breast cancer: potential roles for NF-kappa B2/p52 and for Bcl-3 Oncogene 2000 19 1123 1131 10713699 10.1038/sj.onc.1203412 Nair A Venkatraman M Maliekal TT Nair B Karunagaran D NF-kappaB is constitutively activated in high-grade squamous intraepithelial lesions and squamous cell carcinomas of the human uterine cervix Oncogene 2003 22 50 58 12527907 10.1038/sj.onc.1206043 Lee CT Seol JY Lee SY Park KH Han SJ Yoo CG Han SK Shim YS Kim YW The effect of adenovirus-IkappaBalpha transduction on the chemosensitivity of lung cancer cell line with resistance to cis-diamminedichloroplatinum(II)(cisplatin) and doxorubicin(adriamycin) Lung Cancer 2003 41 199 206 12871783 10.1016/S0169-5002(03)00227-7 Denlinger CE Rundall BK Keller MD Jones DR Proteasome inhibition sensitizes non-small-cell lung cancer to gemcitabine-induced apoptosis Ann Thorac Surg 2004 78 1207 14; discussion 1207-14 15464472 10.1016/j.athoracsur.2004.04.029 Jones DR Broad RM Madrid LV Baldwin ASJ Mayo MW Inhibition of NF-kappaB sensitizes non-small cell lung cancer cells to chemotherapy-induced apoptosis Ann Thorac Surg 2000 70 930 6; discussion 936-7 11016336 10.1016/S0003-4975(00)01635-0 Osaki S Nakanishi Y Takayama K Pei XH Ueno H Hara N Transfer of IkappaBalpha gene increase the sensitivity of paclitaxel mediated with caspase 3 activation in human lung cancer cell J Exp Clin Cancer Res 2003 22 69 75 12725325 Batra RK Guttridge DC Brenner DA Dubinett SM Baldwin AS Boucher RC IkappaBalpha gene transfer is cytotoxic to squamous-cell lung cancer cells and sensitizes them to tumor necrosis factor-alpha-mediated cell death Am J Respir Cell Mol Biol 1999 21 238 245 10423407 Milligan SA Nopajaroonsri C Inhibition of NF-kappa B with proteasome inhibitors enhances apoptosis in human lung adenocarcinoma cells in vitro Anticancer Res 2001 21 39 44 11299767 Oyaizu H Adachi Y Okumura T Okigaki M Oyaizu N Taketani S Ikebukuro K Fukuhara S Ikehara S Proteasome inhibitor 1 enhances paclitaxel-induced apoptosis in human lung adenocarcinoma cell line Oncol Rep 2001 8 825 829 11410792 Jones DR Broad RM Comeau LD Parsons SJ Mayo MW Inhibition of nuclear factor kappaB chemosensitizes non-small cell lung cancer through cytochrome c release and caspase activation J Thorac Cardiovasc Surg 2002 123 310 317 11828291 10.1067/mtc.2002.118684 Finco TS Westwick JK Norris JL Beg AA Der CJ Baldwin ASJ Oncogenic Ha-Ras-induced signaling activates NF-kappaB transcriptional activity, which is required for cellular transformation J Biol Chem 1997 272 24113 24116 9305854 10.1074/jbc.272.39.24113 Arsura M Mercurio F Oliver AL Thorgeirsson SS Sonenshein GE Role of the IkappaB kinase complex in oncogenic Ras- and Raf-mediated transformation of rat liver epithelial cells Mol Cell Biol 2000 20 5381 5391 10891479 10.1128/MCB.20.15.5381-5391.2000 Jiang Y Cui L Yie TA Rom WN Cheng H Tchou-Wong KM Inhibition of anchorage-independent growth and lung metastasis of A549 lung carcinoma cells by IkappaBbeta Oncogene 2001 20 2254 2263 11402320 10.1038/sj.onc.1204293 Madrid LV Mayo MW Reuther JY Baldwin ASJ Akt stimulates the transactivation potential of the RelA/p65 Subunit of NF-kappa B through utilization of the Ikappa B kinase and activation of the mitogen-activated protein kinase p38 J Biol Chem 2001 276 18934 18940 11259436 10.1074/jbc.M101103200 Madrid LV Wang CY Guttridge DC Schottelius AJ Baldwin ASJ Mayo MW Akt suppresses apoptosis by stimulating the transactivation potential of the RelA/p65 subunit of NF-kappaB Mol Cell Biol 2000 20 1626 1638 10669740 10.1128/MCB.20.5.1626-1638.2000 Mukhopadhyay T Roth JA Maxwell SA Altered expression of the p50 subunit of the NF-kappa B transcription factor complex in non-small cell lung carcinoma Oncogene 1995 11 999 1003 7675461 Wang CY Guttridge DC Mayo MW Baldwin ASJ NF-kappaB induces expression of the Bcl-2 homologue A1/Bfl-1 to preferentially suppress chemotherapy-induced apoptosis Mol Cell Biol 1999 19 5923 5929 10454539 Wang CY Mayo MW Korneluk RG Goeddel DV Baldwin ASJ NF-kappaB antiapoptosis: induction of TRAF1 and TRAF2 and c-IAP1 and c-IAP2 to suppress caspase-8 activation Science 1998 281 1680 1683 9733516 10.1126/science.281.5383.1680 Zong WX Edelstein LC Chen C Bash J Gelinas C The prosurvival Bcl-2 homolog Bfl-1/A1 is a direct transcriptional target of NF-kappaB that blocks TNFalpha-induced apoptosis Genes Dev 1999 13 382 387 10049353 Grad JM Zeng XR Boise LH Regulation of Bcl-xL: a little bit of this and a little bit of STAT Curr Opin Oncol 2000 12 543 549 11085453 10.1097/00001622-200011000-00006 Ekedahl J Joseph B Grigoriev MY Muller M Magnusson C Lewensohn R Zhivotovsky B Expression of inhibitor of apoptosis proteins in small- and non-small-cell lung carcinoma cells Exp Cell Res 2002 279 277 290 12243753 10.1006/excr.2002.5608 Pryhuber GS Huyck HL Staversky RJ Finkelstein JN O'Reilly MA Tumor necrosis factor-alpha-induced lung cell expression of antiapoptotic genes TRAF1 and cIAP2 Am J Respir Cell Mol Biol 2000 22 150 156 10657935 Ferreira CG van der Valk P Span SW Jonker JM Postmus PE Kruyt FA Giaccone G Assessment of IAP (inhibitor of apoptosis) proteins as predictors of response to chemotherapy in advanced non-small-cell lung cancer patients Ann Oncol 2001 12 799 805 11484955 10.1023/A:1011167113067 Dai Z Zhu WG Morrison CD Brena RM Smiraglia DJ Raval A Wu YZ Rush LJ Ross P Molina JR Otterson GA Plass C A comprehensive search for DNA amplification in lung cancer identifies inhibitors of apoptosis cIAP1 and cIAP2 as candidate oncogenes Hum Mol Genet 2003 12 791 801 12651874 10.1093/hmg/ddg083 Hofmann HS Simm A Hammer A Silber RE Bartling B Expression of inhibitors of apoptosis (IAP) proteins in non-small cell human lung cancer J Cancer Res Clin Oncol 2002 128 554 560 12384799 10.1007/s00432-002-0364-z Ambrosini G Adida C Altieri DC A novel anti-apoptosis gene, survivin, expressed in cancer and lymphoma Nat Med 1997 3 917 921 9256286 10.1038/nm0897-917 Lam S leRiche JC Zheng Y Coldman A MacAulay C Hawk E Kelloff G Gazdar AF Sex-related differences in bronchial epithelial changes associated with tobacco smoking J Natl Cancer Inst 1999 91 691 696 10218506 10.1093/jnci/91.8.691 Travis WD Colby TV Corrin B Shimosato Y Brambilla E Histologic and graphical text slides for the histological typing of lung and pleural tumors. World Health Organization Pathology Panel: International Classification of Tumors 1999 3rd Berlin, Springer-Verlag 5 Mountain CF Revisions in the International System for Staging Lung Cancer Chest 1997 111 1710 1717 9187198 Shi SR Key ME Kalra KL Antigen retrieval in formalin-fixed, paraffin-embedded tissues: an enhancement method for immunohistochemical staining based on microwave oven heating of tissue sections J Histochem Cytochem 1991 39 741 748 1709656 Adams JC Heavy metal intensification of DAB-based HRP reaction product J Histochem Cytochem 1981 29 775 7252134 Nachmias B Ashhab Y Ben-Yehuda D The inhibitor of apoptosis protein family (IAPs): an emerging therapeutic target in cancer Semin Cancer Biol 2004 14 231 243 15219616 10.1016/j.semcancer.2004.04.002 Nakayama H Ikebe T Beppu M Shirasuna K High expression levels of nuclear factor kappaB, IkappaB kinase alpha and Akt kinase in squamous cell carcinoma of the oral cavity Cancer 2001 92 3037 3044 11753981 10.1002/1097-0142(20011215)92:12<3037::AID-CNCR10171>3.0.CO;2-# Alkan S Izban KF Immunohistochemical localization of phosphorylated AKT in multiple myeloma Blood 2002 99 2278 2279 11902142 10.1182/blood-2001-01-0317 Soria JC Lee HY Lee JI Wang L Issa JP Kemp BL Liu DD Kurie JM Mao L Khuri FR Lack of PTEN expression in non-small cell lung cancer could be related to promoter methylation Clin Cancer Res 2002 8 1178 1184 12006535 Hosoya Y Gemma A Seike M Kurimoto F Uematsu K Hibino S Yoshimura A Shibuya M Kudoh S Alteration of the PTEN/MMAC1 gene locus in primary lung cancer with distant metastasis Lung Cancer 1999 25 87 93 10470842 10.1016/S0169-5002(99)00052-5 Forgacs E Biesterveld EJ Sekido Y Fong K Muneer S Wistuba II Milchgrub S Brezinschek R Virmani A Gazdar AF Minna JD Mutation analysis of the PTEN/MMAC1 gene in lung cancer Oncogene 1998 17 1557 1565 9794233 10.1038/sj.onc.1202070 Perkins ND NF-kappaB: tumor promoter or suppressor? Trends Cell Biol 2004 14 64 69 15102437 10.1016/j.tcb.2003.12.004 Romashkova JA Makarov SS NF-kappaB is a target of AKT in anti-apoptotic PDGF signalling Nature 1999 401 86 90 10485711 10.1038/43474 Meng F Liu L Chin PC D'Mello SR Akt is a downstream target of NF-kappa B J Biol Chem 2002 277 29674 29680 12052823 10.1074/jbc.M112464200 Norris JL Baldwin ASJ Oncogenic Ras enhances NF-kappaB transcriptional activity through Raf-dependent and Raf-independent mitogen-activated protein kinase signaling pathways J Biol Chem 1999 274 13841 13846 10318790 10.1074/jbc.274.20.13841 Shiozaki EN Shi Y Caspases, IAPs and Smac/DIABLO: mechanisms from structural biology Trends Biochem Sci 2004 29 486 494 15337122 10.1016/j.tibs.2004.07.003	16332260	PMC1325242	CC BY	2021-01-04 16:39:11	no	BMC Cancer. 2005 Dec 6; 5:155	utf-8	BMC Cancer	2,005	10.1186/1471-2407-5-155	oa_comm
==== Front RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-781635954910.1186/1742-4690-2-78EditorialIt's the virus, stupid – part 2 Berkhout Ben [email protected] Department of Human Retrovirology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands2005 16 12 2005 2 78 78 15 12 2005 16 12 2005 Copyright © 2005 Berkhout; licensee BioMed Central Ltd.2005Berkhout; 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 editorial presents Retrovirology's choice for the best basic science "retrovirus paper of the year 2005". ==== Body Chronic HIV-1 infection is characterized by a steady but generally slow loss of CD4+ T cells. A central puzzle of HIV-1 research in the early 90's has been how the virus could cause AIDS, when it infects a trivial number of T cells. It took until 1995 before it was realized that massive virus replication occurs during chronic infection [1]. A model was proposed in which HIV-1 kills healthy cells slightly faster than the human body is able to replenish. David Ho, in a paraphrase of former US president Bill Clinton, said "It's the virus, stupid" to underscore the primary role of the virus in the pathogenesis of AIDS. This finding sets the stage for the implementation of viral load tests in routine diagnostics. The new concept also provided the rationale for attempts to block the furiously replicating virus with antivirals, instead of resolving the disease by modulating the immune system. Nevertheless, the frequency of infected peripheral CD4+ T cells in the chronic phase of HIV-1 infection is too low (0.01 – 1%) to account for the ongoing depletion of these cells by viral infection. In addition, the mechanism for the massive and rapid loss during the acute phase of infection remains unknown. Recent work with HIV-1 and SIV in the macaque model demonstrated that acute infection is accompanied by a dramatic and selective loss of memory CD4+ T cells predominantly from the mucosal surfaces, but the mechanism underlying this depletion was not clarified. Two 2005 papers in Nature clearly repeated the old message "It's the virus, stupid – part 2". I find that two of the best basic science retrovirus papers of 2005 are the works from Joseph Mattapallil and colleagues and Gingsheng Li and colleagues describing that HIV-1 infects and kills the memory CD4+T cells, a T-cell subset responsible for remembering previous infections [2], [3]. This initial assault may determine the outcome of the lengthy battle between SIV-HIV and its host. Mattapallil et al. used a technique that can detect a single copy of SIV DNA to show that 30–60% of all memory CD4+ T cells were infected within 10 days of viral challenge. Most of these cells in the infected rhesus macaque had disappeared 4 days later. There is an exclusive loss of memory CD4+T cells not only from gut-associated mucosal tissues, but also from organized lymph nodes and peripheral blood. Li et al. characterized the activation status of the SIV-producing cells in tissue sections, which turn out to be predominantly resting lymphocytes. This result may be surprising because this virus is known to replicate more efficiently in activated cells, but mucosal lymphocytes are probably better described as "recently activated". Mattapallil suggests that the high rate of infection of these cells is a sufficient mechanism to account for their loss during acute infection; no bystander mechanisms need to be invoked. But Li suggests that indirect mechanisms also contribute to T-cell death in acute infection. Recent studies have provided compelling evidence that the rapid and profound memory T cell depletion is not a peculiarity of the SIV-macaque model, but is also seen in the gut of acutely infected patients. Much remains to be learned on how the substantial depletion of memory CD4+ T cells during acute infection influences the gradual depletion of CD4+ T cells during the chronic phase, and how the removal of CCR5-positive memory cells relates to the coreceptor switch towards CXCR4 that occurs in some patients. The two papers clearly demonstrate that our understanding of the pathogenesis of AIDS has been biased by reliance on examination of peripheral blood, which may provide a limited and perhaps even misleading view. The papers also have implications for antiretroviral treatment and the development of vaccines. For instance, mucosal immunity will be essential for designing an effective AIDS vaccine. Early intervention drug therapy and vaccines should ideally prevent the massive destruction of the CD4+memory T cell compartment. Countermeasures should be developed to attack the virus at or shortly after transmission. The fight with antivirals or microbicides should be targeted in the cervico-vaginal tissues, the predominant portal of viral entry in the HIV-1 pandemic. It seems particularly important that vaccines should block HIV-1 before the initial burst of viral replication in mucosal lymphocytes. The best way to prevent this is by a vaccine that induces potent neutralizing antibodies, but that is easier said than done. The new insights may also be relevant for studies that focus on biological differences between the different HIV-1 subtypes. A major determinant of subtype-specific differences in host cell tropism and possibly pathogenesis is encoded by the long terminal repeat (LTR) promoter of HIV-1 [4]. Mireille Centlivre and colleagues inserted the HIV-1 subtype LTRs in SIV and followed viral dissemination in macaque body compartments [5]. In direct virus competition experiments, subtype C replicated preferentially in the gut-associated lymphoid tissue and subtype B was found predominantly in peripheral blood mononuclear cells during acute infection. The different chemokine/interleukin environments are likely to differentially influence the subtype LTR promoters that encode distinct transcription factor binding sites. For instance, the gut-associated lymphoid tissue is an IL-7 rich environment, which specifically activates subtype C replication through enhancement of LTR transcription [6]. Such differences in the pathogenesis of acute HIV-SIV infection may translate into differential disease progression. Acknowledgements I thank Kuan-Teh Jeang and Mireille Centlivre for reading and commenting on this writing. ==== Refs Ho DD Neumann AU Perelson AS Chen W Leonard JM Markowitz M Rapid turnover of plasma virions and CD4 lymphocytes in HIV-1 infection Nature 1995 373 123 126 7816094 10.1038/373123a0 Mattapallil JJ Douek DC Hill B Nishimura Y Martin M Roederer M Massive infection and loss of memory CD4+ T cells in multiple tissues during acute SIV infection Nature 2005 434 1093 1097 15793563 10.1038/nature03501 Li Q Duan L Estes JD Ma ZM Rourke T Wang Y Reilly C Carlis J Miller CJ Haase AT Peak SIV replication in resting memory CD4+ T cells depletes gut lamina propria CD4+ T cells Nature 2005 434 1148 1152 15793562 Van Opijnen T Jeeninga RE Boerlijst MC Pollakis GP Zetterberg V Salminen M Berkhout B Human immunodeficiency virus type 1 subtypes have a distinct long terminal repeat that determines the replication rate in a host-cell-specific manner J Virol 2004 78 3675 3683 15016888 10.1128/JVI.78.7.3675-3683.2004 Centlivre M Sommer P Michel M Fang RH Gofflo S Valladeau J Schmitt N Thierry F Hurtrel B Wain-Hobson S Sala M HIV-1 clade promoters strongly influence spatial and temporal dynamics of viral replication in vivo J Clin Invest 2005 115 348 358 15690084 10.1172/JCI200522873 Centlivre M Sommer P Michel M Ho Tsong Fang R Gofflo S Valladean J Schmitt N Wain-Hobson S Sala M The HIV-1 clade C promoter is particulary well adapted to replication in the gut in primay infection AIDS 2006	16359549	PMC1325243	CC BY	2021-01-04 16:24:29	no	Retrovirology. 2005 Dec 16; 2:78	utf-8	Retrovirology	2,005	10.1186/1742-4690-2-78	oa_comm
==== Front BMC PsychiatryBMC Psychiatry1471-244XBioMed Central London 1471-244X-5-451633666310.1186/1471-244X-5-45Research ArticleRating of personality disorder features in popular movie characters Hesse Morten [email protected] Sanna [email protected] Rasmus R [email protected] Centre for Alcohol and Drug Research, Aarhus University, Købmagergade 26E, 1150 Copenhagen, Denmark2 Department of Psychology, University of Copenhagen, Øster Farimagsgade 5, 1353 København K2005 8 12 2005 5 45 45 21 6 2005 8 12 2005 Copyright © 2005 Hesse et al; licensee BioMed Central Ltd.2005Hesse 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 Tools for training professionals in rating personality disorders are few. We present one such tool: rating of fictional persons. However, before ratings of fictional persons can be useful, we need to know whether raters get the same results, when rating fictional characters. Method Psychology students at the University of Copenhagen (N = 8) rated four different movie characters from four movies based on three systems: Global rating scales representing each of the 10 personality disorders in the DSM-IV, a criterion list of all criteria for all DSM-IV personality disorders in random order, and the Ten Item Personality Inventory for rating the five-factor model. Agreement was estimated based on intraclass-correlation. Results Agreement for rating scales for personality disorders ranged from 0.04 to 0.54. For personality disorder features based on DSM-IV criteria, agreement ranged from 0.24 to 0.89, and agreement for the five-factor model ranged from 0.05 to 0.88. The largest multivariate effect was observed for criteria count followed by the TIPI, followed by rating scales. Raters experienced personality disorder criteria as the easiest, and global personality disorder scales as the most difficult, but with significant variation between movies. Conclusion Psychology students with limited or no clinical experience can agree well on the personality traits of movie characters based on watching the movie. Rating movie characters may be a way to practice assessment of personality. ==== Body Background Personality disorders and personality traits represent a major challenge to many professionals dealing with psychiatric patients. Personality disorders impact treatment for substance use disorders [1,2], and mood disorders [3-6], bipolar disorder [7], and increase the risk of violence and other crime [8], as well as increase the risk of family conflict [9]. Precise and effective assessment of personality disorder is difficult, and the literature is fraught with research showing the difficulties in assessing personality disorders [10]. Most clinical psychologists make use of clinical observation and deduction based on behaviour in the clinic and clients narratives [11], and some researchers argue for the utility of clinical observations [12], and use clinical observations in studies of the psychometric properties of personality disorders [13,14]. Other researchers argue for the use of self-report instruments such as the Millon Clinical Multiaxial Inventory [15], or semi-structured interviews, such as the Structured Interview for the DSM-III-R Personality disorders [SIDP-R] [16], or the Structured Clinical Interview for the DSM-IV [SCID-II] [7]. Yet other researchers argue for the integration of information from a range of sources as the gold standard of personality disorder research [17,18]. Each of these approaches have their strengths and weaknesses, and may be appropriate in some situations, but not in others, depending on the purpose of assessment, client motivation, time and resources, or scientific questions in a research study [19]. However, in order for personality assessment to be clinically useful, mental health professionals working with patients must be able to understand and identify the various aspects of personality. Reading textbooks and hearing lectures on personality and personality disorder may provide basic knowledge, but future clinicians must also get a more specific idea of what personality assessment is. How do you observe the kinds of behaviour and patterns that are listed as criteria in the DSM-IV or the ICD-10? How do you evaluate such behaviours and integrate it into one or more diagnoses? We suggest that one way to begin to learn personality observation may be through watching fictional movies and rate personality features based on observation of one or more main characters. This approach may be useful when practicing personality assessment in a situation where there are no real patients, such as when training medical students or psychology students. However, because there is no research to support that such observation is reliable, we present inter-rater agreement based on 4 movie characters. Method Procedure Subjects were a convenience sample of psychology students (N = 8) at Copenhagen University who agreed to participate in a study of the inter-rater agreement of personality traits and personality disorders. Four were males, and four females. All had completed obligatory courses in personality psychology, psychiatry and clinical psychology. Subjects came to the Centre for Alcohol and Drug Research on four nights one week apart. Two of the authors (S.S. and R.R.T.) led all participants into a common room, gave instructions, and started the movie. Subjects were instructed to rate the main character, and were informed about which character to focus on, before the movie started. At the end of the movie all participants went to separate rooms to complete the questionnaires about the movie on their own, and afterwards returned the questionnaires to the two authors. After the end of the data collection procedure, the participants were invited to receive feedback about the results of the study. Instruments Subjects rated the movie characters on three different personality measures: global rating scales for personality disorder representing each of the personality disorders listed in the DSM-IV [14,20], a list of the 79 criteria for the personality disorders listed in the DSM-IV in random order, and the Ten Item Personality Inventory [21]. Rating scales The ten rating scales representing personality disorders have previously been used in two studies, one of inter-rater agreement of personality disorders [14], and one of convergent validity of personality disorders [20]. The rating scales range from 0 to 100, with scores from 0–29 representing the absence or very mild forms of the personality disorder, scores from 30–69 representing a moderate degree of the disorder, and scores of 70 or above represent marked presence. Each personality disorder is presented with three keywords as prompts. The instruction was to circle the appropriate number to indicate the degree to which the character was similar to the personality disorder mentioned in each row. The DSM-IV criteria The 79 items of the DSM-IV were listed in random order to avoid halo bias, following the work of Blais and his colleagues [13]. Halo bias denotes the situation, where raters tend to be influenced by previous responses, so that for example if the rater has rated the first criterion for paranoid personality disorder as present, he or she will also be more likely to rate the presence of the second criterion. Each item was rated as (0) absent, [1] mildly or periodically present, or [2] present and causing significant distress. DSM-IV criteria count was calculated as the number of criteria rated as 2. The instructions indicated that subjects must read each sentence, and then circle a number from 0–2 to indicate the degree to which the criterion was true. Two criteria were changed: a brief description was given of the criterion of inadequate affect for schizotypal personality disorder, and a reference in the criterion of self-harm for personality disorder was changed, so that the exclusion of suicide attempts did not refer to another point in the list. The Ten Item Personality Inventory [TIPI] The TIPI is a very brief measure of the Big Five of personality: openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism [21]. Each scale is measured by only to items rated on a 7 point Likert scale. Sample items are "1. Extraverted, enthusiastic. 2. Critical, quarrelsome." In the self-report version, the inventory instructs the respondent to "...write a number next to each statement to indicate the extent to which you agree or disagree with that statement. You should rate the extent to which the pair of traits applies to you, even if one characteristic applies more strongly than the other." This instruction was reworded to match the situation were a movie character was being rated. The test-retest reliability and the convergent validity with other instruments of the TIPI has been reported to be good. The Danish translation was made by two independent Danish translators, and retranslated into English by 2 different persons with English as their first language. Any observed differences were discussed and the final translation based on the feedback from the English-speaking translators. Other information Subjects also reported what courses they had finished in psychology of relevance (e.g., psychiatry, clinical psychology, personality psychology), and whether they had seen or heard about the film they were scheduled to see. Movies Four movies were selected by the first author (M.H.). The movies were selected based on the following criteria: • The personality and inner life of the main character must be important to the plot of the film. • The main character should not undergo a complete transformation of personality (although some development in character was acceptable). • The film should not directly be about psychiatric conditions or substance abuse, or represent a completely one-dimensional picture of a person (e.g., a comedy). • The film should be in English, to allow for other researchers to attempt to replicate the findings. The four movie characters that were selected were (in that order): Sarah Morton in Swimming Pool, directed by Francois Ozon, played by Charlotte Rampling; Aileen Wuornos in Monster, directed by Patty Jenkins, played by Charlize Theron; Suzanne Stone in To Die for, directed by Gus Van Sant, played by Nicole Kidman; and Coleman Silk in The Human Stain, directed by Robert Benton, played by Anthony Hopkins. Statistical analyses Inter-rater agreement was calculated through random effects analysis of variance. Intraclass correlations were calculated as the proportion of variance unique to each movie character relative to the total variance in a given scale. This measure of agreement is equivalent to kappa in interpretation [22]. A limitation to the ICC is that it is highly affected by variance, because if the total variance is small, then the unique variance of each rated target must necessarily be even smaller. This is similar to the way that the kappa statistic is limited by low base-rates when calculating agreement. Analysis of variance was used to assess the multivariate and univariate difference between characters, using the SPSS GLM multivariate ANOVA module. Both the movie character and the rater were entered as factors in the model, and the scales from each instrument were then entered as dependent variables in separate analysis. The interaction between the two was not entered (as that would have resulted in 32 cells with n = 1). Bonferroni adjustments were made for all p-values for ratings to adjust for family-wise type 1 error (with 28 tests of inter-rater reliability, 3 multivariate and 25 univariate, all p-values were multiplied by 28). Differences in the experienced difficulty of rating the characters were also analyzed using analysis of variance. Contrast analysis was reported for linear trend, and Bonferroni post hoc comparisons of the difficulty of the movies. Graphs and partial intraclass correlations were produced with STATISTICA for Windows, V. 6.0 [23], and ANOVA was calculated on SPSS for Windows v. 11.5 [24]. Results Inter-rater agreement The results of the analyses of inter-rater agreement are summarized in table 1, 2, 3. Table 1 ANOVA tables and intra-class correlations for the TIPI Degrees of Freedom F-value P-value Eta2 Intraclass correlation TIPI Multivariate 15,47.33 12.92 0.000 0.777 - Openness 3,21 1.54 0.999 0.181 0.05 Conscientiousness 3,21 22.04 0.000 0.759 0.68 Extraversion 3,21 65.49 0.000 0.903 0.88 Agreeableness 3,21 9.89 0.002 0.585 0.50 Neuroticism 3,21 6.83 0.240 0.494 0.37 Notes: F-values and p-values represent the effect of the character (i.e., the target being rated). TIPI: the Ten Item Personality Inventory [21]. All effects are controlled for between-rater variance. Partial intraclass correlation calculated based on intraclass correlations method 1 [22]. Table 2 ANOVA tables and intra-class correlations for the personality disorder rating scales Degrees of Freedom F-value P-value Eta2 Intraclass correlation Multivariate 30,35.89 4.17 0.001 0.770 - Paranoid 3,28 3.30 0.999 0.306 0.20 Schizoid 3,28 5.73 0.151 0.446 0.36 Schizotypal 3,28 3.94 0.246 0.418 0.28 Antisocial 3,28 10.29 0.001 0.656 0.54 Borderline 3,28 7.55 0.016 0.557 0.46 Histrionic 3,28 10.56 0.006 0.600 0.54 Narcissistic 3,28 8.15 0.008 0.584 0.48 Avoidant 3,28 4.73 0.339 0.399 0.31 Dependent 3,28 1.15 0.999 0.167 0.04 Compulsive 3,21 8.08 0.015 0.561 0.47 Notes: F-values and p-values represent the effect of the character (i.e., the target being rated). Rating scales representing each personality disorder [14]. All effects are controlled for between-rater variance. Partial intraclass correlation calculated based on intraclass correlations method 1 [22]. Table 3 ANOVA tables and intra-class correlations for the personality disorder criteria counts Degrees of Freedom F-value P-value Eta2 Intraclass correlation Multivariate 30,35.90 14.62 0.000 0.920 - Paranoid 3,21 4.08 0.999 0.368 0.24 Schizoid 3,21 6.19 0.999 0.469 0.56 Schizotypal 3,21 9.93 0.069 0.587 0.51 Antisocial 3,21 28.54 0.000 0.803 0.75 Borderline 3,21 19.04 0.006 0.731 0.65 Histrionic 3,21 24.73 0.001 0.779 0.72 Narcissistic 3,21 49.30 0.000 0.876 0.89 Avoidant 3,21 7.58 0.136 0.520 0.41 Dependent 3,21 14.11 0.025 0.668 0.62 Compulsive 3,21 15.11 0.016 0.683 0.63 Notes: F-values and p-values represent the effect of the character (i.e., the target being rated). Criteria rated in random order (after 13). All effects are controlled for between-rater variance. Partial intraclass correlation calculated based on intraclass correlations method 1 [22]. The rating scales for severity of personality disorders resulted in a significant overall model (F(30,35.89) = 4.17, p < 0.001, eta2 = 0.770). The intraclass correlations ranked from poor agreement (all cluster A, and avoidant and dependent personality disorder, ICC ranging from 0.04 to 0.36) to moderate agreement (all cluster B and obsessive-compulsive personality disorder, ICC ranging from 0.46–0.54). The five-factor model likewise resulted in a significant overall model (F(15,47.33) = 12.92, p < 0.001, eta2 = 0.777). The intraclass correlations ranged from no agreement (openness to experience and neuroticism, ICC = 0.05 to 0.37) to excellent agreement (extraversion, ICC = 0.88). Finally, when the DSM-IV criteria were used, the multivariate effect was 0.938 (F(30,32.96) = 17.83, p < 0.001). Criteria count for one personality disorder had poor agreement, paranoid personality disorder (ICC = 0.24). The remaining had either moderate agreement (schizoid, schizotypal, avoidant, dependent, obsessive-compulsive), or excellent agreement (antisocial, borderline, histrionic and narcissistic personality disorder criteria counts). The perceived difficulty of rating We asked about the difficulty of rating each type of instrument for each film. Generally, we expected that the rating would become easier with each film. This was true of the perceived difficulty of the specific criteria for personality disorders, which declined significantly from film to film (estimate of linear trend = -0.53, 95% confidence interval [CI] = -0.95;-0.11). Sarah Morton in "Swimming Pool", the film seen first, was rated more difficult to rate on the criteria than those in the to last films (Suzanne Stone and Coleman Silk). The TIPI, which was designed to be easy to understand and fill in, became more difficult to rate (estimate = 0.56, 95% CI = 0.14;0.91). The TIPI was perceived as easiest to rate for Sarah Morton from Swimming Pool, easier than Suzanne Stone or Aileen Wuornos (p < 0.05), and Suzanne Stone was rated more difficult than Coleman Silk (p < 0.05). Description of the movie characters The descriptions of the movie characters in the following are based on the criteria count and the TIPI (see additional file 1 for descriptive statistics). Note that in terms of diagnoses, personality disorder diagnoses require satisfaction of half the criteria listed or more. The descriptions start with a very brief description of the characters as they are in the movies followed by description of the results of the ratings. Sarah Morton [SM] In the movie, "Swimming Pool", Sarah Morton is a writer of detective novels who is temporarily in a crisis. She agrees to borrow her publishers house in France for the summer to work there. While she lives there, his French-speaking daughter arrives and her promiscuous life-style both shocks and fascinates Sarah. Sarah does not appear to have many close relationships. She was described in the ratings as the highest scorer on paranoid, schizoid, schizotypal, avoidant and obsessive-compulsive personality disorder of all the characters (see figure 2). She is not prototypical for any of these personality disorders, and only her score on schizoid would approach (but not reach) a diagnosis of personality disorder. In terms of the five-factor model, her profile is much more clear-cut: she is the most conscientious of all the characters, by far the least extroverted, quite disagreeable and a medium-high scorer on neuroticism. Had this very introverted person been evaluated for personality disorder diagnosis and the shown symptom counts had been the result, she could either get a diagnosis of personality disorder not otherwise specified, or of not being personality disordered, depending on her clinical state at the time. Aileen Wuornos [AW] In the film "Monster", Aileen is a prostitute who falls in love with a young lesbian woman. Shortly after a man rapes her and tries to kill her, but she succeeds in killing him instead, and after that starts to kill men whom she contacts as a prostitute. Please note that the Aileen Wuornos described in this paragraph is the Aileen of the film as seen by the raters in this study – not the real character. AW was perceived as a person with co-morbid borderline and antisocial personality disorder. Her scores on other personality disorders are well below diagnostic cut-offs. In terms of the five-factor model, she is less conscientious than the others, and a medium-high scorer on neuroticism. She would clearly be diagnosed with co-morbid borderline and antisocial personality disorder, had this been a diagnostic evaluation (indeed, the real Aileen Wuornos was diagnosed with borderline and antisocial personality disorder by clinicians who saw her) [25]. Suzanne Stone [SS] Suzanne Stone in the film "To Die For" is a young woman who wants to be on television at any cost. She marries a young man, but soon begins to have affairs with TV producers to accomplish her main goal: to become a news-reporter at a major TV station. When her husband tries to persuade her to settle down and have children, she decides to have him killed instead, taking advantage of three troubled youths, whom she has met while trying to make a TV production. SS was seen as a prototypical narcissistic person by the raters: on average, she satisfied 8 of 9 criteria for narcissistic personality disorder, some histrionic personality disorder criteria, and relatively few others. In terms of the five-factor model, she is as open to experience as the others, as conscientious as the others (except for AW), as extraverted (except for SM), as disagreeable (except CS), and a low-scorer on neuroticism. Had she been evaluated for personality disorders, she would receive a diagnosis of narcissistic personality disorder. Coleman Silk [CS] Coleman Silk in the movie "The Human Stain" is a middle-aged university professor who gets fired from his job under suspicion of racism against two African-American students, whom he has never met (and thus, he does not know the colour of their skin). The loss of his jobs leads to a chain of events that reveals much about his difficult life as well as lead him into new experiences. Among other things, it turns out that he has grown up in a family of African-Americans, and has kept this fact a secret for various reasons. CS was rated as the least disordered of all the characters, scoring very low on all personality disorder scales. In terms of the five-factor model, he was seen as more agreeable than the others, and the lowest scorer on neuroticism. Had he been evaluated for personality disorders, he would not be diagnosed with a personality disorder. Discussion The findings overall indicate that various personality features of movie characters can be rated reliably by relatively untrained raters with only very basic knowledge about the nature of personality and personality disorders. It was found that the perceived difficulty of rating personality disorder criteria declined somewhat over the movies. The use of specific criteria resulted in a better discrimination than the use of global rating scales, but even with global rating scales, agreement could be detected beyond chance, accounting for more the three-quarters of the observed variance. In comparison with real-life patients, all raters were able to observe exactly the same behaviour for each target, yet the amount of situations and contexts that raters could observe were far more varied than would usually be available. In real-life settings, various clinicians will often have to observe patients within a single limited setting (e.g., counselling, therapy, group therapy, milieu therapy), and that setting may not even be the same for different clinicians (e.g., observing the patient on different days). However, rating movie characters also differ from another common situation, in which a co-interviewer rates criteria based on a semi-structured clinical interview. In that situation, inter-rater agreement is expectedly much higher than what was observed in this study. Thus, a strength of this study is to show that the inference of personality disorder traits can be done, even in the absence of clear questions about the specific criteria. The rating of movie characters may serve as illustrations of many points in the assessment of personality disorders. First of all, raters get the point that some people, whilst clearly disturbed and stressed as a result of life-long patterns of maladaptive behaviour, such as Aileen Wuornos and Sarah Morton, do not fit any single prototypical personality disorder profile [26-28]. Thirdly, raters get an impression of the relative precision (and lack thereof) of ratings of personality. Some students may perceive personality disorder diagnoses as nearly arbitrary or prejudice-driven labels, and the experience that agreement can be achieved may help them understand that there is something more than a label to personality disorders. Others with an unrealistic faith in the diagnostic system may experience that rating personality disorders is more difficult, and that although agreement is substantial, sometimes behaviours and reactions in the same person is experienced differently by different raters, even when using the same diagnostic system to rate the behaviours. Several limitations to this study must be acknowledged. The four movies selected did not represent a wide range of different personality traits and personality disorders. For instance, the variance of the TIPI Openness to Experience scale is nearly nil, and there are no high scorers on avoidant or dependent personality disorder, either by the rating scales or by the criteria. Therefore, the level of agreement that could be reached for these traits and disorders is limited by the limited range they represent. Thus, although it is tempting to suggest that differences in agreement between features are due to difference between the degree to which behaviours are easily observable (e.g., conscientiousness, extraversion, Cluster B personality disorder feature), an equally justifiable interpretation is that the amount of variance for some personality traits (e.g., openness to experience, paranoid, schizoid, avoidant and dependent personality disorder) was simply too small to assure reasonable agreement. Also, the difficulty in rating various areas must be interpreted with caution, given that all raters saw the films in the same order. Had the order been varied across raters, changes in perceived difficulty over time would have been much more reliable. Thirdly, we do not know whether these raters were better or worse than the average psychology student at rating personality features. What we do know is that they did not consider themselves experts. The absence of a gold standard for the ratings makes it difficult to conclude anything about the movie characters, beyond simply stating that 8 independent raters reached similar results when rating these 4 movies. Another limitation is the use of relatively untrained psychology students. With regard to agreement, however, it seems likely that experts could do little better in terms of cluster B personality disorder features, where it is unlikely that agreement can be better than ICC ranging from 0.46–0.54 for global rating scales, and ICC ranging from 0.75–0.89 for criteria counts. A next step in evaluating the use of fiction as a tool for practicing assessment of personality disorder would be to conduct a study of the inter-rater agreement on real patients before and after practice with fiction, and identify factors in movies that foster the learning experience of rating personality features in movies. For instance, is it more helpful to assess ambiguous characters, easily rated characters, or a mixture of ambiguous or easily rated characters? Conclusion Raters converge on their rating of personality traits in movie characters. Fiction movies may be useful for training observers in recognizing personality pathology and personality traits. Table 4 Perceived difficulty by movie character PD scales PD criteria TIPI Mean Standard deviation Mean Standard deviation Mean Standard deviation SM 3.10 0.74 3.00 0.67 1.50 1.08 AW 2.56 0.88 2.67 0.50 2.56 1.13 SS 2.80 0.92 2.00 0.82 2.80 0.79 CS 2.50 1.08 2.40 0.97 2.30 0.95 Total 2.74 0.91 2.51 0.82 2.28 1.07 Bonferroni Post hoc SR>SS SR<SS,AW Notes: PD: Personality disorder. Difficulty scores could ranged 0–4. SM: Sarah Morton. AW: Aileen Wuornos. SS: Suzanne Stone. CS: Coleman Silk Pre-publication history The pre-publication history for this paper can be accessed here: Supplementary Material Additional File 1 Descriptive statistics for the 4 characters. Click here for file ==== Refs Kosten TA Kosten TR Rounsaville BJ Personality disorders in opiate addicts show prognostic specificity Journal of Substance Abuse Treatment 1989 6 163 168 2795705 10.1016/0740-5472(89)90003-2 Darke S Williamson A Ross J Teeson M Lynskey M Borderline personality disorder, antisocial personality disorder and risk-taking among heroin users: findings from the Australian Treatment Outcome Study (ATOS) Drug Alcohol Depend 2004 74 77 83 15072810 10.1016/j.drugalcdep.2003.12.002 Grilo CM Sanislow CA Shea MT Skodol AE Stout RL Gunderson JG Yen S Bender DS Pagano ME Zanarini MC Morey LC McGlashan TH Two-year prospective naturalistic study of remission from major depressive disorder as a function of personality disorder comorbidity J Consult Clin Psychol 2005 73 78 85 15709834 10.1037/0022-006X.73.1.78 Gunderson JG Morey LC Stout RL Skodol AE Shea MT McGlashan TH Zanarini MC Grilo CM Sanislow CA Yen S Daversa MT Bender DS Major depressive disorder and borderline personality disorder revisited: longitudinal interactions Journal of Clinical Psychiatry 2004 65 1049 1056 15323588 Hayden EP Klein DN Outcome of dysthymic disorder at 5-year follow-up: the effect of familial psychopathology, early adversity, personality, comorbidity, and chronic stress American Journal of Psychiatry 2001 158 1864 1870 11691693 10.1176/appi.ajp.158.11.1864 Black DW Goldstein RB Nasrallah A Winokur G The prediction of recovery using a multivariate model in 1471 depressed inpatients European Archives of Psychiatry and Clinical Neuroscience 1991 241 41 45 1832304 10.1007/BF02193753 Bieling PJ MacQueen GM Marriot MJ Robb JC Begin H Joffe RT Young LT Longitudinal outcome in patients with bipolar disorder assessed by life-charting is influenced by DSM-IV personality disorder symptoms Bipolar Disorders 2003 5 14 21 12656933 10.1034/j.1399-5618.2003.00014.x Hiscoke UL Langstrom N Ottosson H Grann M Self-reported personality traits and disorders (DSM-IV) and risk of criminal recidivism: a prospective study. Journal of Personality Disorder 2003 17 293 305 10.1521/pedi.17.4.293.23966 Johnson JG Chen H Cohen P Personality Disorder Traits During Adolescence and Relationships With Family Members During the Transition to Adulthood J Consult and Clin Psychol 2004 72 923 932 15612840 10.1037/0022-006X.72.6.923 Clark LA Livesley WJ Morey L Personality disorder assessment: the challenge of construct validity Journal of Personality Disorder 1997 11 205 231 Westen D Divergences between clinical and research methods for assessing personality disorders: implications for research and the evolution of axis II. American Journal of Psychiatry 1997 154 895 903 9210738 Westen D Shedler J Revising and assessing axis II, Part I: developing a clinically and empirically valid assessment method Am J Psychiatry 1999 156 258 272 9989563 Blais MA Norman DK A psychometric evaluation of the DSM-IV personality disorder criteria Journal of Personality Disorder 1997 11 168 176 Hesse M Social workers' ratings of comorbid personality disorders in substance abusers Addictive Behaviors 2005 30 1241 1246 15925133 10.1016/j.addbeh.2004.12.002 Craig RJ Prevalence of Personality Disorders among Cocaine and Heroin Addicts Substance Abuse 2000 21 87 94 12466649 10.1023/A:1007711112790 Cacciola JS Rutherford MJ Alterman AI McKay JR Mulvaney FD Long-term test-retest reliability of personality disorder diagnoses in opiate dependent patients Journal of Personality Disorder 1998 12 332 337 Kranzler HR Tennen H Babor TF Kadden RM Rounsaville BJ Validity of the longitudinal, expert, all data procedure for psychiatric diagnosis in patients with psychoactive substance use disorders Drug Alcohol Depend 1997 45 93 104 9179511 9179511 10.1016/S0376-8716(97)01349-5 Pilkonis PA Heape CL Ruddy J Serrao P Validity in the diagnosis of personality disorders: the use of the LEAD standard Psychological Assessment 1991 3 46 54 10.1037/1040-3590.3.1.46 Rogers R Standardizing DSM-IV Diagnoses: The Clinical Applications of Structured Interviews Journal of Personality Assessment 2003 81 220 225 14638446 10.1207/S15327752JPA8103_04 Hesse M Personality disorders in substance abusers: Validation of the DIP-Q through principal components factor analysis and canonical correlation analysis BMC Psychiatry 2005 5 24 15910688 10.1186/1471-244X-5-24 Gosling SD Rentfrow PJ Swann Jr. WB A very brief measure of the Big-Five personality domains Journal of Research in Personality 2003 37 504 528 10.1016/S0092-6566(03)00046-1 Shrout PE Spitzer RL Fleiss JL Quantification of agreement in psychiatric diagnosis revisited Archives of General Psychiatry 1987 44 172 177 3813814 StatSoft STATISTICA (data analysis software system) 2003 6 , StatSoft SPSS SPSS for Windows 2002 11.5.1 Chicago, SPSS Inc. Myers WC Gooch E Meloy JR The role of psychopathy and sexuality in a female serial killer J Forensic Sci 2005 50 652 657 15932102 Shedler J Westen D Refining Personality Disorder Diagnosis: Integrating Science and Practice American Journal of Psychiatry 2004 161 1350 1365 15285958 10.1176/appi.ajp.161.8.1350 Fossati A Maffei C Bagnato M Battaglia M Donati D Donini M Fiorilli M Novella L Prolo F Patterns of covariation of DSM-IV personality disorders in a mixed psychiatric sample Comprehensive Psychiatry 2000 41 206 215 10834630 10.1016/S0010-440X(00)90049-X Oldham JM Skodol AE Kellman HD Hyler SE Rosnick L Davies M Diagnosis of DSM-III-R personality disorders by two structured interviews: patterns of comorbidity Am J Psychiatry 1992 149 213 220 1734742	16336663	PMC1325244	CC BY	2021-01-04 16:33:02	no	BMC Psychiatry. 2005 Dec 8; 5:45	utf-8	BMC Psychiatry	2,005	10.1186/1471-244X-5-45	oa_comm
==== Front BMC Health Serv ResBMC Health Services Research1472-6963BioMed Central London 1472-6963-5-751631863810.1186/1472-6963-5-75Research ArticleConsumer involvement in Quality Use of Medicines (QUM) projects – lessons from Australia Kirkpatrick Carl MJ [email protected] Elizabeth E [email protected] Gregory R [email protected] Susan E [email protected] School of Pharmacy, University of Queensland, Brisbane, Australia2 Quality Use of Medicines and Pharmacy Research Centre, School of Pharmacy and Medical Sciences, University of South Australia, Adelaide, Australia2005 1 12 2005 5 75 75 1 6 2005 1 12 2005 Copyright © 2005 Kirkpatrick et al; licensee BioMed Central Ltd.2005Kirkpatrick et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background It is essential that knowledge gained through health services research is collated and made available for evaluation, for policy purposes and to enable collaboration between people working in similar areas (capacity building). The Australian Quality Use of Medicine (QUM) on-line, web-based project database, known as the QUMmap, was designed to meet these needs for a specific sub-section of health services research related to improving the use of medicines. Australia's National Strategy for Quality Use of Medicines identifies the primacy of consumers as a major principle for quality use of medicines, and aims to support consumer led research. The aim of this study was to determine how consumers as a group have been represented in QUM projects in Australia. A secondary aim was to investigate how the projects with consumer involvement fit into Australia's QUM policy framework. Method Using the web-based QUMmap, all projects which claimed consumer involvement were identified and stratified into four categories, projects undertaken by; (a) consumers for consumers, (b) health professionals for consumers, (c) health professionals for health professionals, and (d) other. Projects in the first two categories were then classified according to the policy 'building blocks' considered necessary to achieve QUM. Results Of the 143 'consumer' projects identified, the majority stated to be 'for consumers' were either actually by health professionals for health professionals (c) or by health professionals for consumers (b) (47% and 40% respectively). Only 12 projects (9%) were directly undertaken by consumers or consumer groups for consumers (a). The majority of the health professionals for consumers (b) projects were directed at the provision of services and interventions, but were not focusing on the education, training or skill development of consumers. Conclusion Health services research relating to QUM is active in Australia and the projects are collated and searchable on the web-based interactive QUMmap. Healthcare professionals appear to be dominating nominally 'consumer focussed' research, with less than half of these projects actively involving the consumers or directly benefiting consumers. The QUMmap provides a valuable tool for policy analysis and for provision of future directions through identification of QUM initiatives. ==== Body Background Many of the findings of health services research are never published in peer-reviewed journals nor presented at a conference [1,2]. There may be many reasons for this, including that such projects are often small, the researchers often hold clinical or service positions and their research time is minimal or non-existent, or the results did not eventuate 'as intended' [1]. A lot of potential 'lessons learned' are lost to future researchers, there is a risk that an unsuccessful study may be replicated, the researchers themselves lose an opportunity to make contact with other researchers with similar interests or similar studies, and input into policy direction or development is lost [2]. Opportunities for capacity building by teams linking up together are also lost. As discussed below, the quality use of medicines (QUM) on-line project database, known as the QUMmap [3], seeks to address some of these issues and provide decision makers with an interactive, accessible web-based database to search for previous and on-going QUM initiatives. Initiatives to improve use of medicines are being implemented around the world in an effort to maximise health outcomes, reduce adverse events and keep health costs within affordable limits. While prescribing behaviour is one factor that can be targeted to improve use of medicines, it is widely acknowledged that consumer behaviour also influences medication use and that involvement of consumers in strategies to improve use of medicines is necessary in any country's attempts to promote rational drug use [4]. Australia has a comprehensive strategy for promoting rational drug use, known as the National Strategy for QUM [5]. A key principle of the strategy is the primacy of consumers in any initiative to promote quality use of medicines. Another key principle is that multidisciplinary, collaborative approaches are necessary to improve use of medicines and that these approaches should involve all stakeholder groups (ie. doctors, nurses, pharmacists, consumers etc) from the beginning of an initiative's development. This has demonstrated effectiveness particularly in the modification of drug use. [6,7] A WHO publication on developing and implementing national drug policies stated the importance of consumers in improving use of medicines [4]. To assist in evaluating the implementation of the National Strategy for Quality Use of Medicines, the QUM Mapping Project was commissioned by the Australian Department of Health and Ageing, in conjunction with the Pharmaceutical Health and Rational use of Medicines (PHARM) committee [2]. The aim was to set up a database of QUM projects and initiatives in Australia. The mapping project website has been on-line since February 1999, and has collected over 1000 reports of QUM projects in Australia. The approach to improving medicine use in Australia is often referred to as a systems approach, indicating that behaviours to support quality use of medicines must be developed, while at the same time a supportive environment must be created. In developing appropriate behaviours, there is a need to implement initiatives that raise awareness of quality use of medicines as an issue, that develop appropriate skills and knowledge, and that reinforce and maintain appropriate behaviours [8]. This is conceptualised as Figure 1. The National Strategy identifies six 'building blocks' as essential components to achieve optimal quality of medicine use, raising levels of awareness and action across all sectors; Figure 1 Levels of the QUM pyramid – the faces of the three dimensional pyramid representing all the partners required to achieve optimal quality use of medicines (consumers, government, health professionals, industry). Adapted from the National Strategy for Quality Use of Medicines – Plain English Edition [8] • Policy development and implementation; • Facilitation and co-ordination of quality use of medicines initiatives; • Provision of objective information and assurance of ethical promotion of medicines; • Education and training; • Provision of services and appropriate interventions; and • Strategic research, evaluation and routine data collection. This strategy also has international relevance, as the health care systems of many countries strive to improve the use of medications and to increase consumer involvement in this process [9-11]. The consumer movement in Australia was a driving force behind the establishment of the National Strategy for Quality Use of Medicines and Australia's National Medicines Policy. In 1988, the Consumers Health Forum of Australia circulated a discussion paper, entitled "Developing a rational medicinal drug policy for Australia – What does it mean?" [12]. This document was published in its newsletter and distributed to Commonwealth, State and Territory governments, and to the organisations which represented pharmacists, medical practitioners, the pharmaceutical industry, clinical pharmacologists, alcohol and drug societies and patient support groups. In 1989, the Consumers Health Forum of Australia moved the process forward again when they produced a report entitled "Towards a National Medicinal Drug Policy" [13]. This identified a number of components considered, from a consumer perspective, to be essential within a national medicines policy. The lack of any overall policy on quality medication use was noted. In addition, maintaining services developed by and for consumers was seen as an important component of any national strategy. The implementation of Australia's National Strategy for Quality Use of Medicines, which began in 1992, was supported with approximately $2 million annually in research funding. This research funding was available to all groups and aimed to ensure multidisciplinary research, which included consumer involvement from project conceptualisation. There were also efforts to support consumer led research so that services and resources appropriate to consumer needs for improving use of medicines were developed. More recently, Australia's National Health and Medical Research Council have published a statement on consumer and community participation in health and medical research [14]. Consumer involvement in research has also been investigated using surveys and questionnaires to determine the nature and extent of their involvement in funding bodies and in randomised clinical trials [15,16]. Objectives The main aim of this study was to evaluate how consumers, as a group, have been represented in the QUM projects in Australia. A secondary aim was to investigate how these projects relate to the QUM policy framework, in particular, to the six key QUM building blocks defined by Australia's National Strategy for QUM. Methods Using the standard SQL search engine available on the QUMmap [3], all projects were identified which either claimed that "consumers" were the target group (this is a tick box available as one choice when projects are entered on the database) or included a text word "consumer". The term "consumer" is widely used in reference to healthcare in Australia and has a broad definition to include any user or potential user of health care [17]. The QUMmap has this group noted as a specific tick box target group. The search was undertaken to be as inclusive as possible of all research involving consumers in any way, for consultation, for collaboration or as initiators of the research. The database search was undertaken for the period February 1999 and December 2004 using the publicly accessible searching facilities provided on the website [3]. All identified projects were evaluated and sorted into four groups by CK, in consultation with ER and ST, using definitions as follows: (a) Consumers for consumers – any consumer or consumer group undertaking research/initiatives directly for consumers; (b) Health professionals for consumers – health professionals (doctors, nurses, pharmacists etc) undertaking research/initiatives which directly involve consumers as part of the initiative (eg health professionals working with consumer groups to provide educational material); (c) Health professionals for health professionals – a QUM research project/initiative undertaken by health professionals to improve health professional services which may indirectly provide assistance to the consumer. (d) Other – does not fit into any of these categories, although consumers had been nominated as a target group or the term consumer was included in the project description. Categories (c) and (d) were discarded from further analysis because they did not involve consumers directly in QUM activities or research. This left the consumer for consumer for consumer (n = 12) (category (a)) and health professional for consumer (n = 69) (category (b)) projects (n = 81, total). These remaining projects were then allocated by CK, in consultation with ER and ST, to one of the three levels of the QUM pyramid (Figure 1) [8] and analysed to see which level the projects were achieving in the QUM framework. They were also allocated to one of the six key building blocks of Australia's QUM strategy [8]. A Kruskal-Wallis test was used (difference considered significant if P < 0.05) to investigate differences between project costs in the categories, where these were recorded. Results In the period from February 1999 to December 2004, 143 projects on the QUMmap [3] were identified by the search strategy. However, the majority of projects stated to be for consumers, were either undertaken by health professionals for health professionals (category (c)) or health professionals for consumers (category (b)). Only 12 projects could be identified as being directly undertaken by consumers or consumer groups specifically for consumers (category (a)) (Table 1). Table 1 Categories of projects on the QUMmap identifying consumers as their target, or having consumer as a searchable term Category Number Percent Health professionals for consumers (b) 69 47 Health professionals for health professionals (c) 57 40 Consumers for consumers (a) 12 9 Other (d) 5 4 The geographical distribution of projects sorted by categories (a), (b) and (c) is presented in Table 2. The majority of projects in the consumers for consumers (a) category were undertaken in the state of New South Wales (8/12), Australia's most populous state where many head offices of consumer organisations are found. Five projects were undertaken by the Australian Pensioners' & Superannuants' Federation, one by the Council of the Aging, and one by the Combined Pensioners and Superannuants Association. Table 2 State distribution of selected categories of project on the QUM Map Number (%) State Consumers for consumers (a) Health professionals for consumers (b) Health professionals for health professionals (c) NSW 8 (67) 17 (25) 18 (32) SA 1 (8) 17 (25) 15 (26) VIC 2 (17) 12 (17) 7 (12) ACT - 5 (7) 4 (7) QLD - 8 (12) 9 (16) WA - 7 (10) 3 (5) TAS - 2 (3) 1 (2) NT 1 (8) 1 (1) - The median cost of the 9 consumers for consumers (category (a)) projects for which funding data were available was AUD $59,000 (range 26,000 – 240,000). This was similar to the median cost of $57,000 for health professionals for consumers (category (b)) projects (n = 45 with funding data available; range $400 – $668,000). The median cost for health professionals for health professionals (category (c)) projects was AUD $83,000 (n = 27 with funding data available, range $5,000 – $392,000). These were not statistically significantly different (P = 0.18). The QUM pyramid levels for the health professionals for consumers (category (b)) and the consumers for consumers (category (a)) projects are presented in Table 3. Most of the consumers for consumers (category (a)) projects were about informing or educating consumers about QUM, with a small number of projects introducing skill levels to make appropriate QUM decisions. In comparison, the health professionals for consumers (category (b)) projects covered all levels of the QUM pyramid. It is interesting to note that more of the health professionals for consumers (category (b)) projects were located in the upper part of the pyramid, with a greater focus on action and evaluation. Table 3 QUM pyramid level as defined in Australia's National Strategy for Quality Use of Medicines (QUM) for selected projects on the QUMmap QUM Pyramid level Raising awareness Knowledge and skill development Reinforcement and maintenance Consumers for consumers (a) 8 4 0 Health professionals for consumers (b) 24 20 25 Analysis of the projects according to the QUM building block classification demonstrated that the majority of health professionals for consumers (category (b)) projects were about the provision of services and interventions. The health professionals for consumers (category (b)) projects did not focus on education and training or building skill development in consumers. Discussion This paper audits the consumer role in QUM projects listed on the Australian QUM Map, asking the question, are consumers adequately involved in the QUM activities in Australia? One hundred and forty three projects that claimed to involve "consumers" were identified using the searches readily available on the QUM map website [3]. Only 81 (57%) of these projects involved consumers either via consumer groups, or with some interaction with healthcare professionals to provide direct benefit to the consumer. Furthermore, only 12 (9%) were projects undertaken by consumers for consumers. The majority of the consumers for consumers projects have been undertaken in New South Wales, with Victoria being the next most prominent state. New South Wales and Victoria are Australia's most populated states. A number of health professional projects (n = 69) are involving consumers (health professionals for consumers (category (b)) projects), with good distribution among all states of Australia. All projects that receive funding by the major supporting bodies for QUM in Australia are currently included on the QUMmap (more details of the projects included and the policy potential for the map have been published previously [2]). The data currently on the QUMmap can be considered to be a sample of projects undertaken in the QUM area in Australia, with the current audit showing a "snap-shot" of the spread of activity. It would appear that health professionals for health professionals projects are receiving greater median funding per project than those directly involving consumers ($83,000 versus $57,000 median funding for health professionals for consumers or versus $59,000 median for consumers for consumers), although these differences were not statistically significant (P = 0.18, comparison of the three categories). Perhaps the level of expertise associated with grant applications by health professionals contrasted to those made by consumers or consumer groups. It could also be that project budgets are formulated differently, depending upon which partners are involved. Evaluation using the levels of the QUM pyramid of the consumers for consumers (category (a)) and health professionals for consumers (category (b)) projects has shown some interesting trends. The consumers for consumers projects are mostly based at the bottom of the pyramid (information/awareness) with some in the middle sections endeavouring to increase awareness and prepare consumers to take action on QUM. In contrast, the health professionals for consumers projects are in the top two thirds of the pyramid – ready to take action and actively monitoring adverse and positive effects of QUM. In addition consumer projects, including those undertaken by health professionals are narrowly focused within the key building blocks, directed mainly to the provision of services and interventions. More activity across all the essential domains of Australia's QUM National Strategy is clearly required. Consumers (or consumer groups) need to be encouraged via financial, infrastructure or other means, to develop QUM projects that move into the upper echelon of the QUM pyramid and cover all key building blocks. It is important that consumers are directly involved in all aspects of the QUM process in order to achieve successful behaviour change. It is well recognised that health professionals, by not directly involving consumers/consumer groups may be perceived as "paternalistic", the outcome being that consumers are resistant to being peripherally involved in activities or are resistant to changing their use of medicines [18-21]. Indeed, it has been shown that involvement of consumer and community groups early in projects improves uptake by consumers, and this has been demonstrated particularly in projects involving the use of sunscreens, cancer screening programs, smoking cessation programs, and in the treatment of asthma [21-26]. Moreover, we found no projects designed and implemented by consumers for health care professionals. Although other countries do not have specific QUM strategies or even defined National Medicines Policies including QUM, the messages presented here for Australia can be taken and developed in other health care systems internationally. Increasing consumer led initiatives is important in all countries [15,16]. The projects audited by this study are available for international perusal on the QUMmap and the QUMmap, or a similar database would be a valuable addition to the health services research effort of other countries. Canada has recently developed a similar database [27]. The major limitation of this present audit is that the QUMmap may not capture all QUM research and projects implemented in Australia. However it does include all those that are funded by General Practice Education Programme (GPEP), QUM Evaluation Programme (QUMEP), National Health and Medical Research Council (NHMRC), Pharmacy Government Agreements, the National Prescribing Service (NPS) and the Safety and Quality Council. It has also achieved wide publicity such that many projects are entered by the investigators themselves [2]. Although this website has been advertised widely through academic, professional and other interested bodies, it is possible that some consumer based QUM projects have not been entered on to the database and therefore have not been included in this audit. The significance of the lack of mandatory reporting of QUM projects on the QUMmap has been highlighted previously [1]. Costings were included only for some projects, which limits the interpretation of the findings and statistical analyses of median funding for the different categories of projects. The projects were classified by CK, in consultation with ER and ST as consensus classifications. Although independent ratings and reliability analyses were not conducted, we consider that the classification system was clear and that misclassifications were unlikely. Conclusion This audit of the QUMmap would indicate that consumers as a group are not well represented in QUM research in Australia, with some states having no consumer driven projects on the QUMmap. Healthcare professionals appear to be dominating consumer based research, with only a little over half of these projects actively involving the consumers or directly benefiting consumers. On this basis, consumers (or consumer groups) need to be encouraged via financial, infrastructure or other means, to develop or be directly involved in the QUM projects in all states of Australia. It is likely that similar conclusions could apply to other health care systems internationally. Competing interests The author(s) declare that they have no competing interests. Authors' contributions All four authors participated in the design of the study and assisted in data analysis and writing the manuscript. CK carried out the database searches and the collation of the results. ER and ST initially conceived the study. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements The Department of Health and Ageing, Canberra, Australia provided funding for the development and maintenance of the QUMmap. ==== Refs Dooley MJ Lu PL Hospital pharmacists fail to record their research in the quality use of medicines mapping project Journal of Pharmacy Practice and Research 2002 32 141 142 Roughead EE Monteith GR Harvey KJ Tett SE Evaluating Australia's National Medicines Policy using geographical mapping Intern Med J 2002 32 66 71 11885845 10.1046/j.1445-5994.2002.00166.x QUMmap http://www.qummap.health.gov.au World Health Organisation How to develop and implement a national drug policy. Second Edition. 2001 Geneva, World Health Organisation Australian Department of Health and Ageing www.nmp.health.gov.au, download available through publications link Bochner F Burgess NG Martin ED Approaches to rationing drugs in hospitals. An Australian perspective Pharmacoeconomics 1996 10 467 474 10163628 Pearce MJ Begg EJ Encouraging consensus cost effective drug therapy: five years experience with a hospital drug utilisation review programme N Z Med J 1997 110 92 95 9137310 Commonwealth Department of Health and Aging The National Strategy for Quality Use of Medicines (Plain English Edition) 2002 First , Commonwealth Department of Health and Aging, Commonwealth of Australia 1 36 Sketris IS Brown MG Murphy AL Policy choices for Pharmacare: The need to examine benefit design, medication management strategies and evaluation HealthcarePapers 2004 4 36 45 15114068 Romanow RJ Building on values: The future of health care in Canada 2002 Ottawa, Canada, Commission on the future of health care in Canada 11861605 Donelan K Blendon RJ Schoen C Binns K Osborn R Davis K The elderly in five nations: the importance of universal coverage Health Aff (Millwood) 2000 19 226 235 10812802 10.1377/hlthaff.19.3.226 Consumers Health Forum of Australia Developing a rational medicinal drug policy for Australia - what does it mean? Health Forum 1988 4 13 Consumers Health Forum of Australia Towards a National Medicinal Drug Policy 1989 Melbourne, NHMRC http://www.nhmrc.gov.au/publications/synopses/r22syn.htm , accessed 24th Aug 2005 O'Donnell M Entwistle V Consumer involvement in research projects: the activities of research funders Health Policy 2004 69 229 238 15212869 10.1016/j.healthpol.2003.12.011 Hanley B Truesdale A King A Elbourne D Chalmers I Involving consumers in designing, conducting, and interpreting randomised controlled trials: questionnaire survey Bmj 2001 322 519 523 11230065 10.1136/bmj.322.7285.519 Consumers Health Forum of Australia www.chf.org.au, accessed 24th August 2005 Macdonald LA Sackett DL Haynes RB Taylor DW Labelling in hypertension: a review of the behavioural and psychological consequences J Chronic Dis 1984 37 933 942 6396317 10.1016/0021-9681(84)90070-5 Logan RA Longo DR Rethinking anti-smoking media campaigns: two generations of research and issues for the next J Health Care Finance 1999 25 77 90 10353092 Korhonen T Urjanheimo EL Mannonen P Korhonen HJ Uutela A Puska P Quit and Win campaigns as a long-term anti-smoking intervention in North Karelia and other parts of Finland Tob Control 1999 8 175 181 10478402 Mudde AN de Vries H Dolders MG Evaluation of a Dutch community-based smoking cessation intervention Prev Med 1995 24 61 70 7740017 10.1006/pmed.1995.1009 Miller DR Geller AC Wood MC Lew RA Koh HK The Falmouth Safe Skin Project: evaluation of a community program to promote sun protection in youth Health Educ Behav 1999 26 369 384 10349574 Duan N Fox SA Derose KP Carson S Maintaining mammography adherence through telephone counseling in a church-based trial Am J Public Health 2000 90 1468 1471 10983211 Derose KP Hawes-Dawson J Fox SA Maldonado N Tatum A Kington R Dealing with diversity: recruiting churches and women for a randomized trial of mammography promotion Health Educ Behav 2000 27 632 648 11009131 Herborg H Soendergaard B Jorgensen T Fonnesbaek L Hepler CD Holst H Froekjaer B Improving drug therapy for patients with asthma-part 2: Use of antiasthma medications J Am Pharm Assoc (Wash) 2001 41 551 559 11486981 Pescatello LS Murphy D Vollono J Lynch E Bernene J Costanzo D The cardiovascular health impact of an incentive worksite health promotion program Am J Health Promot 2001 16 16 20 11575051 Canadian Co-ordinating Office for Health Technology Assessment https://www.ccohta.ca/mpup/index_e.cfm	16318638	PMC1325245	CC BY	2021-01-04 16:31:53	no	BMC Health Serv Res. 2005 Dec 1; 5:75	utf-8	BMC Health Serv Res	2,005	10.1186/1472-6963-5-75	oa_comm
==== Front BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-1101633665710.1186/1471-2334-5-110Research ArticleBacterial isolates from blood cultures of children with suspected septicaemia in Calabar, Nigeria Meremikwu Martin M [email protected] Chukwuemeka E [email protected] Anne E [email protected] Joseph U [email protected] Simon J [email protected] Departments of Paediatrics, Faculty of Clinical Sciences, University of Calabar, Calabar, Nigeria2 Department of Medical Microbiology and Parasitology, Faculty of Laboratory and Allied Health Sciences, University of Calabar, Calabar, Nigeria2005 8 12 2005 5 110 110 9 3 2005 8 12 2005 Copyright © 2005 Meremikwu et al; licensee BioMed Central Ltd.2005Meremikwu 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 Septicaemia is a common cause of morbidity and mortality among children in the developing world. This pattern has changed little in the past decade. Physical signs and symptoms, though useful in identifying possible cases have limited specificity. Definitive diagnosis is by bacteriologic culture of blood samples to identify organisms and establish antibiotic susceptibility. These results are usually not available promptly. Therefore a knowledge of epidemiologic and antimicribial susceptibility pattern of common pathogens is useful for prompt treatment of patients. This report highlights the pattern of bacterial isolates in our environment from a retrospective study of our patients' records. Methods One thousand, two hundred and one blood samples were analysed from children aged 0–15 years, admitted into the children's wards of the University of Calabar Teaching Hospital, Calabar, Nigeria with features suggesting septicaemia. Samples were collected under aseptic conditions and cultured for aerobic and anaerobic organisms. Isolates were identified using bacteriologic and biochemical methods and antibiotic sensitivity determined by agar diffusion method using standard antibiotic discs. Results Bacteria was isolated in 552 (48.9%) of samples with highest rates among newborns (271 : 50.8). The most frequent isolates were Staphylococcal aureus (48.7%) and Coliforms (23.4%). Results showed high susceptibilities to the Cephalosporins (Ceftriazone- 100%:83.2%, Cefuroxime-100%:76.5%) and Macrolides (Azithromycin-100%:92.9%) for S. aureus and coliforms respectively. This study underscores the importance of septicaemia as a common cause of febrile illness in children and provides information on common prevalent aetiologic agents and drug susceptibilities of the commonest pathogens. Conclusion Staphylococcus aureus and coliforms were the leading causes of septicaemia in children in this locality, and the third generation cephalosporins and azithromycin were shown to be effective against these pathogens. ==== Body Background Septicaemia is a common cause of paediatric morbidity and mortality. Deaths from paediatric septicaemia are likely to be higher in low-income settings. Children with septicaemia present with fever, difficult breathing, tachycardia, malaise, inability to feed or lethargy, but those with asymptomatic bacteraemia tend to show no obvious sign of illness. In Nigeria, septicaemia is a major cause of death in neonates and children [1,2]. The outcome of treatment of neonates with septicaemia has remained poor in Nigeria as shown by reports of mortality rate of 33% to 41% from two tertiary hospitals in the country [3,4]. Prompt diagnosis and effective treatment is necessary to prevent death and complications from septicaemia. Physical signs and symptoms are useful in identifying infants and children with septicaemia and other non-localised infections but these have limited specificity [5,6]. Clinical assessment using a combination of symptoms and signs are useful guides to provisional diagnosis of septicaemia. For instance, a seven-item, weighted, clinical score system comprising grunting, abdominal distension, increased pre-feed aspirates, tachycardia, hyperthermia, chest retractions and lethargy showed that these criteria were sensitive for identifying newborns with septicaemia [7]. Rapid immunological techniques like C-Reactive Proteins (CRP) assays may help in the preliminary diagnostic assessment of suspected septicaemia. However, they lack the capacity to detect specific pathogens. Bacteriological culture to isolate the offending pathogen remains the mainstay of definitive diagnosis of septicaemia. The results of bacteriological cultures and antibiotic susceptibility tests take about a week, necessitating initial empirical treatment of suspected septicaemia. Knowledge of epidemiological and anti-microbial susceptibility pattern of common pathogens in a given area helps to inform the choice of antibiotics. We report the pattern of bacterial isolates in children with clinical diagnosis of septicaemia seen at the University of Calabar Teaching Hospital in South-eastern Nigeria. Methods The study included all consecutive blood cultures in children aged 0–15 years admitted to the University of Calabar Teaching Hospital (UCTH) from July 1996 to December 2002. The indication for blood cultures were clinical features adjudged by the attending clinician to be indicative of sepsis especially fever without localized lesion and absence of asexual malaria parasite in peripheral blood. Presence of malaria parasitaemia did not preclude suspicion of septicaemia when illness was severe or risk factors of sepsis such as severe malnutrition, lethargy and abdominal distension were present. Specimens were collected into cooked meat broth under aseptic conditions using sterile, disposable hypodermic needle and syringe and transported to the laboratory within 30 minutes of collection. Aerobic cultures were mounted on the same day and sub-cultured within 48 hours if there was any indication of growth. Eosin methylene blue and blood agar plates were used for sub-culture and incubated at 37°C aerobically and in candle extinction jar respectively. Full identification of organisms was done with standard bacteriological and biochemical methods [8]. Antibiotic sensitivity patterns of bacterial isolates were determined by agar diffusion method using antibiotic discs [9] Results A total of 1,201 children with suspected septicaemia were studied. The age and sex distribution of the patients are shown in Table 1. There were 539 (44.9%) females and 662 (55.1%) males. The majority of the patients were newborns (533; 44.4%) and infants (252; 21.0%). Table 1 Age and sex distribution of 1201 children with suspected septicaemia in Calabar Nigeria Age Number of Children Examined Number (%) with positive bacterial isolates Female Male Total < 1 mo 230 303 533 271 (50.8) 1 mo – 1 yr 111 141 252 113 (44.8) 2 yr – 5 yr 126 133 259 113 (43.6) 6 yr – 10 yr 40 55 95 39 (41.0) 11 yr – 18 yr 32 30 62 16 (25.8) Total 539 662 1201 552 (46.0) Bacteria were isolated in 552 (45.9%) of the 1,201 patients studied. The types and pattern of bacterial isolates in age groups is shown on Table 2. The rate of isolation was highest among newborns (271/533: 50.8%). The overall rate of isolation reduced with increasing age but the types of organisms cultured did not vary with age. The most frequent isolates were Staphylococcus aureus (48.7%) and Coliforms (23.4%). Unidentified gram-negative rods (8.0%), Pseudomonas aeruginosa spp (5.8%), Streptococcal spp (4.7%) and Chromobacterium spp (4.5%) were also fairly frequently identified. We did not isolate anaerobes. Our laboratory techniques may not have been sensitive enough to detect obligate anaerobes. Table 2 Age distribution of 552 bacterial isolates from blood cultures of 1201 children with suspected septicaemia in Calabar Nigeria Types of bacterial isolates Number of children with positive bacterial cultures in age groups (N = 552) ≤ 1 mo. 2 mo. – 1 yr 2 yr – 5 yr 6 yr – 10 yr 11 – 18 yr Sub-total (% N) S. aureus 138 52 51 22 6 269 (48.7) Coliforms 63 24 33 7 2 129 (23.4) Unidentified Gram-negative rods. 18 11 10 3 2 44 (8.0) Pseudomonas spp. 16 5 9 2 0 32 (5.8) β-haemolytic Streptococci 5 3 2 1 1 12 (2.2) Other Streptococci 8 4 2 0 0 14 (2.5) Chromobacterium spp. 15 5 2 2 1 25 (4.5) Salmonella typhi 0 2 0 0 2 4 (0.7) Other Salmonella spp. 5 3 0 1 0 9 (1.6) CNS* 1 4 2 0 2 9 (1.6) Proteus spp. 1 1 2 1 0 5 (0.9) Sub-total (% N) 271 (49.1) 113 (20.5) 113 (20.5) 39 (7.1) 16 (2.9) 552 (100) CNS: Coagulase-negative staphylococci Table 3 shows the anti-microbial sensitivity pattern of Staphylococcus aureus and Coliforms that made up 72% of all isolates. Antibiotic sensitivity data for the other isolates are not presented because they are too few. Staphylococcus aureus had the highest susceptibility to Ceftriazone (100%), Cefuroxime (100%), Azithromycin (100%), Erythromycin (90.1%) and Gentamicin (86.6%). Coliforms were most susceptible to Ceftazidime (78.8%), Ceftriaxone (83.3%), Cefuroxime (76.5%) and Azithromycin (92.9%). Table 3 Antimicrobial sensitivity pattern of Staphylococcus aureus and Coliforms, the commonest bacterial isolates of children with septicaemia in Calabar Nigeria Types of organism % of isolates susceptible to antimicrobial agents A C E Se Ge CFT CFZ CEX Z Staphylococcus aureus 4.4 57.6 90.1 28.5 86.6 66.9 100 100 100 (136) (159) (182) (123) (179) (115) (30) (65) (19) Coliforms 11.4 34.1 - 23.6 61.6 78.8 83.3 76.5 92.9 (44) (91) - (72) (86) (66) (18) (34) (14) *Figures in parenthesis show number of isolates tested. Legend: A = Ampicillin C = Chloramphenicol E = Erythromycin Se = Cotrimoxazole Ge = Gentamicin CFZ = Ceftriazone CFT = Ceftazidine CEX = Cefuroxime Z = Azithromycin Discussion The present study includes children of all age groups but neonates are in the majority (44.4%). It reveals a rather high rate (44.9%) of isolation of bacterial pathogens from blood cultures of children with provisional diagnosis of septicaemia. The isolation rate is comparable to rates reported in other studies of Nigerian neonates with suspected septicaemia in Calabar (50.6%) [10], Ilorin (30.8%) [11] and Ife (55%) [12]. In all these reports, the organisms were S. aureus and gram-negative rods (Pseudomonas aeruginosa and Escherichia coli). The present study shows a similar pattern. This observation is in consonance with reports from two other developing countries [6,13] This suggests that infections by these agents constitute a significant threat to child survival in this locale and other developing country settings. The in vitro susceptibility tests of these two most common isolates (coliforms and S. aureus) showed high levels of resistance to such commonly used antibiotics as Ampicillin, Chloramphenicol and Co-trimoxazole. While 86.6% of S. aureus isolates were sensitive to Gentamicin, only 61.6% of coliforms were sensitive to the same antibiotic, showing higher levels of resistance than reported in the same hospital about a decade ago, which showed a sensitivity of 89.7% [10,14]. The local antibiotics policy informed by the result of that study recommends the use of gentamicin as the sole agent for initial therapy in neonatal septicaemia. This policy needs to be reviewed to include a third generation cephalosporin in keeping with the result of the present study. Susceptibility of S. aureus to Erythromycin remains high in the current study (90.1%). Sensitivity of the few isolates of S. aureus tested against Azithromycin is 100%, comparable to the findings of another Nigerian study [11]. This highlights the variable nature of antibiotic susceptibility pattern both in time and location within the same country. Susceptibility of coliforms and S. aureus to third generation Cephalosporins (namely Ceftazidine, Cefuroxime and Ceftriazone), and Azithromycin were quite good. A recent report from Pakistan, another developing country showed high levels of resistance of gram-negative organisms to ceftazidime (71.6%) and cefotaxime (55.2%) [13]. Improved commitment to rational use of these antibiotics in Calabar is needed to sustain this relatively high level of susceptibility. Conclusion This study has shown that S. aureus and gram-negative rods (Pseudomonas spp and coliform) are the leading causes of septicaemia in children in South-east Nigeria, a pattern similar to that of other low income countries. Observed decline in susceptibility of these common pathogens to common antibiotics calls for increased efforts to ensure more rational use of these drugs. Epidemiological surveillance studies such as the current one should provide useful information base to guide practice and policies on rational use of anti-infective agents. Competing interests The author(s) declare that they have no competing interests. Authors' contributions MMM, CEN and JUO participated in the clinical assessment and treatment of the patients. AEA and SJU supervised the laboratory procedures. CEN and MMM analysed the data. All the authors participated in the preparation of the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements We are grateful to M. Arisa and C. Ugorji for helping with data work. We gratefully appreciate the help of laboratory personnel of the microbiology department of the University of Calabar Teaching Hospital. ==== Refs Asindi AA Ekanem AD Neonatal septicaemia in Calabar Nigeria East African Medical Journal 1988 65 335 41 3168870 Asindi AA Ibia EO Udo JJ Mortality pattern among Nigerian children in the 1980s J Trop Med Hyg 1991 94 152 155 2051519 Omene JA Neonatal septicaemia in Benin City (Nigeria). A review of 74 cases Tropical Geographical Medicine 1979 31 35 39 Adejuyigbe EA Adeodu OO Ako-Nai KA Taiwo O Owa JA Septicaemia in high risk neonates at a teaching hospital in Ile-Ife, Nigeria East Afr Med J 2001 78 540 3 11921599 Hague RA Eastman EJ Lee RE Cant AJ Resolution of hepatic abcess after interferon gamma in chronic granulomatous disease Arch Dis child 1993 69 443 5 8259876 Weber MW Carlin JB Gatchalian S Lehmann D Muhe L Mulholland EK WHO Young Infants Study Group Predictors of neonatal sepsis in developing countries Pediatr Infect Dis J 2003 22 711 7 12913772 Singh SA Dutta S Narang A Predictive clinical scores for diagnosis of late onset neonatal septicemia J Trop Pediatr 2003 49 235 9 12929886 10.1093/tropej/49.4.235 Finegold SM Markin WJ Scott EJ Bailey & Scott's Diagnostic Microbiology 1978 5 St. Louis: The CV Mosby Company Oxoid Ltd, Wade, Road Basingstoke Hampshire, RG24 8PW England Antia-Obong OE Utsalo SJ Udo JJ Udo KT Neonatal septicaemia in Calabar, Nigeria Cent Afr J Med 1992 38 161 5 1394397 Mokuolu AO Jiya N Adesiyun OO Neonatal septicaemia in Ilorin: bacterial pathogens and antibiotic sensitivity pattern Afr J Med Med Sci 2002 31 127 30 12518907 Ako-Nai AK Adejuyigbe EA Ajayi FM Onipede AO The bacteriology of neonatal septicaemia in Ile-Ife, Nigeria J Trop Pediatr 1999 45 146 51 10401192 10.1093/tropej/45.3.146 Aurangzeb B Hameed A Neonatal sepsis in hospital-born babies: bacterial isolates and antibiotic susceptibility patterns J Coll Physicians Surg Pak 2003 13 629 32 14700488 Antia-Obong OE Utsalo SJ Bacterial agents in neonatal septicaemia in Calabar, Nigeria: review of 100 cases Trop Doct 1991 21 169 70 1746038	16336657	PMC1325246	CC BY	2021-01-04 16:28:17	no	BMC Infect Dis. 2005 Dec 8; 5:110	utf-8	BMC Infect Dis	2,005	10.1186/1471-2334-5-110	oa_comm
==== Front BMC Cell BiolBMC Cell Biology1471-2121BioMed Central London 1471-2121-6-441633664310.1186/1471-2121-6-44Research ArticleCommon themes and cell type specific variations of higher order chromatin arrangements in the mouse Mayer Robert [email protected] Alessandro [email protected] Hase Johann [email protected] Timm [email protected] Thomas [email protected] Steffen [email protected] Ludwig-Maximilians-Universität München, Department Biologie II, Groβhaderner Str 2, 82152 Planegg-Martinsried, Germany2 Kirchhoff Institut für Physik, Universität Heidelberg, Germany3 Institute of Stem Cell Research, GSF – National Research Center for Environment and Health, Neuherberg, Germany2005 7 12 2005 6 44 44 14 9 2005 7 12 2005 Copyright © 2005 Mayer et al; licensee BioMed Central Ltd.2005Mayer 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 Similarities as well as differences in higher order chromatin arrangements of human cell types were previously reported. For an evolutionary comparison, we now studied the arrangements of chromosome territories and centromere regions in six mouse cell types (lymphocytes, embryonic stem cells, macrophages, fibroblasts, myoblasts and myotubes) with fluorescence in situ hybridization and confocal laser scanning microscopy. Both species evolved pronounced differences in karyotypes after their last common ancestors lived about 87 million years ago and thus seem particularly suited to elucidate common and cell type specific themes of higher order chromatin arrangements in mammals. Results All mouse cell types showed non-random correlations of radial chromosome territory positions with gene density as well as with chromosome size. The distribution of chromosome territories and pericentromeric heterochromatin changed during differentiation, leading to distinct cell type specific distribution patterns. We exclude a strict dependence of these differences on nuclear shape. Positional differences in mouse cell nuclei were less pronounced compared to human cell nuclei in agreement with smaller differences in chromosome size and gene density. Notably, the position of chromosome territories relative to each other was very variable. Conclusion Chromosome territory arrangements according to chromosome size and gene density provide common, evolutionary conserved themes in both, human and mouse cell types. Our findings are incompatible with a previously reported model of parental genome separation. ==== Body Background The existence of chromosome territories as restricted volumes in which the DNA of only one chromosome is spatially arranged during interphase is now established for about 20 years [1,2]. The distribution of individual territories within the nucleus has come into focus more recently. Although the side-by-side arrangement of chromosome territories can change from one cell cycle to the next [3,4], the radial organization of chromatin in the nucleus in general is not random. So far, the most widespread detected principle of functional nuclear architecture is the specific positioning of chromatin with different replication time points in S-phase. From single cell eukaryotes [5] to distantly related multicellular organisms like Hydra [6], chicken [7], humans [8,9] and plants [10], a layer of chromatin replicating in mid S-phase was found at the nuclear periphery and around nucleoli, while early replicating chromatin was distributed in interior nuclear zones between the perinucleolar and perinuclear compartments. For chromosome territories and some chromosomal subregions non-random radial distributions have been described. Several studies have shown that in spherical nuclei of both quiescent and cycling human lymphocytes, gene rich chromosomes are statistically more often centrally located while gene poor chromosomes are preferentially at the periphery [11-13]. Such an arrangement has also been reported for the gene rich human chromosome 19 homologs and gene poor chromosome 18 homologs in other primates [14]. Controversial results have been reported for flat human fibroblast nuclei. While one group described gene rich chromosome territories to be centrally and gene poor ones peripherally located [11,13,15], other groups described a size dependent radial distribution where large chromosomes are preferentially peripheral and small chromosomes internal [12,16,17]. While for spherical nuclei of cells growing in suspension all sites at the nuclear periphery are topologically indistinguishable from each other, a flat ellipsoidal nucleus as in fibroblasts possesses a unique outer rim defined by the intersection of the horizontal mid-plane with the nuclear border. In recent work on fibroblast nuclei of Homo sapiens (HSA) we found that territories of both, gene poor chromosome HSA18 and gene rich chromosome HSA19, stay close to the nuclear center, remote from the outer rim just described [17]. Accordingly, they were often neighbors. HSA18 territories, however, were on average located closer than HSA19 territories to the top and bottom part of the envelope of structurally preserved nuclei. In contrast, in spherical lymphocyte nuclei gene rich HSA19 territories are typically located in the nuclear interior while gene poor HSA18 territories are associated with the nuclear envelope and thus away from HSA19. The shape of nuclei thus apparently plays a role in territory positioning. Differences in the distribution of some chromosome territories in different cell types have also been described in mouse [18]. Interestingly, cell types from related differentiation pathways like large and small lung cells, were found to have more similar chromosomal distribution patterns than unrelated cell types. Only chromosomes with low to average gene density from unsynchronized cells were investigated [18] and gene density and cell cycle topics were not addressed. In chicken, large, gene poor chromosomes have been found peripheral and small, gene rich ones more centrally in flat fibroblast nuclei, in semi-spherical neuroblast nuclei and spherical nuclei of some blood cell types [7,19]. While repetitive probes for individual centromeric regions have been used in many studies to identify the position of specific chromosomes [20-22] centromeres as functional entities also have been a subject of interest. Changes in centromere distribution have been associated with differentiation processes [23-26]. Clusters of centromeric and pericentromeric regions, so called chromocenters, were shown to be involved in silencing of several genes in various hematopoietic cell types [27,28] and MeCP2, a protein binding to methylated DNA, was shown to induce clustering of these regions during mouse myotube differentiation [29]. These findings indicate that in order to understand the rules which govern nuclear functions, we need to understand chromatin structure at all levels, from the organization of the nucleosome fiber carrying individual active and silent genes to the architecture and arrangements of chromosome territories in the nuclear space. In the present study we tested the spatial organization of chromocenters and chromosome territories for specific distribution patterns in six mouse cell types. We also asked whether chromosome distribution in lymphocytes and fibroblasts is conserved from man to mouse. An obvious difference between human and mouse karyotypes is that all chromosomes of Mus musculus have the centromere near one telomere (telocentric) with large heterochromatic blocks nearby (Figure 1a) while human chromosomes are divided by the centromere in two arms of approximately similar size (metacentric) or a rather short and rather long arm (acrocentric) or ratios in between (submetacentric) (Figure 1b). Differences in gene content and size between chromosomes are also smaller in mouse than in humans (Figure 1). Disregarding human and mouse Y-chromosomes, human chromosome sizes vary about fivefold (245 – 47 Megabasepairs, Mbp [30] and human chromosome gene densities vary more than sixfold (23 – 3.5 genes/Mbp [30]). Mouse chromosomes (Figure 1a) vary in gene density only about twofold (15.9 – 7.5 genes/Mbp) and in size about threefold (195 – 61 Mbp). We tested the radial distribution of mouse chromosome territories for size-dependence and gene-density dependence in cell types with nuclear shapes ranging from spherical to flat ellipsoidal. We also investigated the position of homologous and pairs of heterologous chromosomes relative to each other. We selected 6 chromosomes to cover big, small, gene rich and gene poor examples (Figure 1a). For chromocenters, the degree of clustering and the nuclear distribution was measured in each cell type. All examinations were made on nuclei with a defined cell cycle stage. Figure 1 Sizes and gene densities of mouse (a) and human (b) chromosomes. The length of chromosomes is scaled according to their genomic size. Gene density in genes/Mbp for each chromosome is indicated in a box. The distance of the box to the x-axis is scaled according to the gene density. Centromeres are indicated in black. Chromosomes with NORs are indicated with an asterisk. Mouse chromosomes investigated in this study are surrounded with the color used in Figure 3 for their representation. (a) Mouse chromosomes have an average size of 124 ± 32 Mbp (standard deviation) and an average gene density of 10.1 ± 2.3 genes/Mbp. For both parameters the variation in human chromosomes (b) is thus much larger. They have an average size of 128 ± 56 Mbp and an average gene density of 7.3 ± 4.2 genes/Mbp. Values are disregarding the Y-chromosomes. Data are from Ensemble Genome Browser [30]. Results Experimental design We performed dual color fluorescence in situ hybridization on formaldehyde fixed, morphologically preserved nuclei (3D-FISH) of Mus musculus (MMU) cells (Figure 2). Embryonic stem (ES) cells, in vitro differentiated macrophages, primary fibroblasts and stimulated primary lymphocytes were hybridized with the following combinations of whole chromosome paint probes: chromosomes 1 and 14, 2 and 9, 11 and X. Chromosomes 11 and X were additionally hybridized to unstimultated lymphocytes, myoblasts and myotubes. The rationale for choosing these chromosomes reflects their differences in size and gene content (Figure 1). MMU1 is the largest mouse chromosome (195 Mbp). MMU14 is comparatively small (119 Mbp) and both are gene poor (8.4 and 8.7 genes/Mbp, respectively). MMU2 is large (182 Mbp) compared to MMU 9 (124 Mbp) and both are more gene rich than the first pair (10.7 and 11.7 genes/Mbp, respectively), being the 6th and 5th gene richest chromosome. MMU11 (122 Mbp) is similar in size to MMU9 and MMU14 and the most gene rich chromosome in the mouse karyotype (15.9 genes/Mbp). MMUX has the third lowest gene density of all mouse chromosomes (7.9 genes/Mbp) and is rather large (164 Mbp). Only X-chromosomes from male cells were investigated. The smaller chromosomes 15–19 were not included in our analysis since they as well as MMU12 may contain nucleolar organizing regions [31,32] which are tethered to nucleoli and thus spatial distribution would be biased. Figure 2 3D-FISH on structurally preserved mouse nuclei. Mouse chromosome pairs MMU1/MMU14 (a, d, g, j), MMU2/MMU9 (b, e, h, k) and MMU11/MMUX (c, f, i, l, m, n, o) were detected in S-phase lymphocytes (a, b, c), fibroblasts (d, e, f) embryonic stem cells (g,h,i), macrophages (j, k, l), myoblasts (m), myotubes (n) and G0 lymphocytes (o). TO-PRO-3 (pseudocolored blue) was used as DNA counterstain. Maximum intensity projections of confocal image stacks are shown. All images are shown to scale. Scalebar: 5 μm To minimize potential influences of cell cycle stages [15,33-35] we only recorded nuclei in a defined stage. In proliferating cell types, cultures were labeled with a short BrdU pulse to identify S-phase cells, and only BrdU positive cells were recorded. Post-mitotic macrophages were identified by the absence of a BrdU signal after 24 hours of incubation. Unstimulated lymphocytes and myotubes were considered as G0 since in control experiments they did not incorporate BrdU even if incubation times >24 h were used. After counterstaining for DNA, confocal image stacks were recorded and quantitatively evaluated for the parameters described in the following sections. For each chromosome combination and cell type we typically recorded 30 nuclei. All quantitative evaluations were carried out in 3D. The radial distribution of chromosome territories is cell type specific A total of 962 three-dimensional image stacks with chromosome territories plus the respective image stacks of counterstained nuclear DNA was analyzed. The radial distribution of painted chromosome territories was measured in each nucleus relative to the nuclear radius and averaged over the set of recorded nuclei (Figure 3). In mouse lymphocyte nuclei we found territories of the most gene rich chromosome MMU11 typically much more internally located than all other chromosomes, the most prominent difference in radial distribution in our study (see Figure 4 for p-values). MMU2 and MMU9 occupied intermediate positions and gene poor MMU1, MMU14 and MMUX were found more towards the periphery (Figure 3a,b,c,f). Thus the distribution of tested chromosomes was in agreement with a gene density related positioning. Linear regression analysis (Figure 3b) revealed a steeper line for lymphocyte nuclei than for other cell types, indicating a stronger dependency on this factor in this cell type. Linear regression resulted in a correlation coefficient of 0.62, indicating a good fit of the data (Figure 3b). The distribution of MMU11 in lymphocytes was significantly more internal compared to other cell types (see Figure 4 for p-values). No significant difference was found between G0 and S-phase lymphocytes. Figure 3 Three-dimensional relative radial nuclear distribution of mouse chromosome territories from a total of 481 nuclei with two labeled pairs of chromosomes each. Each nucleus was divided in 25 shells with equal spacing. The relative amount of the signal from a given chromosome paint probe or of the DNA counterstain (cst) in each of the shells was measured and averaged over all nuclei of the respective cell type (see Methods for details). Graphs on the left (a,d,g,j,m) show the percentage of each signal in the 25 shells. Error bars in (a-o) show the standard error of the mean. For each nucleus and territory signal, the median radius of the relative radial distribution was calculated (see additional data file 1 for complete listing). Graphs in the center (b,e,h,k,n) show the averages of these medians plotted against the gene density of the painted chromosome. In the graphs on the right (c,f,i,l,o) these averages of medians are plotted against chromosomal size. For cell types with investigated territories from six types of chromosomes, the black lines show the linear regression. The box shows the correlation coefficient R and the p-value indicating the probability that there is no correlation but that the observed relation was a chance result. Both values were calculated from individual medians from all nuclei while the regression line shown was fitted to the six average values of the medians represented by the dots. (a-c) Lymphocyte nuclei in S-phase (continuous lines in (a), circles in (b), all six investigated chromosomes) and in G0 (broken lines in (a), squares in b,c, MMU11 and MMUX only). In (b,c), data points for average medians in G0 lymphocytes fall exactly on data points in S-phase and are thus difficult to distinguish. G0 lymphocytes were not included in the linear regression analysis. (d-f) Fibroblast S-phase nuclei. (g-i) ES-cell S-phase nuclei. (j-l) Macrophage G0 nuclei. Color code shown in (a) also applies to (b-l). (m-o) Myoblast (continuous lines) and myotube nuclei (broken lines). Correlation coefficients for myoblasts (0.26, p = 0.032) and myotubes (0.68, p < 0.001) are not directly comparable since only two chromosomes were investigated. Note that in all cell types there is a correlation for gene poor as well as large chromosome territories to be more peripheral, although for some chromosome territories average medians diverge from this pattern. Figure 4 Statistical comparison of relative radial distributions of chromosome territories. Significance levels are shown as p-values from a two-sided Kolmogorov-Smirnov test and are color coded as indicated. Chromosome names are highlighted according to their gene density as indicated. (a) Pair wise comparison of the relative radial distributions of chromosomes in a given cell type. Lymphocytes in S-phase and in G0 are listed separately. (b) Pair wise comparison of the relative radial distributions of a given chromosome territory between cell types. Primary mouse fibroblasts differed from other cell types in that the radial distribution curves for all investigated chromosomes were rather similar (Figure 3d). A tendency for the location of gene rich chromosomes towards internal and of gene poor chromosomes towards the nuclear border was found (Figure 3e) but except for MMU1 differences between chromosomes were not significant (Figure 4). A tendency for a more peripheral location of large chromosomes was also noted (Figure 3f). Although ES cell nuclei have a shape very similar to lymphocyte nuclei, some chromosome territories were distributed quite differently. MMU11 was located less internal than in lymphocyte nuclei (p < 0.001) although it was again the most internal chromosome (Figure 3g,h,i, Figure 4). The small, gene poor MMU14 territories were found significantly more internal than in lymphocyte nuclei. The correlation coefficient was similar for gene density and size dependent chromosome territory distribution. Post-mitotic macrophages were differentiated in vitro from ES-cells. They differed significantly from ES cells in the radial nuclear distribution of MMU1 which was now observed further outside (Figure 3j,k,l, Figure 4b). In this cell type, distribution of territories showed a better correlation with chromosomal size than with gene density (Figure 3k,l). MMU11 distribution changed significantly during myoblast-to-myotube differentiation (Figure 4b). MMUX territories were significantly more peripheral in S-phase myoblasts and in particular in postmitotic myotubes than in other cell types (Figure 3m,n,o, Figure 4b). In a subset of nuclei from each cell type we visually analyzed whether chromosome territories were in contact with the nuclear border or not. In ES-cells (n = 28), macrophages (n = 28), myoblasts (n = 35) and myotubes (n = 21) all inspected MMU11 and MMUX territories had contact with the nuclear periphery. In 7 of 31 lymphocyte nuclei (23%), one of the two MMU11 homologs apparently did not touch the nuclear border while the other MMU11 and all MMUX territories did. Chromosome territories do not have fixed positions relative to each other Confocal image stacks were also used to determine the positioning of chromosome territories relative to each other. Three-dimensional angles between chromosome territories were calculated using the intensity gravity centers of the chromosome territories and the geometrical center of the segmented nucleus as point of origin. If homologous chromosome territories were touching each other we attempted their separation by increasing the threshold. For those nuclei where a separation could not be achieved in this way, the angle between the homologs was set to zero. Homologous association of chromosome territories would lead to small angles, while a parental separation of haploid sets with an antiparallel order of chromosomes as suggested by Nagele et al. [36] would lead to angles close to 180 degrees for homologs. In disagreement with both models, cumulative distribution curves show that angles cover the whole range between 0° and 180° for all investigated chromosomes in all investigated cell types (Figure 5c,e,g,i). While we observed a high frequency of 0° (= inseparable territories) in some cases, this was typically compensated by the sparse occurrence of small angles between 1° and 40°. An exception are MMU1 chromosomes in macrophages where 45% of the nuclei showed inseparable territories (9 of 20 nuclei). In the investigated cell types, only the comparison between MMU1 and MMU11 homologous association in macrophages revealed a highly significant difference (Table 1). Mean values also are not compatible with predictions of either model (Table 1). We conclude that measured angles between homologous chromosomes are incompatible with both, non-random homologous association and parental genome separation with the possible exception of MMU1 in macrophage nuclei. Figure 5 Angles between chromosome territories. (a) Scheme of fixed chromosomal angles between two pairs of chromosomes as predicted by the model of genome separation with antiparallel orientation [36]. In this example, specific angles are assumed to be about 60° and 120°. A cumulative distribution plot of homologous angles in this model has a steep increase from 0 to 100% near 180° (not shown). (b) Cumulative distribution of heterologous angles in this model example. For other chromosome pairs, angles and respective sharp increases of cumulative graphs would be at other values. Cumulative plots are shown for the indicated cell types. (c-j) Experimentally observed cumulative distributions of angles between homologous (c,e,g,i) or heterologous (d,f,h,j) chromosome territories in the indicated cell types. Table 1 Angles between homologous chromosomes. Mean values (mv), median, standard deviation (std) and numbers of evaluated nuclei (n) are listed. In addition, p-values derived from the comparison of homologous angles between pairs of homologous chromosomes are shown. p-values <0.05 are underlined. lymphocytes mv median std n MMU2 MMU9 MMU11 MMU14 MMU1 85.0 83.9 48.5 34 0.254 0.608 0.728 0.185 MMU2 95.3 105.6 46.9 32 0.964 0.176 0.044 MMU9 99.5 105.5 44.8 32 0.146 0.163 MMU11 79.4 89.4 44.4 41 0.532 MMU14 71.7 69.9 50.5 34 fibroblasts mv median std n MMU2 MMU9 MMU11 MMU14 MMU1 77.5 69.8 58.4 26 0.023 0.137 0.061 0.303 MMU2 103.4 118.0 43.0 30 0.586 0.512 0.682 MMU9 102.5 103.1 46.5 30 0.783 0.434 MMU11 112.2 113.2 42.9 31 0.331 MMU14 92.2 97.4 43.7 26 ES cells mv median std n MMU2 MMU9 MMU11 MMU14 MMU1 101.4 99.2 49.1 25 0.605 0.564 0.182 0.468 MMU2 82.0 90.8 55.0 30 0.388 0.494 0.981 MMU9 98.8 102.0 40.9 30 0.761 0.729 MMU11 97.5 100.6 32.5 22 0.899 MMU14 89.0 90.5 50.6 25 macrophages mv median std n MMU2 MMU9 MMU11 MMU14 MMU1 46.8 57.6 46.5 20 0.169 0.18 0.003 0.035 MMU2 84.2 87.1 44.2 18 0.999 0.081 0.866 MMU9 81.1 81.0 52.9 17 0.184 0.929 MMU11 109.1 126.3 52.9 30 0.059 MMU14 89.7 106.1 52.4 20 The model of parental genome separation with a deterministic antiparallel order of chromosomes in the parental chromosome sets [36] requests that angles between given heterologous chromosome territories vary within narrow limits. In a comparison of all "heterologous" angles between two pairs of chromosomes, we would expect a distinct bimodal distribution of these heterologous angles. A fixed smaller angle (Figure 5a, A1-B1 and A2-B2) and a fixed larger angle (A1-B2 and A2-B1). In a cumulative frequency distribution histogram such peaks would cause two sharp increases in the curve at the respective angles (Figure 5b). For different pairs of chromosome territories, fixed but different heterologous angles would be expected, reflecting the different positions of individual chromosome territories in haploid sets with antiparallel order. We measured angles between heterologous chromosomes for the pairs MMU1-MMU14, MMU2-MMU9 and MMU11-MMUX. For a given nucleus, all four heterologous angles (two where MMUX was involved) were calculated if all four territories (three with MMUX) could be segmented. In contrast to the Nagele model [36], all curves for measured heterologous angles showed a very steady increase (Figure 5d,f,h,j), arguing for a very variable side-by-side distribution of chromosome territories. Angle distributions between heterologous chromosomes are very similar for different pairs of chromosomes and we could not detect significant differences in the studied cell types (p > 0.1). We conclude that measured angles between heterologous chromosomes are incompatible both with the hypothesis of parental genome separation and with an antiparallel order of the two haploid sets. The spatial organization of chromocenters is dynamic and cell type specific In mouse interphase nuclei centromere regions cluster to a different degree according to cell type [23-25,29,37], thereby generating so called chromocenters. The mouse major satellite is a pericentromeric satellite DNA that is present in all mouse chromosomes except the Y-chromosome [38]. In nuclei subjected to FISH with the mouse major satellite as a probe (Figure 6) we first counted the number of chromocenters per nucleus. The strongest clustering reflected by the smallest numbers of chromocenters was found in lymphocyte nuclei, the least clustering was observed in fibroblast nuclei (Figure 6, Figure 7; difference highly significant, p < 0.001). Serum-starved G0 fibroblasts did not show a significant difference when compared to S-phase fibroblasts (p > 0.2). During the differentiation of ES cells to postmitotic macrophages, the number of chromocenters decreased significantly (p < 0.01) from an average of 14.7 to 10.2. Post-mitotic macrophages were identified by the presence of CD11b surface antigen and the absence of BrdU incorporation after a 24 hour incubation. We occasionally observed adherent CD11b positive cells that had incorporated BrdU. Some of these cycling macrophage precursors were possibly in their last round of S-phase before entering the postmitotic stage. When we counted the number of chromocenters (average 14.7; Figure 7e) we found a significant difference to postmitotic macrophages (p < 0.001) but not to ES-cells (p > 0.2). This suggests that increased clustering of chromocenters in macrophages occurred during postmitotic terminal differentiation. We previously described a highly significant (p < 0.001) reduction of chromocenter numbers during the differentiation from myoblasts (mean number of 20.4) via post-mitotic myocytes (14.5 chromocenters) to myotubes (11.1 chromocenters) [29]. Figure 6 3D-FISH with the mouse major satellite DNA probe. For each cell type, maximum intensity xy-projections (top) and 3D-reconstructions (bottom) are shown. In the projections, DNA counterstain (TO-PRO3) is shown in red and FISH signals are false-colored in green. Overlap of both signals in chromocenters leads to the intense yellow color. Scale bars represent 5 μm in the respective projection. The 3D-reconstructions (not to scale) are shown together with xy, xz and yz maximum intensity projections in the background. Chromocenters in the reconstructions are shown as solid green structures, while the nuclear border is presented as a transparent shell. Note the differences in number and size of chromocenters in the various cell types and the differences in nuclear shape. Figure 7 Numbers of chromocenters per nucleus. Cell types are as indicated. mv = mean value, the standard deviation and the number of nuclei (n) are also given. In contrast to human cell nuclei, chromocenters in mouse cell nuclei are easily identifiable by their extremely bright fluorescence in formaldehyde-fixed, structurally preserved cells counterstained with DAPI or TO-PRO-3 (Figure 2). Our images revealed that these bright areas were identical with the chromocenter FISH signal (Figure 6), except for occasional small parts of counterstained chromatin which escaped detection by FISH. This finding opened the opportunity to investigate the relative radial distribution of chromocenters in the same nuclei that were used for the analysis of the radial chromosome territory distribution. For this purpose we applied a threshold to segment the nucleus and another much higher threshold to segment the intensely stained chromocenters. In all cell types studied, chromocenters had a more internal average position than total nuclear counterstain (Figure 8). The most internal average position was found in fibroblast nuclei, the most peripheral in lymphocyte nuclei. Chromocenter-DNA in quiescent lymphocytes (G0) showed a tendency for a more internal nuclear location compared with S-phase lymphocytes but the difference was not significant (p = 0.68). Differences between both lymphocyte populations and any other cell type were highly significant (p < 0.001). Chromocenter distribution in ES-cells was significantly different from fibroblasts, myoblasts, macrophages (all p = 0.001 or smaller) and myotubes (p = 0.024). A significant difference was also found between myoblasts and myotubes (p = 0.035). Other comparisons revealed no significant differences (p > 0.1). Results obtained with counterstained chromocenters were confirmed by analysis of FISH-labeled chromocenters (Figure 8). Figure 8 Three-dimensional relative radial distribution of chromocenters (continuous lines) compared to DNA counterstain (broken lines; see legend to Figure 3 and Methods for details). Chromocenters identified by high thresholds of the DNA-counterstain TO-PRO-3 (left) gave results very similar to chromocenters labeled by FISH with a mouse major satellite probe in independent experiments (right). To verify whether chromocenters are in contact with the nuclear border we determined the position of FISH labeled chromocenters (Figure 6) by visual inspection of light optical sections. Each chromocenter was classified to be either in contact with the nuclear border (peripheral), the nucleolus (perinucleolar), both these structures or neither of them (="internal"; Figure 9). In all investigated cell types, the majority of chromocenters were in contact with the nuclear border (64%–97%). A variable fraction was either additionally (14%–37%) or exclusively (3%–30%) touching a nucleolus. The percentage of chromocenters belonging to this perinucleolar fraction varied substantially between cell types (16%–58%). Only a minor fraction of chromocenters was located "internally", i.e. associated neither with the nuclear border nor with a nucleolus (0.3%–6%). Figure 9 Association of chromocenters with the nuclear periphery and the nucleolus. Pie slices represent the percentages of chromocenters at the respective intranuclear locus in the indicated cell type. See main text for details. Nuclear Shapes depend on cell types Nuclei of the cell types investigated in this study differed substantially in shape (Figure 6). While nuclei from lymphocytes and separately growing ES-cells were approximately spherical, nuclei from fibroblasts and myoblasts resembled flat ellipsoids. Nuclear shape differences may be a factor influencing higher order chromatin arrangements. To relate our above results to nuclear shape we determined the nuclear flatness of all analyzed cell types. For this purpose, we measured in a subset of nuclei the length of the longest nuclear axis in xy-projections which was defined as x-axis, the longest axis that was perpendicular to the x-axis (defined as y-axis) and the z-axis (measured on xz and yz projections of the nucleus). The flatness of the nuclei was then calculated according to (√(xy))/z (Table 2). A sphere has a value of 1 whereas larger values are obtained for flat structures. As expected, nuclei of ES-cells and lymphocytes revealed the lowest and the very flat fibroblast nuclei revealed the highest values. The difference observed between S-phase and G0 fibroblast nuclei did not result from a difference in nuclear height but from a reduced xy-size of G0 compared to S-phase nuclei. (Table 2). Notably, nuclei may become more spherical or flatter during cell differentiation. While nuclei from in vitro differentiated macrophages were flatter than those from their ES-cell precursors, myotube nuclei were rounder than those from their myoblast precursors. Table 2 Shape parameters of nuclei in different cell types, sorted by increasing flatness. All lengths are in μm and the standard deviation of the mean is given. See text for details n= x-axis y-axis z-axis flatness (√(xy))/z ES cells (S-phase) 37 11.5 ± 0.7 10.6 ± 0.7 10.9 ± 1.1 1.01 Lymphocytes (G0) 40 9.2 ± 0.4 8.7 ± 0.4 8.6 ± 0.6 1.04 Myotubes (G0) 31 12.5 ± 2.9 8.2 ± 1.0 7.4 ± 1.1 1.37 Macrophages (G0) 33 13.1 ± 1.4 9.7 ± 1.2 5.8 ± 0.7 1.94 Myoblasts (S-phase) 30 16.2 ± 1.9 11.2 ± 1.6 4.9 ± 0.7 2.75 Fibroblasts (G0) 20 16.0 ± 2.7 11.5 ± 1.6 3.7 ± 0.6 3.67 Fibroblasts (S-phase) 20 19.6 ± 2.5 15.4 ± 1.9 3.8 ± 0.5 4.57 Discussion The radial distribution of chromosome territories In lymphocyte nuclei of humans, gene rich chromosome territories were shown to locate to internal regions of the nucleus while gene poor ones are more peripheral [11-13,39,40], a distribution also found for the homologs of human chromosomes 18 and 19 in primates [14]. Our data provide the first report for a gene density dependent radial chromosome territory arrangement in lymphocytes of a non-primate animal, suggesting that this ordering principle in the lymphocyte nucleus has been evolutionary conserved since a common ancestor of mice and humans lived some 87 million years ago [41]. The finding that differences are less pronounced in mouse than in humans is consistent with much smaller differences in chromosomal gene density in the mouse karyotype. The evaluation method used here was previously applied in studies on lymphocytes of humans and other primates and results are thus comparable. The most gene rich mouse chromosome MMU11 (peak at 66% of the nuclear radius, Figure 3a) is not as centrally located as the most gene rich human chromosome, HSA19 (peaks at 40–50%) [12,14,40], or the HSA19 homologs in ten primate species (peaks between 40 and 60%) [14] but it comes close to the second most gene rich human chromosome, HSA17 (peak at 58%) [39], which consists of about 3/4 of sequences syntenic to MMU11 [30]. Gene density of individual chromosomes was not the only theme of radial nuclear order, since in addition we observed a correlation with chromosome size (Figure 3). Interestingly, in mouse lymphocyte nuclei the correlation coefficient was higher for a gene density dependent arrangement while in macrophage nuclei it was higher for size dependent arrangement. This indicates a level of cell type specific differences in chromosome territory arrangements whose functional significance can now be explored. Studies finding the same transgene arrays more internal when transcriptionally active than when inactive [42,43] suggest that "gene density sorting" may be correlated to transcriptional activity rather than gene content per se. Current evidence argues against movement of chromosome territories during interphase but repositioning of chromosomes relative to each other was observed during mitosis [3,4]. A comparison of the radial positioning of two chromosomal subregions between human ES and lymphoblastoid cells revealed a slightly significant difference (p < 0.04) for the p-arm of HSA12 but not the p-arm of HSA6 [44]. In our study, between mouse ES cell and lymphocytes nuclei we found many significant differences in the radial distribution of the six tested chromosomes. Differences in fibroblast and macrophage nuclei were much less pronounced. For human fibroblast nuclei, both, a gene density related distribution [11,13,15] and a chromosome size dependent distribution [12,16,17] have been reported. In a recent study [17], we reconciled these seemingly conflicting data by evidence that both, gene density and size related features of chromosome territory positioning can be observed (see Introduction). Territories of small HSA18 and HSA19 were both found close to the 3D nuclear center in spite of the large differences in gene density between them. HSA1 is about 3.5 times larger than HSA18 and HSA19. The largest ratio in the current study was only 1.64 (MMU1 vs. MMU14). A similar factor is reached for example by the human chromosome combinations HSA1 and HSA8 or HSA12 and HSA18. Both combinations were not found to produce significant radial positioning differences in human fibroblasts [17]. Assuming that chromosome size differences play an important role in chromosome territory positioning in both, human and mouse nuclei, the much smaller size differences between mouse chromosomes compared to human chromosomes may explain the lack of significant radial distribution differences in mouse fibroblasts. Linear regression analysis showed a slightly better fit for a gene-density related distribution than for a size related distribution in this cell type (Figure 3e,f). A chromosome territory distribution related to both, size and gene density was also reported for chicken cell nuclei [7,19]. This fits with the fact that chicken microchromosomes show a much higher gene density than macrochromosomes. In addition, in species ranging from humans and other mammals to chicken, a layer of chromatin at the nuclear periphery and around nucleoli is replicated in mid to late S-phase and consists of gene poor sequences. Gene dense chromatin replicates early in S-phase and is distributed in interior nuclear zones between the perinucleolar and perinuclear compartments [7-9]. Two recent studies found early replicating chromatin also in the interior of Hydra cell nuclei [6] and of micronuclei of a Ciliate [5], while a zone of mid-late replicating chromatin was noted in close association with the nuclear envelope. While it is not known at present whether early and mid-late replicating chromatin in Hydra and Ciliates differ in gene density to the same extend as observed in higher animals, present data support the hypothesis that non-random radial chromatin arrangements have been evolutionary conserved possibly since the formation of the first eukaryotic cells. This hypothesis, if it can be further substantiated, argues for a still unknown adaptive value of this radial order [17,45]. Despite relatively small gene density and size differences between mouse chromosomes we found significant variations in distribution from one cell type to another. The strongest case was provided by the comparison of ES cell nuclei with lymphocyte nuclei where we detected significant differences for MMU9, MMU11 and MMU14. Both mouse cell types have very similar nuclear shapes. We therefore can exclude that these distribution differences are strictly dependent on a single factor, be it nuclear shape, chromosomal size or gene density. More complex mechanisms must therefore be implicated. A possibility that is now open for experimental tests are cell type specific differences of gene expression pattern along a given chromosome. A more similar chromosomal distribution in related cell types than in unrelated ones provides circumstantial evidence for such an assumption [18]. In mouse large and small lung cells the distribution of all tested chromosomes was similar and mouse lymphocytes and myeloblasts showed only one significant difference [18]. In our study we found highly significant differences between two hematopoietic cell types, lymphocytes and macrophages, suggesting that terminal differentiation implies cell type specific changes of chromosome positioning, possibly in response to transcriptional changes. Restrictions for the spatial distribution of chromosome territories may come from the arrangement of specific chromosomal subregions such as pericentromeric heterochromatin which may be involved in the development of cell type specific higher order chromatin arrangements [46]. Arrangements of chromosome territories follow probabilistic rules Radial distributions as discussed above were derived as a mean of the positions found in individual nuclei. In individual nuclei, chromosome territories can occupy a position quite different from this mean, reflecting the dynamic nuclear organization of the genome. Notably, chromosome territories were considerably more variable arranged with respect to each other than with respect to their radial nuclear order. Our data are neither compatible with a general association of homologous chromosomes nor with a separation of the genome in two parental haploid sets. A spatial separation of paternal and maternal haploid chromosome sets together with an antiparallel chromosome order in each set resulting in homologous chromosomes typically positioned opposed to each other was reported [36,47-49]. Other studies, however, did not find evidence for these claims in human cell types [17,50]. The present study provides substantial evidence against separation of paternal and maternal chromosome sets in several mouse cell types. Instead, our data support a very variable distribution of chromosome territories with respect to each other, in agreement with a study of radiation induced chromosome translocations in human lymphocytes [51]. Our findings, however, do not exclude preferential neighborhoods of certain chromatin regions in specific cell types. A number of publications have reported individual examples for such a non-random proximity of particular chromosomes [18,52], centromeres [53,54] or genes [55-59] in some cell types but not in others, including homologous pairing of specific chromosomes in some examples [60-62]. This study suggests a more frequent association of MMU1 in macrophage nuclei compared to other cell types. Organization of chromocenters Our results confirm previous studies on other mouse cell types showing characteristic cell type specific patterns of chromocenter distribution [20,24-26,29,37,63,64]. Similar observations have been made in rat [65] and human cells [44,53,54,60,66,67]. The extend of centromere clustering is, however, also species specific. Human fibroblasts, lymphocytes, and ES cells, revealed more than 30 centromere signals in cycling cells for the 46 human chromosomes [44,67] and thus much less clustering than the respective mouse cell types in our study. In addition to the number of chromocenters, the present study also provides data about their radial distribution, their association with the nuclear border or the nucleolus and the shape of the harboring nuclei. As for chromosome territories, we found common themes. Cell types with spherical nuclei revealed a more peripheral relative radial distribution of chromocenters while flat nuclei showed a more internally located one. With the exception of gene richest MMU11 in some cell types, radial distributions of investigated chromosome territories were more peripheral than the distributions of counterstained nuclear DNA (Figure 3a,d,g,j,m), including chromosomes MMU2 and MMU9, the fifth and sixth gene richest chromosomes in mouse. This raised the question which chromosomes or parts thereof account for the internally located DNA. The 14 chromosome pairs not investigated in this study including six pairs of NOR bearing chromosomes come into question as well as chromosome regions not detected by FISH with chromosome paint probes. In chromosome painting experiments, repetitive sequences that would cross-hybridize to other chromosomes are suppressed. As a consequence, tandem repetitive sequences contained in centromeric and pericentromeric regions stay unlabeled. Indeed, chromocenters were more internal than the average painted chromosome territories from the same cell nuclei and also than total counterstained DNA. Considering the large size of chromocenters, this finding is compatible with the observation that in all cell types 64% – 97% of the chromocenters touched the nuclear border. 3D-reconstructions (Figure 6) illustrate several examples where chromocenters touch the nuclear border but also reach deep into the nuclear interior. This figure also suggests a reason for the tendency of flat cells to have more and smaller chromocenters (Table 2, Figure 7). The average chromocenter in fibroblast nuclei contained pericentromeric regions from two chromosomes. For geometrical reasons, the number of chromosome territories of which centromeres can associate within single chromocenters may be more constrained in flat nuclei compared to spherical nuclei. Nuclear shape however cannot be the only reason for differences of higher order chromatin arrangements between cell types since nuclei with similar shape but from different cell types such as ES cells and lymphocytes show marked differences. Also, the number of chromocenters is not always larger in flatter nuclei. When ES cells were in vitro differentiated to macrophages their flatness increased while the number of chromocenters decreased. The finding that in this case centromere clustering happens during a postmitotic stage, argues for a differentiation related process. In the differentiation pathways investigated in the present study (myoblasts to myotubes and ES cells to macrophages), we found a decrease in the number of chromocenters. Such a relation was noted in an early study using Giemsa staining on different tissues of mouse [37] and also found in in vitro differentiation experiments [29,65,66]. Generally, non-cycling cells often show fewer chromocenters than their cycling counterparts [33,67]. The extreme case is reached in certain neuronal cells of the mouse retina were all centromeres cluster into a single chromocenter (I. Solovei, personal communication). Our observations suggest, however, that cell differentiation in other cases may also imply a de-clustering of centromeres. Fibroblast and myoblasts nuclei showed larger numbers of chromocenters than ES cell nuclei. More direct evidence is available for postmitotic mouse Purkinje neurons where clustering of centromeric regions is dynamic during postnatal development. After a transient increase in clustering combined with a more central location 3 days after birth, a fraction of centromeric regions split up again together with some centromere movements back to the nuclear periphery [23-25]. Conclusion We report common themes of higher order chromatin arrangement as well as cell type specific differences in mouse cell nuclei. A common theme detected here as well as in previous studies of human cell nuclei [11-13] is the preferential radial distribution of chromosome territories, that describes the distance of territories to the nuclear center. In both, human and mouse lymphocyte nuclei, gene rich chromosome territories are distributed to more internal regions than gene poor chromosome territories, indicating evolutionary conservation of this ordering principle at least since the separation of primate and rodent ancestors. In all investigated mouse cell types, we observed a tendency for such a gene density dependent distribution of chromosome territories as well as a preference of large territories to be more peripheral than small ones. Cell type specific differences however were noted with respect to the predominance of gene density or size related correlations. In addition, individual chromosome territories showed cell type specific variations in radial distribution. Cell type specific higher order chromatin arrangements could not be explained by differences in nuclear shape and thus other yet unknown factors must be implicated. In contrast to the radial distribution of chromosome territories in the nucleus, their side-by-side arrangements (neighborhoods) were highly variable. Our data are not compatible with a reported model of separation of haploid parental chromosome sets with an antiparallel order of chromosomes [36]. Depending on cell type, clustering of centromeric regions into larger chromocenters was either increased or decreased compared to precursor cells. In general, we found stronger clustering in further differentiated cells as well as in spherical nuclei when compared to flat nuclei but exceptions occurred. Cell type dependent variations also included differences in radial nuclear distribution of chromocenters. A common theme was contact of a majority of chromocenters with the nuclear border. Materials and methods Cell culture, fixation procedure and FISH-pretreatments EB-5 ES-cells (40, XY) were cultivated in DMEM with 15% FCS (tested for ES-cells) with additional supplements as described elsewhere [68] under 5% CO2. The ES-cells grew in gelatinized flasks without feeder cells. For 3D-preparations, glass cover slips were coated with gelatine (pork skin gelatine, Sigma, Deisenhofen, Germany) by incubation with a 1% solution in water for 20 min and air drying. ES-cell suspension was incubated for 1 h to allow attachment. When ES cells grow on a surface for extended periods of time they start to form colonies in which the cells can have nuclei of highly irregular shape. For technical reasons we limited our evaluations to single cells with round nuclei. Differentiation of ES cells to macrophages was started by co-cultivation of ES cells on OP9 stroma cells [69] as described in [70]. At day 8 of differentiation, suspension cells were transferred to cell culture flasks using medium containing macrophage colony-stimulating factor (MCSF) and interleukin 3 (IL-3). Cytokines were obtained by cultivation of L-cells and X63 AG-653 cells, transgenically expressing and secreting M-CSF or IL-3, respectively [71]. On day 12, the culture contained many adherent macrophages. Cells were transferred onto glass coverslips and fixed the following day. Terminally differentiated macrophages were identified by detection of the surface antigen CD11b, by cell shape and by the absence of BrdU incorporation (see below). Mouse embryonic fibroblasts (40, XX and 40, XY, kindly provided by Dr. Alexander Pfeifer, Institut für Pharmakologie, Ludwig-Maximilians-Universität München) were cultured in DMEM with 10% fetal bovine serum under 5% CO2 to 80% confluency on glas coverslips. Mouse lymphocytes from pooled peripheral blood (kindly provided by Dr. Manuela Mohr, Lehrstuhl für molekulare Tierzucht und Biotechnologie, Ludwig-Maximilians-Universität München) were isolated on a Ficoll gradient. Cultivation was in RPMI with 15% fetal bovine serum. Stimulation was with 12 μg/ml concanavalin A for 72 h. After centrifugation, cells were resuspended in 50% FCS/ 50% RPMI. Glass cover slips (18 × 18 mm, 170 μm thick) were coated with poly-L-lysine (MW 300 000, Sigma, Deisenhofen, Germany) by incubation with a 0.1 mg/ml solution for 40 min, washed with water and air-dried. The cell suspension was incubated for 1 h or longer to allow for attachment. Pmi28 primary mouse myoblasts were kindly provided by A. Starzinski-Powitz [72] and cultured and differentiated to myotubes as described [73]. For the identification of cells in S-phase, BrdU at a final concentration of 5 μM was added to the culture medium 30–60 min before fixation except for macrophages and myotubes where incubation time was 24 h. Fixation was performed with 4 % formaldehyde freshly made from paraformaldehyde [74] and buffered in PBS for 10 min. For ES-cells, macrophages, fibroblasts and myotubes, formaldehyde was in 0.75 × PBS, for myoblasts in 1 × PBS. Lymphocytes were incubated in 0.3 × PBS for 40 sec prior to fixation and also fixed in 0.3 × PBS to prevent shrinkage of the nucleus that otherwise occurs in this cell type. Permeabilization steps for all cells included 10 min in 0.5% Triton-X 100, 60 min. incubation in 20% glycerol in PBS followed by five freezing/thawing cycles in liquid nitrogen and a 10 min incubation in 0.1 M HCl. Slides were kept at 4°C in 50% formamid/2 × SSC until hybridization. Air-drying of nuclei was carefully avoided at all steps. To avoid obstruction due to mixed results from active and inactive X-chromosomes, we investigated only active X-chromosomes from male cells. ES cells, unstimulated lymphocytes and myoblasts were from male sources, thus they as well as in vitro differentiated macrophages and myotubes contained only an active X-chromosome. Fibroblasts and stimulated lymphocytes were from mixed female and male sources. When we labeled X-chromosomes in these cell types, only nuclei with a single one-chromosome-size territory (male cells) were recorded. DNA probes and FISH Mouse chromosome paint probes, produced by DOP-PCR [75] from sorted chromosomes, were kindly provided by N. Carter, Cambridge, UK [76]. Labeling of chromosome paints was done by DOP-PCR using biotin-dUTP or digoxigenin-dUTP. 10 μl of both chromosome paint probes and 80 μl mouse C0t1-DNA (1 μg/μl, Invitrogen) were precipitated and solved in 5 μl deionized formamide. The same volume of 20% dextransulphate in 2 × SSC was added. Simultaneous denaturation of probes and target was at 75°C for 3.5 min. Hybridization was performed at 37°C for 2–3 days. To exclude influences from the labeling scheme we switched biotin and digoxigenin so that half of the evaluated nuclei had one labeling scheme and the other half the other one. FISH with the mouse major satellite specific probe was performed as described [29]. Detection After hybridization washing steps with 2 × SSC at 37°C and 0.1 × SSC at 60°C were performed. Biotin was detected with avidin-Alexa-488 (Molecular Probes, USA) and goat-anti-avidin-FITC (Vector Laboratories, USA). Digoxigenin was detected with rabbit-anti-Dig (Sigma) and goat-anti-rabbit-Cy3 (Amersham Pharmacia, UK). BrdU detection was in PBS with mouse-anti-BrdU (Roche, Mannheim, Germany) and goat-anti-mouse-Alexa-350 (Molecular Probes, Eugene, Ore.). TO-PRO-3 (1 μM; Molecular Probes) was used as a DNA counterstain. Confocal microscopy and image analysis Stacks of optical sections were collected on Leica TCS 4D (100x, N.A 1.4 Plan Apo Objective) and on Zeiss LSM 410 (63x/1.4 Plan Apo) confocal microscopes. Voxel size was 80 nm or below in xy and 240 nm or below in z. Where necessary, individual image stacks were processed with ImageJ [77] e.g. to clip other nuclei from the images. The program used to determine the relative radial distributions of chromosomes and chromocenters is described in detail elsewhere [12]. Briefly, it segments each nucleus in 25 equally spaced "shells". The outermost shell is fitted to the surface of the segmented nucleus and inner shells are adapted accordingly. On any ray from the nuclear center to the surface, each shell has the same width, resulting in increasing volumes for outer shells. The percentage of a given signal in each shell is then calculated. Due to the limited resolution of light microscopy and a Gaussian filtering, the edge of the nucleus does not appear as a sharp border but blurred, with intensity decreasing to zero over a small region. Nuclear segmentation will include some of it. This is the reason for the decreasing amounts of DNA in the outermost shells in the curves. Angles and distances between chromosome territories were measured with a newly developed program. Thresholds for nuclei and territories were determined interactively. The gravity centers of the resulting objects and the geometrical center of the nucleus were used for calculations. 3D reconstructions shown in Figure 6 were made with AMIRA (TGS Europe, now available from Mercury Computer Systems, Merignac, France). Graphs were made in Microsoft Excel. Final figures were assembled in Adobe Photoshop (Adobe Systems, San Jose, CA, USA). Statistical analysis To determine whether differences between relative radial distributions were significant, we used the median of the distribution in each nucleus. These and other values like angles were compared using the two-sided Kolmogorov-Smirnov test in the Software Package SPSS 12 (SPSS Inc., Chicago, Ill.). Linear regression analysis was also performed in SPSS. Authors' contributions This study was conceived and supervised by SD and TC. RM and AB performed the experiments and quantitative evaluation. TS helped setting up the ES cell differentiation. JvH wrote the quantitative evaluation programs which were embedded in a shell script environment by SD. Statistical evaluation was performed by AB, RM and SD. The manuscript was written by SD with contributions from TC and help from all coauthors. Supplementary Material Additional File 1 All median values of the relative radial distribution of chromosome territories in individual nuclei (see legend to Figure 3). Values in this text file mediantable.txt are separated by tabs. Click here for file Acknowledgements We are grateful to scientists who have provided research material for this study as detailed in the methods section. We thank our colleagues Irina Solovei and Christian Lanctôt for critical reading of the manuscript. This work was supported by the Deutsche Forschungsgemeinschaft by a grant to T. Cremer and S. Dietzel (Cr59/21-1). ==== Refs Schardin M Cremer T Hager HD Lang M Specific staining of human chromosomes in Chinese hamster x man hybrid cell lines demonstrates interphase chromosome territories Hum Genet 1985 71 281 287 2416668 10.1007/BF00388452 Manuelidis L Individual interphase chromosome domains revealed by in situ hybridization Hum Genet 1985 71 288 293 3908288 10.1007/BF00388453 Lucas JN Cervantes E Significant large-scale chromosome territory movement occurs as a result of mitosis, but not during interphase Int J Radiat Biol 2002 78 449 455 12065049 10.1080/09553000110097190 Walter J Schermelleh L Cremer M Tashiro S Cremer T Chromosome order in HeLa cells changes during mitosis and early G1, but is stably maintained during subsequent interphase stages J Cell Biol 2003 160 685 697 12604593 10.1083/jcb.200211103 Postberg J Alexandrova O Cremer T Lipps HJ Exploiting nuclear duality of ciliates to analyze topological requirements for DNA replication and transcription J Cell Sci 2005 118 3973 3983 16129882 10.1242/jcs.02497 Alexandrova O Solovei I Cremer T David CN Replication labeling patterns and chromosome territories typical of mammalian nuclei are conserved in the early metazoan Hydra Chromosoma 2003 112 190 200 14615892 10.1007/s00412-003-0259-z Habermann FA Cremer M Walter J Kreth G von Hase J Bauer K Wienberg J Cremer C Cremer T Solovei I Arrangements of macro- and microchromosomes in chicken cells Chromosome Res 2001 9 569 584 11721954 10.1023/A:1012447318535 O'Keefe RT Henderson SC Spector DL Dynamic organization of DNA replication in mammalian cell nuclei: spatially and temporally defined replication of chromosome-specific alpha-satellite DNA sequences J Cell Biol 1992 116 1095 1110 1740468 10.1083/jcb.116.5.1095 Sadoni N Langer S Fauth C Bernardi G Cremer T Turner BM Zink D Nuclear organization of mammalian genomes. Polar chromosome territories build up functionally distinct higher order compartments J Cell Biol 1999 146 1211 1226 10491386 10.1083/jcb.146.6.1211 Mayr C Jasencakova Z Meister A Schubert I Zink D Comparative analysis of the functional genome architecture of animal and plant cell nuclei Chromosome Res 2003 11 471 484 12971723 10.1023/A:1024978711705 Croft JA Bridger JM Boyle S Perry P Teague P Bickmore WA Differences in the Localization and Morphology of Chromosomes in the Human Nucleus J Cell Biol 1999 145 1119 1131 10366586 10.1083/jcb.145.6.1119 Cremer M von Hase J Volm T Brero A Kreth G Walter J Fischer C Solovei I Cremer C Cremer T Non-random radial higher-order chromatin arrangements in nuclei of diploid human cells Chromosome Res 2001 9 541 567 11721953 10.1023/A:1012495201697 Boyle S Gilchrist S Bridger JM Mahy NL Ellis JA Bickmore WA The spatial organization of human chromosomes within the nuclei of normal and emerin-mutant cells Hum Mol Genet 2001 10 211 219 11159939 10.1093/hmg/10.3.211 Tanabe H Müller S Neusser M von Hase J Calcagno E Cremer M Solovei I Cremer C Cremer T Evolutionary conservation of chromosome territory arrangements in cell nuclei from higher primates Proc Natl Acad Sci U S A 2002 99 4424 4429 11930003 10.1073/pnas.072618599 Bridger JM Boyle S Kill IR Bickmore WA Re-modelling of nuclear architecture in quiescent and senescent human fibroblasts Curr Biol 2000 10 149 152 10679329 10.1016/S0960-9822(00)00312-2 Sun HB Shen J Yokota H Size-dependent positioning of human chromosomes in interphase nuclei Biophys J 2000 79 184 190 10866946 Bolzer A Kreth G Solovei I Köhler D Saracoglu K Fauth C Müller S Eils R Cremer C Speicher MR Cremer T Three-dimensional maps of all chromosome positions demonstrate a probabilistic order in human male fibroblast nuclei and prometaphase rosettes PLoS Biol 2005 3 e157 15839726 10.1371/journal.pbio.0030157 Parada L McQueen P Misteli T Tissue-specific spatial organization of genomes Genome Biology 2004 5 R44 15239829 10.1186/gb-2004-5-7-r44 Stadler S Schnapp V Mayer R Stein S Cremer C Bonifer C Cremer T Dietzel S The architecture of chicken chromosome territories changes during differentiation BMC Cell Biol 2004 5 44 15555075 10.1186/1471-2121-5-44 Manuelidis L Different central nervous system cell types display distinct and nonrandom arrangements of satellite DNA sequences Proc-Natl-Acad-Sci-U-S-A 1984 81 3123 3127 6587343 van Dekken H Hulspas R Spatial analysis of intranuclear human repetitive DNA regions by in situ hybridization and digital fluorescence microscopy Histochem-J 1993 25 173 182 8473197 10.1007/BF00163812 Dietzel S Weilandt E Eils R Münkel C Cremer C Cremer T Three-dimensional distribution of centromeric or paracentromeric heterochromatin of chromosomes 1, 7, 15 and 17 in human lymphocyte nuclei studied with light microscopic axial tomography Bioimaging 1995 3 121 133 10.1002/1361-6374(199509)3:3<121::AID-BIO3>3.3.CO;2-5 Manuelidis L Indications of centromere movement during interphase and differentiation Ann N Y Acad Sci 1985 450 205 221 3860180 Martou G de Boni U Nuclear topology of murine, cerebellar Purkinje neurons: changes as a function of development Exp Cell Res 2000 256 131 139 10739660 10.1006/excr.1999.4793 Solovei I Grandi N Knoth R Volk B Cremer T Postnatal changes of pericentromeric heterochromatin and nucleoli in postmitotic Purkinje cells during murine cerebellum development Cytogenetics and Genome Research 2004 105 142 150 Kim SH McQueen PG Lichtman MK Shevach EM Parada LA Misteli T Spatial genome organization during T-cell differentiation Cytogenet Genome Res 2004 105 292 301 15237218 10.1159/000078201 Brown KE Guest SS Smale ST Hahm K Merkenschlager M Fisher AG Association of transcriptionally silent genes with Ikaros complexes at centromeric heterochromatin Cell 1997 91 845 854 9413993 10.1016/S0092-8674(00)80472-9 Brown KE Amoils S Horn JM Buckle VJ Higgs DR Merkenschlager M Fisher AG Expression of alpha- and beta-globin genes occurs within different nuclear domains in haemopoietic cells Nat Cell Biol 2001 3 602 606 11389446 10.1038/35078577 Brero A Easwaran HP Nowak D Grunewald I Cremer T Leonhardt H Cardoso MC MeCP2 induces large-scale chromatin reorganization during terminal differentiation J Cell Biol 2005 169 733 743 15939760 10.1083/jcb.200502062 Mouse data based on NCBI assembly m34, May 2005, Genebuild: Ensembl, Sep 2005. Human data based on NCBI assembly 35, July 2004, Genebuild: Ensembl, Sep 2005 Dev VG Tantravahi R Miller DA Miller OJ Nucleolus organizers in Mus musculus subspecies and in the RAG mouse cell line Genetics 1977 86 389 398 69563 Winking H Nielsen K Gropp A Variable positions of NORs in Mus musculus Cytogenet Cell Genet 1980 26 158 164 6156053 Weimer R Haaf T Krüger J Poot M Schmid M Characterization of centromere arrangements and test for random distribution in G0, G1, S, G2, G1, and early S' phase in human lymphocytes Hum Genet 1992 88 673 682 1551672 10.1007/BF02265296 Ferguson M Ward DC Cell cycle dependent chromosomal movement in pre-mitotic human T-lymphocyte nuclei Chromosoma 1992 101 557 565 1521500 10.1007/BF00660315 Vourc'h C Taruscio D Boyle AL Ward DC Cell cycle-dependent distribution of telomeres, centromeres, and chromosome-specific subsatellite domains in the interphase nucleus of mouse lymphocytes Exp Cell Res 1993 205 142 151 8453988 10.1006/excr.1993.1068 Nagele R Freeman T McMorrow L Lee HY Precise spatial positioning of chromosomes during prometaphase: evidence for chromosomal order Science 1995 270 1831 1835 8525379 Hsu TC Cooper JE Mace MLJ Brinkley BR Arrangement of centromeres in mouse cells Chromosoma 1971 34 73 87 4105528 10.1007/BF00285517 Pardue ML Gall JG Chromosomal localization of mouse satellite DNA Science 1970 168 1356 1358 5462793 Weierich C Brero A Stein S von Hase J Cremer C Cremer T Solovei I Three-dimensional arrangements of centromeres and telomeres in nuclei of human and murine lymphocytes Chromosome Res 2003 11 485 502 12971724 10.1023/A:1025016828544 Cremer M Küpper K Wagler B Wizelman L von Hase J Weiland Y Kreja L Diebold J Speicher MR Cremer T Inheritance of gene density-related higher order chromatin arrangements in normal and tumor cell nuclei J Cell Biol 2003 162 809 820 12952935 10.1083/jcb.200304096 Springer MS Murphy WJ Eizirik E O'Brien SJ Placental mammal diversification and the Cretaceous-Tertiary boundary Proc Natl Acad Sci U S A 2003 100 1056 1061 12552136 10.1073/pnas.0334222100 Tumbar T Belmont AS Interphase movements of a DNA chromosome region modulated by VP16 transcriptional activator Nat Cell Biol 2001 3 134 139 11175745 10.1038/35055033 Dietzel S Zolghadr K Hepperger C Belmont AS Differential large-scale chromatin compaction and intranuclear positioning of transcribed versus non-transcribed transgene arrays containing beta-globin regulatory sequences J Cell Sci 2004 117 4603 4614 15331668 10.1242/jcs.01330 Wiblin AE Cui W Clark AJ Bickmore WA Distinctive nuclear organisation of centromeres and regions involved in pluripotency in human embryonic stem cells J Cell Sci 2005 118 3861 3868 16105879 10.1242/jcs.02500 Tanabe H Habermann FA Solovei I Cremer M Cremer T Non-random radial arrangements of interphase chromosome territories: evolutionary considerations and functional implications Mutat Res 2002 504 37 45 12106644 Manuelidis L A view of interphase chromosomes Science 1990 250 1533 1540 2274784 Nagele RG Freeman T Fazekas J Lee KM Thomson Z Lee HY Chromosome spatial order in human cells: evidence for early origin and faithful propagation Chromosoma 1998 107 330 338 9880766 10.1007/s004120050315 Koss LG Characteristics of chromosomes in polarized normal human bronchial cells provide a blueprint for nuclear organization Cytogenet Cell Genet 1998 82 230 237 9858826 10.1159/000015109 Nagele RG Freeman T McMorrow L Thomson Z Kitson-Wind K Lee H Chromosomes exhibit preferential positioning in nuclei of quiescent human cells J Cell Sci 1999 112 525 535 9914164 Allison DC Nestor AL Evidence for a relatively random array of human chromosomes on the mitotic ring J Cell Biol 1999 145 1 14 10189364 10.1083/jcb.145.1.1 Cornforth MN Greulich-Bode KM Loucas BD Arsuaga J Vázquez M Sachs RK Brückner M Molls M Hahnfeldt P Hlatky L Brenner DJ Chromosomes are predominantly located randomly with respect to each other in interphase human cells J Cell Biol 2002 159 237 244 12403811 10.1083/jcb.200206009 Parada LA McQueen PG Munson PJ Misteli T Conservation of relative chromosome positioning in normal and cancer cells Curr Biol 2002 12 1692 1697 12361574 10.1016/S0960-9822(02)01166-1 Alcobia I Dilão R Parreira L Spatial associations of centromeres in the nuclei of hematopoietic cells: evidence for cell-type-specific organizational patterns Blood 2000 95 1608 1615 10688815 Alcobia I Quina AS Neves H Clode N Parreira L The spatial organization of centromeric heterochromatin during normal human lymphopoiesis: evidence for ontogenically determined spatial patterns Exp Cell Res 2003 290 358 369 14567993 10.1016/S0014-4827(03)00335-5 Lukasova E Kozubek S Kozubek M Kjeronska J Ryznar L Horakova J Krahulcova E Horneck G Localisation and distance between ABL and BCR genes in interphase nuclei of bone marrow cells of control donors and patients with chronic myeloid leukaemia Hum Genet 1997 100 525 535 9341866 10.1007/s004390050547 Neves H Ramos C da Silva MG Parreira A Parreira L The nuclear topography of ABL, BCR, PML, and RARalpha genes: evidence for gene proximity in specific phases of the cell cycle and stages of hematopoietic differentiation Blood 1999 93 1197 1207 9949162 Nikiforova MN Stringer JR Blough R Medvedovic M Fagin JA Nikiforov YE Proximity of chromosomal loci that participate in radiation-induced rearrangements in human cells Science 2000 290 138 141 11021799 10.1126/science.290.5489.138 Roix JJ McQueen PG Munson PJ Parada LA Misteli T Spatial proximity of translocation-prone gene loci in human lymphomas Nat Genet 2003 34 287 291 12808455 10.1038/ng1177 Spilianakis CG Lalioti MD Town T Lee GR Flavell RA Interchromosomal associations between alternatively expressed loci Nature 2005 435 637 645 15880101 10.1038/nature03574 Arnoldus EPJ Peters ACB Bots GTAM Raap AK van der Ploeg M Somatic pairing of chromosome 1 centromeres in interphase nuclei of human cerebellum Hum-Genet 1989 83 231 234 2793166 10.1007/BF00285162 Arnoldus EP Noordermeer IA Peters AC Raap AK Van der Ploeg M Interphase cytogenetics reveals somatic pairing of chromosome 17 centromeres in normal human brain tissue, but no trisomy 7 or sex-chromosome loss Cytogenet Cell Genet 1991 56 214 216 2055120 Chandley AC Speed RM Leitch AR Different distributions of homologous chromosomes in adult human Sertoli cells and in lymphocytes signify nuclear differentiation. J Cell Sci 1996 109 773 776 8718668 Choh V De Boni U Spatial repositioning of centromeric domains during regrowth of axons in nuclei of murine dorsal root ganglion neurons in vitro J Neurobiol 1996 31 325 332 8910790 10.1002/(SICI)1097-4695(199611)31:3<325::AID-NEU5>3.0.CO;2-C Cerda MC Berríos S Fernández-Donoso R Garagna S Redi C Organisation of complex nuclear domains in somatic mouse cells Biol Cell 1999 91 55 65 10321023 10.1016/S0248-4900(99)80031-8 Chaly N Munro SB Centromeres reposition to the nuclear periphery during L6E9 myogenesis in vitro Exp Cell Res 1996 223 274 278 8601404 10.1006/excr.1996.0082 Beil M Durschmied D Paschke S Schreiner B Nolte U Bruel A Irinopoulou T Spatial distribution patterns of interphase centromeres during retinoic acid-induced differentiation of promyelocytic leukemia cells Cytometry 2002 47 217 225 11933011 10.1002/cyto.10077 Solovei I Schermelleh L Düring K Engelhardt A Stein S Cremer C Cremer T Differences in centromere positioning of cycling and postmitotic human cell types Chromosoma 2004 112 410 423 15197559 10.1007/s00412-004-0287-3 Faust N Huber MC Sippel AE Bonifer C Different macrophage populations develop from embryonic/fetal and adult hematopoietic tissues Exp Hematol 1997 25 432 444 9168065 Nakano T Kodama H Honjo T Generation of lymphohematopoietic cells from embryonic stem cells in culture Science 1994 265 1098 1101 8066449 Schroeder T Fraser ST Ogawa M Nishikawa S Oka C Bornkamm GW Honjo T Just U Recombination signal sequence-binding protein Jkappa alters mesodermal cell fate decisions by suppressing cardiomyogenesis Proc Natl Acad Sci U S A 2003 100 4018 4023 12655061 10.1073/pnas.0438008100 Faust N Bonifer C Wiles MV Sippel AE An in vitro differentiation system for the examination of transgene activation in mouse macrophages DNA Cell Biol 1994 13 901 907 7917012 Kaufmann U Kirsch J Irintchev A Wernig A Starzinski-Powitz A The M-cadherin catenin complex interacts with microtubules in skeletal muscle cells: implications for the fusion of myoblasts J Cell Sci 1999 112 55 68 9841904 Cardoso MC Joseph C Rahn HP Reusch R Nadal-Ginard B Leonhardt H Mapping and use of a sequence that targets DNA ligase I to sites of DNA replication in vivo J Cell Biol 1997 139 579 587 9348276 10.1083/jcb.139.3.579 Dernburg AF Sedat JW Berrios M Mapping three-dimensional chromosome architecture in situ Methods in Cell Biology: Nuclear structure and function 1998 53 , Academic Press 187 233 9348510 Telenius H Pelmear AH Tunnacliffe A Carter NP Behmel A Ferguson-Smith MA Nordenskjold M Pfragner R Ponder BA Cytogenetic analysis by chromosome painting using DOP-PCR amplified flow-sorted chromosomes Genes Chromosomes Cancer 1992 4 257 263 1382568 Rabbitts P Impey H Heppell-Parton A Langford C Tease C Lowe N Bailey D Ferguson-Smith M Carter N Chromosome specific paints from a high resolution flow karyotype of the mouse Nat Genet 1995 9 369 375 7795642 10.1038/ng0495-369 ImageJ ImageJ	16336643	PMC1325247	CC BY	2021-01-04 16:31:30	no	BMC Cell Biol. 2005 Dec 7; 6:44	utf-8	BMC Cell Biol	2,005	10.1186/1471-2121-6-44	oa_comm
==== Front Aust New Zealand Health PolicyAustralia and New Zealand Health Policy1743-8462BioMed Central London 1743-8462-2-281630574210.1186/1743-8462-2-28ResearchRecent developments in targeting access to high cost medicines in Australia Lu Christine Y [email protected] Jan [email protected] Kenneth M [email protected] Richard O [email protected] School of Medical Sciences, University of New South Wales, Sydney, Australia2 Department of Clinical Pharmacology and Toxicology, St Vincent's Hospital, Sydney, Australia3 School of Public Health and Community Medicine, University of New South Wales, Sydney, Australia2005 23 11 2005 2 28 28 9 8 2005 23 11 2005 Copyright © 2005 Lu et al; licensee BioMed Central Ltd.2005Lu et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background In Australia, the Pharmaceutical Benefits Scheme (PBS) has developed a set of arrangements to control access to high-cost medicines to ensure their use is cost-effective. These medicines include the tumour necrosis factor-alpha inhibitors (TNFIs) for the treatment of rheumatoid arthritis. The aim of this first phase of a qualitative study was to explore basic views on the restricted access to TNFIs in order to confirm where further investigation should take place in the next phase. Methods Semi-structured interviews were conducted in 2004 with a member of the four relevant stakeholder groups. Participants were asked their opinions about features of the establishment, process and effects of the system of restricted access to TNFIs. Views on the collaboration between stakeholder groups in the decision-making process were also collected. Results The principle of 'controlled access' to TNFIs was supported in general. There were concerns regarding some of the specific eligibility criteria. Wider and more transparent stakeholder consultation was judged desirable. Some flexibility around prescribing of TNFIs by physicians, and regular review of the arrangements were proposed. These themes will inform the next phase of the study. Conclusion This first phase highlighted a range of issues associated with the PBS arrangements restricting access to TNFIs. Timely review and report of issues and concerns associated with such policy developments that arose in practice are essential. There is a need for a more comprehensive exploration across a wide range of stakeholders with different perspectives that will in turn be helpful in guiding policy and practice around national arrangements to manage access to high-cost medicines. Access to medicinestumour necrosis factor inhibitorspharmaceutical benefits schemedrug reimbursement ==== Body Background Expenditure on pharmaceuticals has increased dramatically in all countries both in the developed and the developing world in recent years [1]. Provision of an increasing number of effective but expensive pharmaceuticals by drug reimbursement systems is challenging because the costs per patient are high. Payers in the public and private health systems face similar challenges due to increasing consumer expectations to access these medicines in the context of substantive cost constraints. The Pharmaceutical Benefits Scheme (PBS) is Australia's national program which provides subsidised access to prescriptions medicines for the community [2]. Decisions on drug reimbursement ('listing') under the PBS are based on assessment by the Pharmaceutical Benefits Advisory Committee (PBAC), which evaluates efficacy, safety and incremental cost-effectiveness of new pharmaceutical products compared to other existing treatments [3]. Australia was the first country to introduce an explicit requirement for economic analysis to select pharmaceuticals for a publicly funded formulary [4]. This system has attracted considerable attention worldwide as Australian pharmaceutical prices are markedly lower than those in other comparable countries [5]. The national goal, as expressed in Australia's National Medicines Policy and, in particular, the Quality Use of Medicines component, states that limited resources should be utilised in such a way that there is provision of needed, effective and safe medicines that are affordable for the individual in order to achieve optimal health outcomes [6]. Biological agents licensed in Australia for treating rheumatoid arthritis include three tumour necrosis factor-alpha inhibitors (TNFIs), etanercept (Enbrel®), infliximab (Remicade®), and adalimumab (Humira®), and an interleukin-1 receptor antagonist, anakinra (Kineret®). These new medicines markedly reduce disease activity in a majority of patients, but there are concerns about their high cost (approximately AUD$20,000 per patient per year), in particular given the annual growth in government expenditure on pharmaceuticals averaging 10.5% between 1992–93 and 2002–03 (representing an increase from AUD$1.88 billion to AUD$5.12 billion) [7]. Further, there are uncertainties regarding their long-term safety, including serious and opportunistic infections and the risk of lymphoma [8,9]. The PBS has evolved a set of arrangements to control access to high-cost medicines in an attempt to maintain the viability of the PBS. Representative, but most innovative of this set of arrangements are those established for the TNFIs, implemented since August 2003. The decision to subsidise TNFIs was based on a collaborative decision-making model to enable the listing of the TNFIs on the PBS. This involved the PBAC which consulted the relevant key stakeholders: the respective pharmaceutical companies, and rheumatologists via the Australian Rheumatology Association Therapeutics Subcommittee. The arrangements for access under the PBS (PBS-restrictions), formulated to secure the most cost-effective use of these expensive agents, were agreed upon after extensive negotiations [3,10]. The respective consumer organisation, the Arthritis Foundation of Australia, also coordinated a strong lobbying effort by consumer representatives to the government. This collaborative model has set a new paradigm for future PBS decisions. Subsidised access was limited to a subset of patients whose disease "has not been adequately controlled using conventional anti-rheumatic drugs and these patients must meet strict criteria for both starting and continuing biological therapy" [11]. Patients are required to sign a Patient Acknowledgement Form which specifies that continuation of therapy beyond four months "will only be approved if objective, substantial response is achieved". Prescribing rights are limited to specialists with expertise in the management of rheumatoid arthritis. Also agreed upon was a risk-mitigation arrangement between the government and the sponsors that established a ceiling level for government outlays annually [10]. Similarly complex arrangements for access are being applied to other high-cost medicines such as imatinib (Glivec®) for the treatment of patients with chronic myeloid leukaemia [11]. Decision makers' and stakeholders' perspectives on prioritising decisions about drug reimbursement and decision-making processes in comparable countries such as Finland, Canada, and the United Kingdom have been described [12-14]. However, to the best of our knowledge, there are no published data on perceptions regarding approaches used in Australia, such as arrangements for access to high-cost medicines operated through the PBS, or studies examining stakeholder consultation processes to reach consensus on arrangements for subsidised access to high-cost medicines. We considered that careful examination of the recent developments in targeting access to high-cost medicines, using TNFIs as an example, would be instructive in informing the debate concerning the principles and processes that might underpin appropriate and ethical access to expensive pharmaceuticals under the PBS or similar access systems. Stakeholders are defined here as groups of people that have the potential to influence the decisions on the arrangements for access, or those who are affected by the PBS-restrictions. The views, attitudes, concerns, and level of support for these arrangements by the stakeholders are critical determinants for a successful implementation of the arrangements. Qualitative techniques are useful to explore perceptions and experiences across a range of relevant stakeholders with respect to restricted access to TNFIs, and to understand the effects of the access criteria when implemented in practice. From a policy perspective, issues and concerns that arise in practice need to be reviewed in a timely manner thereby enabling appropriate management of any implications. Research guiding policy development can often best be undertaken as an iterative approach with each phase building on the one before, in particular where the impact of implementing new access criteria on practice is unknown. The scope and content of such a study are not specified in advance but are developed iteratively. This paper reports on the initial phase of a qualitative study using an iterative approach. This first phase sought to explore the views of one member of each of the four stakeholder groups with a vested interest in any new policy about provision of medicines, in order to confirm where further exploration should take place in the next iteration. Methods Qualitative techniques were used for data collection and analysis. In depth, semi-structured interviews were conducted in Sydney in 2004 with four individuals to explore four different perspectives about access to TNFIs under the PBS. Interviewees were asked their opinions about features of the establishment, process and effects of the arrangements for targeting access to TNFIs. Views on the decision-making process and collaboration between stakeholder groups were also collected. Due to a finite number of key stakeholder representatives who participated in the PBAC decision-making process that formulated the arrangements for access to TNFIs, they were purposively not invited in this initial phase of the study. Purposeful sampling was used to select participants on the basis of their primary membership of different stakeholder groups with respect to the controlled access to TNFIs: a rheumatologist, a health advisor to the government, an employee of a pharmaceutical company, and a patient (who had used a TNFI). Representation of participants was not sought in this first phase of the study which aims to elicit a basic viewpoint from a member of each stakeholder group. The rheumatologist received an invitation to participate through an opinion-leader rheumatologist, purposively selected to have had experiences prescribing TNFIs through the PBS at the time of the study. The health advisor who participated belonged to a like-group of the PBAC, and the pharmaceutical employee was invited to give views of the industry but was not from a company that marketed TNFIs. The patient was nominated by a rheumatologist. Participant information sheets and consent forms were provided to all participants prior to the interviews. These four interviews were conducted by one trained researcher (CL); interviews were of 45 to 60 minutes duration. Interviews were recorded and transcribed verbatim. The transcripts were coded for major concepts. Categories related to the discussion topics were created. Coded segments were then analysed and categorised thematically, thereby developing a thematic framework. NVivo software 2.0 was used to help manage the qualitative data. The transcripts were also coded independently by another investigator (RD or KW). The researchers then met to examine the analyses in order to reach agreement on categories and identified themes. Some differences with regard to labelling the themes were found, but agreement was reached on the central meanings. Interviewees were offered opportunities to review an edited interview transcript to check that their views had been accurately represented. Only one interviewee reviewed the transcript. It was accepted without change. The study was approved by the Human Research Ethics Committees of St Vincent's Hospital Sydney, and the University of New South Wales, Australia. Results and Discussion This initial phase identified a range of concerns, as held by individuals from different stakeholder groups, that arose in practice associated with the recent developments to control access to a group of high-cost medicines in Australia. These are imperative issues that decision makers in Australia and other countries need to consider and manage appropriately and in a timely manner. The pre-determined interview topics and major themes that emerged from the data are listed in Table 1. The four major themes are described below with selected quotes from participants. Table 1 Major themes around restricted access to tumour necrosis factor inhibitors Predetermined Interview topics Perceptions of PBS-restrictions on access Experience with application process Pre-PBS collaboration between the different stakeholder groups Post-PBS collaboration between the different stakeholder groups Sources of information Major themes emerging from data Access to medicines Targeting access Review of the PBS-restrictions Implications of access arrangements in practice Effects on the practice of rheumatology Ethical issues Roles and responsibilities PBS Decision-making process Stakeholder consultation process Transparency Information provision 1. Access to medicines 1.1 Targeting access Interviewees expressed an understanding for the finite resources that can be allocated to treat the possible range of diseases. The principle of 'controlled access' to TNFIs was supported in general. All perceived that the purpose of restricting access to TNFIs was to control costs while targeting those patients identified by the access criteria as most likely to derive benefit. "The government is trying to target the therapy to those who need it ... and control their budget." (advisor) The PBS-restrictions on access, including the limiting of prescribing rights to qualified specialists, were considered to be safe and cost-effective, generally evidence-based and appropriate, given the expense and some lingering uncertainty about long-term safety. "TNF inhibitors are fairly new ... there are no long-term data as yet on their use ... they are not without side effects, and because of their expense, trying existing therapies first is a reasonable approach" (rheumatologist) Participants disagreed with some of the specific eligibility criteria which they interpreted as inflexible and not obviously evidence-based. For example, a number of patients with severe rheumatoid arthritis suffered because of the requirement that rheumatoid factor be positive. "The requirement for a rheumatoid factor is a concern because I don't believe that there's any scientific evidence ... and they [application assessors] seem to be inflexible about that criterion ... There are unfortunate people who would qualify on every criteria except rheumatoid factor alone." (rheumatologist) 1.2 Review of the PBS-restrictions Availability of TNFIs through the PBS, although tightly restricted, was welcomed. However, ongoing review of the PBS-restrictions in accordance with emerging new evidence and analysis of utilisation data was seen as both important and necessary. Interviewees believed the key stakeholder representatives who had participated in the initial stakeholder consultation should again participate in these activities. The risk-mitigation agreement should also be reviewed for its effectiveness and value. "There should be some process of review of criteria, after some agreed period, particularly if they're sharing a [financial] risk, to see whether they still continue as appropriate based on new evidence and data of utilisation, etcetera." (advisor) 2. Implications of access arrangements in practice 2.1 Effects on the practice of rheumatology The arrangements for access to TNFIs requiring evidence of exposure to and failure of a number of anti-rheumatic drugs appear to have promoted re-evaluation of previous treatments and responses in patients with insufficiently controlled rheumatoid arthritis. A proportion of patients may have benefited from this process with better control of their disease without reaching the 'disease activity' specified in the eligibility criteria. "It's forced our rheumatologists to re-look at the treatment that patients have already received and to re-evaluate ... in reassessing the patient and aiming to meet the PBS criteria, they actually get their rheumatoid arthritis under control before requiring a biological..." (rheumatologist) 2.2 Ethical issues One of the ethical issues raised by the interviewees was the possible benefit that might be achieved in some patients if the PBS-restrictions would allow a switch to another TNFI if they failed to respond to the first agent. "If patients didn't meet the criteria for continuing treatment and at the time they stopped, were not bad enough to meet the initial criteria for another agent, they can't switch to another agent, because they are required to meet the initial criteria again." (rheumatologist). The original PBS-restrictions did not address this issue, but these have been altered shortly after the completion of this study's initial phase. The PBS introduced an 'interchangeability' rule from December 2004, allowing patients who meet the eligibility criteria for commencing therapy to trial any of the listed TNFIs as well as anakinra, another anti-rheumatic biological agent [11]. This change demonstrates that some review of the PBS-restrictions is occurring. Despite the 'objective' clinical measures upon which continued access to TNFIs is determined, there are many factors influencing the medical condition of an individual. Concerns were raised about withdrawing patients from an effective treatment when they may have just missed the threshold for continued access as well as a lack of flexibility for physicians to make decisions in cases that are equivocal based on their professional judgement. "There's a disparity between the patients' clinical response and joint count and their laboratory parameters ... the inflammatory markers may be staying up when they actually clinically are a lot better." (rheumatologist) "There's got to be a level of discretion there to the experts because they're the guys who know. It's not some bureaucrat." (patient) The application process was seen to be complex, time-consuming and an administrative burden for everyone involved: "The administrative burden for the doctors, for the Health Insurance Commission...on the pharmacists, and also the patients...and the whole being caught up in that bureaucratic tangle." (pharmaceutical employee) Another ethical concern was that a Patient Information Sheet had not been developed to accompany the Patient Acknowledgement Form at the time these interviews took place, i.e. 12 months after the first TNFI was made available under the PBS. "They're [the patients] signing a contract and they should be able to understand the benefits and also the implications...they need to do it in a manner where they have all the information they need to decide." (advisor) 2.3 Roles and responsibilities Broadly, government is expected to provide affordable access to effective medicines for patients with medical needs while using taxpayers' money wisely and appropriately. A related and necessary responsibility of government is to monitor the PBAC process to ensure appropriate assessment and selection of medicines for reimbursement by the PBS. "The government has an obligation to ensure that the processes that the PBAC uses are rigorous, are regularly evaluated, and are well supported." (advisor) The pharmaceutical industry interviewee felt that information provision by industry was a major responsibility and needed particularly when high-cost medicines subject to complex eligibility criteria become available through the PBS. This interviewee considered that updates should be provided to prescribers about the clinical uses of medicines, including the application procedures for PBS-subsidised treatment; and to promote marketing and publicity with the content being concordant with the PBS-restrictions (A consequence of risk-mitigation agreements is a strong negative incentive for inappropriate marketing and publicity leading to use outside approved PBS-restrictions). The risk-mitigation agreement that is part of the arrangements for access to TNFIs is an indication of collaboration between the government and the pharmaceutical company. Physicians have a responsibility to comply with the PBS-restrictions by law [15]. The PBS-restrictions were worded in such stringent and exhaustive detail that they were seen to demonstrate a lack of faith in the prescribers by the PBS. "There seems ... mistrust from the PBS that rheumatologists could have a loophole for the prescribing ... that they're worried that we will be prescribing the drug for patients who don't strictly have rheumatoid arthritis." (rheumatologist) A view was put that if physicians disagree with any of the restrictions, they should lobby through their professional organisation for revision because they are ultimately responsible for their patients' well-being. More reliance on physicians was considered appropriate and desirable. "Prescribers are obliged to adhere to the PBS arrangements... If we don't like the criteria and have concerns with it then the appropriate thing is to lobby through the ARA [Australian Rheumatology Association]..." (rheumatologist) Concerns were voiced about the need to achieve substantial clinical improvement for continuing access at the follow-up assessment. This requirement may have introduced potential perverse incentives to 'fudge' measurements of disease activity in order to ensure continued access to TNFIs. This raises the issue of the responsibility and accountability of prescribers with respect to their legal obligation to comply with the PBS requirements for PBS-subsidised treatments. However, this does place powerful ethical concerns on doctors in cases where they become the agents whereby patients are denied or withdrawn from effective medicines, particularly if the patient's response is near the threshold level of eligibility. The Australia Rheumatology Association was seen to be responsible for reviewing the PBS-restrictions and providing information to their members, i.e. rheumatologists, as well as the patients. "The ARA [Australia Rheumatology Association] has an obligation to provide information to their own members, it's essential that they support with adequate provision of education and advice and support about this new treatment ... and they also have a role to provide education to the public." (rheumatologist) The role and responsibilities of patient organisations were not explored in detail in this initial phase. Consumer organisations could represent patients in the consultation processes, but qualifications of the staff working in such organisations were thought to be important if they are to contribute to the consultation processes or to produce educational information. Consumer organisations were perceived to primarily interact with patients and to disseminate information. "It's difficult to allocate the Arthritis Foundation a more active role in making recommendations, depending on who is actually staffing the Arthritis Foundation and where those opinions are coming from...the Arthritis Foundation is primarily interfacing with the patients." (rheumatologist) The staff at the Health Insurance Commission, a statutory organisation administering the PBS and other health programs [16], was considered in general to be helpful with respect to queries regarding the application procedure for subsidised TNFI treatments. 3. PBS Decision-making process 3.1 Stakeholder consultation process The interviewees were in substantive accord with the initiative instituted with TNFIs, that negotiations between the stakeholders should take place early in the decision-making process to enhance the likelihood that the final arrangements for access to such medicines are reasonable and can be supported. "Negotiations should start early so that there is clear communication between the expectation and the calibre of evidence needed." (advisor) "...so that the final criteria are the correct criteria, the right patient group is being identified, and also the initiation and continuation rules, etcetera, are workable and manageable" (pharmaceutical employee) However, the interviewees had different views about the stakeholder consultation process that had taken place. "The major strength is the negotiations and the partnering." (advisor) "It seems that each group has their own agenda and it's not really like a collaboration but more each trying to put forward their opinion." (rheumatologist) While the representativeness of the Australian Rheumatology Association Therapeutics Sub-committee in the decision-making process was believed to be appropriate, a wider consultation among ordinary members of the association was wanted: "The therapeutics committee should have sought information from the members of the ARA [Australian Rheumatology Association], or have consulted more widely before they made their decisions because it seems that they made their decisions on our behalf but without consulting us." (rheumatologist) On the subject of representing the views of patients in such processes, the rheumatologist was of the opinion that doctors would appropriately represent the views of their patients. On the contrary, the three other participants felt that patients should have a more direct role in the process, which could provide a humane view; and patients should at least be involved in the development of the Patient Acknowledgement Form. "No one really understands what you go through except you ... not even my rheumatologist or other specialist." (patient) "Patient groups should have a seat at the table...they should be represented formally, rather than by the doctors... We need to have a process where consumers, prescribers, people with interest, come together on an equal footing to discuss what the issues are and what the criteria are and can follow the progress." (advisor) There is increasing need for the voices of patients, their carers and the public to be heard because it is appropriate and crucial for acceptance, change and improvement in the process [17-19]. The representativeness of some consumer groups was raised along with concern that some are funded in part by pharmaceutical companies. This concern may be reduced if conflicts of interests are declared openly, which links to the theme of "transparency". 3.2 Transparency Transparency of the PBAC process and their decisions was a recurrent theme in the interviews. Interviewees thought that wider involvement in the consultation process and greater empowerment of stakeholders to contribute to discussion of the issues would be desirable. Disclosure of the rationale behind the PBAC's selection of access criteria, currently not available in the public domain, was considered to be desirable and a helpful improvement to the system. Greater empowerment (and transparency) by provision of appropriate information was proposed. "A public document which summarised why different decisions were made, and the evidence base on which they were made would be a most fruitful thing to help patients to understand why they have potentially limited access to some of these medicines ... that's about empowerment in information." (advisor) In addition, the issue of the requirement for a positive rheumatoid factor has been the subject of debate since the PBS-restrictions for TNFIs were implemented [20,21]. Greater transparency would be helpful to resolve these concerns and gain increased community support. The recent recommendation by PBAC to remove this criterion [22] is a welcome and rational outcome of the lengthy debate on this issue. The interviewees who participated in this study were 'outsiders' to the consultation process that led to establishment of the arrangements for access. Minimal involvement in the decision-making process and the current level of transparency around PBAC's decisions limited their support for some of the eligibility criteria. Further, their minimal awareness of the negotiations that took place between key stakeholders suggests that there is a need for better communication between those key stakeholder representatives with their constituencies. Increased communication within and between the stakeholder groups is also critical to obtain stronger support for the final PBS-restrictions in practice. 4. Information provision The PBS-restrictions to control access to high-cost medicines, exemplified by TNFIs, are particularly complex, and thus adequate provision of information is all the more important. Information from pharmaceutical companies to the prescribers and general practitioners were noted to be abundant, however, interviewees considered that the content needed to be more balanced. In particular, the rheumatologist and the government advisor expressed a lack of faith in the pharmaceutical companies' ability to supply appropriate materials for patients or the public. "It is not the sponsor's responsibility to talk to patients because they will always get it wrong... direct-to-consumer advertising is a crazy way to go..." (advisor) Some concern was also expressed about industry-funded 'educational' functions: "The way it was presented was one-sided, that the government has put all these barriers forward, the barrier nature was highlighted..." (advisor) Interviewees' had concerns about the understanding of the PBS system and the PBAC process among health professionals and in the community. Better provision of information on the structure of the PBS and the PBAC process was proposed. In particular, interviewees noted that doctors are the most appropriate health professionals to inform patients with regards to the PBS-restrictions, and that they therefore need to be better informed about the cost-effective analyses that underpin the PBAC decisions. For example, patients should be made aware that self-funding is an option, although in the case of TNFIs this is impractical for most people. Interviewees proposed that communication about the need to targeting use of highly specialised medicines should extend to (potential) patients and the public generally, and also to general practitioners who could have an important role and contribution in this area. Increased community awareness and understanding of the principle of quality-use-of-medicines should be part of the overall educational strategy. Interviewees had different views on whose responsibility it was to provide information. The rheumatologist considered that information about the efficacy and safety of TNFIs is most important and that the Australian Rheumatology Association would be appropriate to provide ongoing updates. Other proposed providers of information: pharmacists, the Medicare Australia (previously the Health Insurance Commission), the PBS, the National Prescribing Service (an independent educational organisation funded by the Federal Government) [23], and the consumer organisations. Further investigation The initial phase of the study has explored concerns of individuals from the four relevant stakeholder groups. Our study was not designed to support statistical generalisability. Findings have established the basis for the next phase of the study to explore perceptions of a wide range of stakeholders that will be helpful in gaining a comprehensive and broad societal view of these significant developments of controlling access to high-cost medicines. A written interview guide has been developed on the basis of the findings of this phase for use in subsequent interviews. Diversity in sample characteristics and inclusion of individuals who were engaged in the decision-making process will expand understanding of the issues affecting different members of the stakeholder groups and provide some insight into the negotiations during the decision-making process. Interviewees will include government advisors, participants in the stakeholder consultation process that formulated the arrangements for access to TNFIs (who are in a position to comment on what actually occurred and what could be improved), rheumatologists and patients who have gone through the application process, pharmaceutical company personnel, consumer organisation representatives, and government administration staff. We anticipate that this broad range of interviewees will provide deep and rich insights to decision makers regarding the future decision-making process and development of national arrangements to manage access to medicines from the perspectives of enhancing appropriate access to effective and expensive medicines. Conclusion The PBS has developed a set of arrangements to target access to expensive medicines such as the TNFIs in an attempt to sustain this national drug reimbursement system. Findings from this initial phase of the study highlighted a number of issues associated with the restricted access to TNFIs. The principle of 'controlled access' to TNFIs was supported in general by all interviewees each from a different stakeholder group. However, it is clear that some views, such as those concerning patient direct participation or the varying responsibilities of each stakeholder group, were different. The pre-determined questions appeared to adequately explore the key issues. A thematic framework has now been developed on the basis of the major themes that emerged from the interviews. We believe there is a need for a more comprehensive exploration, through a second phase of the study, to elicit different perspectives across a variety of stakeholders, that will in turn be helpful in guiding policy and practice around national arrangements to manage access to high-cost medicines. Australia's Pharmaceutical Benefits Scheme, as part of the National Medicines Policy, is committed to providing access to cost-effective medicines to the community. When resources are constrained, restricting access is an inevitable outcome. Greater and more trust-based communication and discussion among a wide range of stakeholder groups will enhance community awareness and support to limit access to effective medicines. Acceptable but different approaches to control access and manage the associated implications should be explored with the goal of enhancing health outcomes for all. Declaration of competing interests Christine Lu and A/Prof Jan Ritchie have nothing to declare. A/Prof Ken Williams is a member of the Advisory Board to the sponsor for adalimumab. Prof Ric Day is a member of the Advisory Board to sponsors for adalimumab, infliximab, and anakinra in Australia. A/Prof Williams and Prof Day have also been contracted to undertake clinical trials with etanercept, infliximab, adalimumab, and anakinra. Recompense for these activities received was placed in audited hospital trust funds for use in the research activities of the Clinical Pharmacology Department, St Vincent's Hospital, Sydney. Authors' contributions CL, KW, and RD have made substantial contributions to the conception and design of the study, analysis and interpretation of the data; and drafting and revising the manuscript. JR has been involved in advising on the analysis, drafting the manuscript and revising it critically for intellectual content. All authors read and approved the final manuscript. Acknowledgements The authors thank the interviewees for their participation in this research. ==== Refs Organisation for Economic Co-operation and Development (OECD) OECD Health Data: Statistics and indicator for 30 countries 2005 Commonwealth Department of Health and Aged Care The Australian health care system: an outline Canberra 2000 Sansom L The subsidy of pharmaceuticals in Australia: processes and challenges Australian Health Review 2004 28 194 205 15527399 Henry D Economic analysis as an aid to subsidisation decisions: the development of Australian guidelines for pharmaceuticals Pharmacoeconomics 1992 1 54 67 10147039 Productivity Commission International pharmaceutical price differences: research report 2001 Canberra: AusInfo Sansom L The Australian National Medicinal Drug Policy Journal of Quality Clinical Practice 1999 19 31 35 10.1046/j.1440-1762.1999.00291.x Australian Institute of Health and Welfare (AIHW) Health expenditure in Australia 2002–03 AIHW Cat No HWE 27 (Health and Welfare Expenditure Series No 20) 2004 Canberra: AIHW Symmons DP Silman AJ Anti-tumor necrosis factor alpha therapy and the risk of lymphoma in rheumatoid arthritis: no clear answer Arthritis & Rheumatism 2004 50 1703 1706 15188344 10.1002/art.20312 Fleischmann RM Iqbal I Stern RL Considerations with the use of biological therapy in the treatment of rheumatoid arthritis Expert Opinion on Drug Safety 2004 3 391 403 15335295 10.1517/14740338.3.5.391 Lu CY Williams K Day R March L Sansom L Bertouch J Access to high cost drugs in Australia: Risk sharing scheme may set a new paradigm BMJ 2004 329 415 416 15321884 10.1136/bmj.329.7463.415 Commonwealth Department of Health and Ageing: Schedule of Pharmaceutical Benefits for Approved Pharmacists and Medical Practitioners (1 April 2005) Vuorenkoski L Toiviainen H Hemminki E Drug reimbursement in Finland – a case of explicit prioritising in special categories Health Policy 2003 66 169 177 14585516 10.1016/S0168-8510(03)00042-3 Singer PA Martin DK Giacomini M Purdy L Priority setting for new technologies in medicine: qualitative case study BMJ 2000 321 1316 1319 11090513 10.1136/bmj.321.7272.1316 Wirtz V Cribb A Barber N Reimbursement decisions in health policy – extending our understanding of the elements of decision-making Health Policy 2005 73 330 338 16039351 10.1016/j.healthpol.2004.12.002 Australian Government, The National Health (Pharmaceutical Benefits) Regulations 1960 (made under The National Health Act 1953) Medicare Australia Wiseman V Mooney G Berry G Tang KC Involving the general public in priority setting: experiences from Australia Social Science & Medicine 2003 56 1001 1012 12593873 10.1016/S0277-9536(02)00091-6 Cayton H Patient and public involvement Journal of Health Services Research & Policy 2004 9 193 194 15511317 Florin D Dixon J Public involvement in health care BMJ 2004 328 159 161 14726350 10.1136/bmj.328.7432.159 Lu CY Williams KM March L Bertouch JV Day RO Subsidised access to TNF-alpha inhibitors: is the rationale for exclusion of rheumatoid factor negative patients defensible? Medical Journal of Australia 2004 181 457 458 15487967 Mann EL Subsidised access to TNF-alpha inhibitors: is the rationale for exclusion of rheumatoid factor negative patients defensible? Medical Journal of Australia 2005 182 47 48 15651955 Commonwealth Department of Health and Ageing: Recommendations made by the Pharmaceutical Benefits Advisory Committee in March 2005 Weekes LM Mackson JM Fitzgerald M Phillips SR National prescribing service: creating an implementation arm for national medicines policy British Journal of Clinical Pharmacolology 2005 59 113 116	16305742	PMC1325248	CC BY	2021-01-04 16:25:23	no	Aust New Zealand Health Policy. 2005 Nov 23; 2:28	utf-8	Aust New Zealand Health Policy	2,005	10.1186/1743-8462-2-28	oa_comm
==== Front BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-1231630954910.1186/1471-2458-5-123Research ArticleAn economic way of reducing health, environmental, and other pressures of urban traffic: a decision analysis on trip aggregation Tuomisto Jouni T [email protected] Marko [email protected] Centre for Environmental Health Risk Analysis, National Public Health Institute (KTL), P.O. Box 95, FI-70701, Finland2005 25 11 2005 5 123 123 19 8 2005 25 11 2005 Copyright © 2005 Tuomisto and Tainio; licensee BioMed Central Ltd.2005Tuomisto and Tainio; 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 Traffic congestion is rapidly becoming the most important obstacle to urban development. In addition, traffic creates major health, environmental, and economical problems. Nonetheless, automobiles are crucial for the functions of the modern society. Most proposals for sustainable traffic solutions face major political opposition, economical consequences, or technical problems. Methods We performed a decision analysis in a poorly studied area, trip aggregation, and studied decisions from the perspective of two different stakeholders, the passenger and society. We modelled the impact and potential of composite traffic, a hypothetical large-scale demand-responsive public transport system for the Helsinki metropolitan area, where a centralised system would collect the information on all trip demands online, would merge the trips with the same origin and destination into public vehicles with eight or four seats, and then would transmit the trip instructions to the passengers' mobile phones. Results We show here that in an urban area with one million inhabitants, trip aggregation could reduce the health, environmental, and other detrimental impacts of car traffic typically by 50–70%, and if implemented could attract about half of the car passengers, and within a broad operational range would require no public subsidies. Conclusion Composite traffic provides new degrees of freedom in urban decision-making in identifying novel solutions to the problems of urban traffic. ==== Body Background Personal car traffic is one of the major sources of particulate matter (PM), a major pollutant that is estimated to be responsible for 300000 premature deaths every year in the European Union [1]. In Finland, the estimates are 920 deaths and 960000 restricted activity days for all PM [1]. The direct PM emissions from bus traffic were estimated to be responsible for 12 deaths per year in the Helsinki metropolitan area in 1999, but bus emissions accounted for less than one fifth of the total road traffic PM emission [2]. Although cars have become cleaner during recent years, the growth of car traffic and the location of the emission near the breathing zone mean that these emissions are still ranked high among the environmental health hazards. Traffic is also a major source of CO2 and some other greenhouse gases. In the Helsinki metropolitan area, there are about 4300 traffic accidents that kill 25 persons and injure 1300 persons every year [3]. The health and material costs in the metropolitan area are approximately 1 million € per day (227 million € per year in Helsinki alone [4]). In Finland, traffic accidents are the second most important cause of death in the age group 15–24 years [5]. It is therefore clear that road traffic is a major public health concern. However, it is also clear that one cannot envisage a modern society with no traffic. This does not mean that the problem can be ignored – research needs to be done on alternatives to current road use and ways need to be found to minimize the impact on public health. There have been numerous efforts to reduce different kinds of impacts of road traffic, such as emissions (electric, hybrid, and hydrogen cars [6], natural gas buses [2], catalysts and particle traps [7], and driving style [8]); congestion (traffic control [9], street tolls, public transport subsidies); injuries (airbags, speed limits [10] and need to travel (urban planning [11]). Despite these efforts, the general view is that the environmental, health, and other detrimental effects of car traffic will continue to increase in the future. Although many modes of public transport are more efficient, they cannot compete with the flexibility of the private car. Thus public transport is not used as much as would be optimal for society. Novel systems that are both flexible and efficient should have special interest to municipal planners. Several attempts have been made to develop transport based on trip aggregation. There exists car sharing clubs that rent cars to their members on a pay-as-you-drive basis even for periods as short as one hour. The renting of a car can often be done conveniently on the Internet or using a cell phone with the car being picked up from and returned to dedicated areas that are located around the city. These clubs can be economically viable without subsidies. They attract people that drive less than average, because the fixed costs are clearly lower than with car ownership. Club membership usually reduces car driving by 30–50% and increases the use of public transportation [12]. Such a club exists also in Helsinki. In car pooling, people form groups that travel together instead of using their own individual cars. In many countries, car pooling is encouraged and there are dedicated lanes for cars with several passengers. Car pooling may be a practical solution for commuting in some cases, but it is not feasible for the majority of trips. There is no policy actively encouraging car pooling in Finland. Municipalities have subsidised demand-responsive public transport in some places, such as the outlying areas of the Helsinki metropolitan area, or within the metropolitan area to organise trips for handicapped people. The former system was developed as a response to the disappearance of regular bus routes. Although important for some specific subgroups, these systems have handled only a small subset of all trips, and the low volume prevents efficient aggregation. Indeed, there have been complaints about poor service related to low volume: long waiting times and lengthy routes to the destination. However, also positive results have been obtained from some pilot projects aggregating trips. In Finland, this has been the case with trips that are organised by the government or municipality due to societal reasons, such as some patient trips to and from hospitals, or lengthy school trips. Because society has paid the cost of the trip, it has also directly gained benefits from trip aggregation. The volume of these regional pilot studies has been in the order of 100 trips per day. [13] In 2000 in the Helsinki metropolitan area, there were 2.9 million trips made every working day a region with one million inhabitants (the communities of Helsinki, Espoo, Vantaa, and Kauniainen). These trips were divided as follows, 1.3 million were by private car, 0.8 million by public transport, and 0.8 million on foot or by cycling [14]. In this work, we modelled the impact and potential of a hypothetical large-scale demand-responsive public transport system in the Helsinki metropolitan area, where a centralised system would collect the information on all trip demand online, aggregate the trips with the same origin and destination into public vehicles with eight or four passenger seats, and transmit the trip instructions directly to the passengers' mobile phones. We designate this system as composite traffic, as it represent a composite of the flexibility of the taxi and the efficient trip aggregation of the bus. We studied 1) how effectively trips could be aggregated; 2) what are the various costs and pressures of car and composite traffic; 3) what are the perceived costs for different passengers; 4) what incentives are needed to reach particular composite traffic volumes or areal coverage; and 5) how variation, uncertainty, and multiple decision-makers would affect the decision situation. Methods The model was built using the Analytica 3.1™ program that utilises a graphical interface for creating probabilistic (Monte Carlo) models . It consists of two parts: first, a deterministic trip aggregation model that produces the output tables used in the decision analysis. The calculation of the results takes several days and therefore is calculated separately. In the second part, the trip aggregation results are combined with unit cost functions, emission factors and other uncertain and/or varying variables using a probabilistic simulation. One of our major aims was to create the model in such a way that a non-specialist could follow the logic, reasoning, and conclusions without going into the modelling details (Figures 1, 7, and 8). In addition, he/she should even be able to test the model by using some personal assumptions. To facilitate this, we used a system diagram method denoted pyrkilo (see [2,15] and supplemental material at ). It has been developed in KTL as component of the science-policy interface, i.e. promoting the flow of information and understanding between science and policy, within the field of environmental health. Figure 1 Overview of the model Composite traffic v. 1.0.1. The model calculates health effects and other costs in the Helsinki metropolitan area. The overview of the urban traffic problem utilizes the DPSEEA approach (driving force, pressure, state, exposure, effect, action) [22]. The most important colour and shape symbols are explained in the lower left corner. The principle is to describe an environmental health risk situation in a formal manner utilising system diagrams with causal connections of actions, outcomes, and interconnected variables (for causal diagrams, see [16]). As an example, a pyrkilo diagram may contain items along a causal pathway from abatement strategies for emissions to their dispersion to exposure to health effects (blue ovals, see Figure 1). However, the diagrams also describe parameters other than causal connections such as values, preferences and arguments, and finally conclusions from the examination (orange and yellow blocks). Arrows depict causal or non-causal connections between items. One novel feature of the pyrkilo diagrams is that they include and deal with three different kinds of variables in a single examination. These are physical variables related to a health hazard, political variables related to valuation of the outcome and other issues, and variables related to the assessment process itself, such as scope and conclusions. The pyrkilo diagrams used in this study are based on an assessment of the literature by the authors. Stakeholders were not involved in this first study. However, the problems of and solutions to urban transportation have practical implications in everyday life. Stakeholder studies would therefore be useful in the future to describe and understand important issues and value judgements related to composite traffic and other proposed solutions. The whole metropolitan area was divided into 129 areas. The 129 areas are standard areas used in urban planning and they contain on average 7300 inhabitants (standard deviation 5000 inhabitants). We used a road matrix containing 234 major links between the 129 areas; there was only one link connecting any two neighbouring areas, and there was exactly one, pre-specified route between any two areas. Trip data and trip aggregation We took modelled personal car trips (public transportation, cycling, and walking were excluded from this exercise) for one working day (year 2000) into our model. Trip rates were estimated for each origin-destination pair (129129 pairs) and for each time point (12 min intervals, resulting in 120 time points) based on summary data of trips in the Helsinki metropolitan area [14]. The summary data was disaggregated into smaller areas using numbers of population and jobs in each area [17]. The trips were disaggregated over 24 hours based on time activity data in the traffic (based on diaries) [18]. All scenarios had the same street structure and number of trips with a particular origin, destination, and time. The trips were divided into car trips and composite trips differently in each of the 91 scenarios based on 1) the percentage of the trips that are handled by composite traffic (composite fraction) ranging from 0 to 100% (13 alternatives; default: 50%), and 2) the area where composite traffic would be provided (i.e. the area within which a composite vehicle is guaranteed if desired) ranging from downtown Helsinki (81810 inhabitants) to the whole metropolitan area (944200 inhabitants) (7 alternatives; default: the whole area). The number of cars driving and kilometres driven by personal cars were calculated assuming 1.5 passengers per car. The composite traffic trips were allocated into public vehicles with either eight or four passenger seats. The trips were aggregated if they had the same origin and destination areas (129 areas) and timing (12 min intervals). If there were less than four passengers in an aggregate of trips, the trips were divided into two parts, and the passengers had to transfer into another vehicle at the most busy point along the route. This made it possible to effectively aggregate trips with the same origin but different destination and vice versa. If there were still not enough passengers to fill a vehicle, a non-full vehicle was used; everyone was guaranteed of receiving a ride. We assumed that the composite traffic would be a door-to-door service, or within walking distance comparable to that from a parking lot. The vehicles drove directly from the origin area to the destination area without stopping. The actual trips, modes of transportation, and delays during trips and vehicle transfers were calculated, as well as the kilometres travelled by each type of vehicle and the number of vehicles needed. Decision analysis In the decision-analytic part of the model, several of the outcomes modelled (i.e., pressures) were monetised and combined using probabilistic simulation. We assessed separately the uncertainty of an input value and the variation of the value between individuals in the population. Costs were separately calculated for the passenger and society. Some costs affect these stakeholders differently, such as fine particle and carbon dioxide emissions: these were calculated as societal costs only, not as costs to a passenger. Estimating the costs and benefits of and human behaviour in a hypothetical traffic system is a difficult task. We used current values (including estimates on their uncertainties) whenever available, and wide confidence intervals otherwise. We made a special effort in trying to quantitatively estimate the individual variation within the population. Variation between individuals was separately estimated for three variables: how passengers evaluate the capital costs of owning a car; how passengers are willing to pay for either the right to drive themselves or the right not to drive; and how many passengers are travelling together. The detailed descriptions of cost elements can be found in Table 2. Table 2 The input variables used in calculations of the pressures. In most cases, there is no data available on uncertainty, and it is based on author judgement (AJ). Title, Unit [Reference] Description Definition Accidents, cases/a [3,24] The number of injuries and deaths in traffic accidents in the Helsinki metropolitan area. Poisson distribution is used to describe the uncertainty. Injuries: Poisson(1129) Deaths: Poisson(26) Accident costs, €/d [24,25] The societal costs of traffic accidents were 227 million euro in Helsinki in 2004. The numbers are scaled up from Helsinki to the metropolitan area based on the numbers of people injured in accidents. The uncertainty is based on the standard deviation of the variable Accidents (deaths), which is ca. 20% of the mean. We assume that half of the accidents are attributable to private car traffic, while the other half is attributable to other traffic modes (walking, cycling, public transportation). In addition, the accident risk is proportional to the change in traffic volume, but there is uncertainty about the slope. The expected value is that when traffic volume decreases by 10%, accident risk decreases by 5%; but it could vary between 0% and 10% (the latter being the default assumption in the guidelines for road construction planning). var a:= 227 M((1129)/724)/365; a:= normal(a,a/5) var b:= vehicle_km; b:= (1-b/b[comp_fr=0])triangular(0,0.5,1); b:= (1-b)a0.5vehicle_km Vehicle price, €/vehicle [26] Price of a new vehicle. Note that the interpretation is slightly different with different vehicles. The car price is the price that a random new car would cost, and it has therefore large uncertainty. The price of a composite vehicle is the average price of a taxi-style car in Finland, and the confidence intervals are narrower because there is no individual uncertainty. This is because the price of an individual car affects the costs of individual car trips, while the cost of a composite trip is dependent on the total cost of the fleet to the service provider. The same typical vehicles are used as in the Emission factor. 8-seat vehicle: 39520Triangular(0.75,1,1.25) 4-seat vehicle: 22600Triangular(0.75,1,1.25) personal car: lognormal(19100,1.5) [median, geometric standard deviation for lognormal distribution] Vehicle lifetime, a Author judgement (AJ) Expected operation time of a new vehicle. 8-seat vehicle: 7Triangular(0.75,1,1.25) 4-seat vehicle: 5Triangular(0.75,1,1.25) personal car: 9Triangular(0.7,1,1.3) Cap variab, fraction (AJ) The value a car-owner gives to capital costs of the car as a fraction of the true costs. Each row represents one possibility for the distribution of individual valuations in the population. Probability distributions are used to represent this variation within the population. Three possible distributions of variation within the population: A: Uniform(0,1) B: Triangular(0,0,1) C: Bernoulli(0.2)) Cap uncert, – (AJ) The uncertainty between several valuation distributions about Cap variab on the population level. A: 1/3 B: 1/3 C: 1/3 Trips per car, trips/d/car (AJ) Average number of trips per car per day, i.e. the cumulative number of passengers that use the car during the day. This value is used to calculate the vehicle capital costs. uniform(4,10) Parking space, €/d/parking space [25] Cost of a parking space to society due to the loss of the land, and maintenance costs. The average price of development land in Helsinki is around 300 €/m2, and one parking space requires ca. 20 m2. The standard values in road planning are 30 years for scope and 5%/a for discount. Opportunity cost for land is calculated based on these values; in addition, it is assumed that 50% of composite traffic parking places can be located in areas where the parking cost is negligible. 9.1lognormal(1,1.3) Parking price, €/trip [27] The cost of 30 min parking in zones 1, 2, 3 in Helsinki. It is assumed that each car trip involves 30 min of parking during daytime, while during evening and night, the parking is free. Also daytime parking at home is included in these estimates, although it is difficult to price. In any case, it is common to pay at least 5–10 euro per month for a parking place (or more for a garage), which is 15–30 cents per day. Due to the uncertainties, the confidence intervals are large. Downtown: 2.40.5Triangular(0,1,2) Other centre: 1.20.5Triangular(0,1,2) Suburb: 0.60.5Triangular(0,1,2) Emission factor, g/km [26,28] Fine particle and carbon dioxide unit emissions for average vehicles. Fine particle emissions are taken from the Lipasto model using average (mixed gasoline and diesel) values for personal car and diesel EURO3 (applied since 2000) values for composite vehicles. For CO2, typical emissions of a new car were used based on the Finnish Vehicle Administration AKE. The following vehicles are used as typical examples of the class: 8-seat vehicle: Toyota Hiace 2.5 D4D 100 4 door long DX bus (diesel) 4-seat vehicle: Toyota Corolla 2.0 90 D4D Linea Terra 5 door Hatchback (diesel) Car: Toyota Corolla 1.6 VVT-i Linea Terra 5 door Hatchback (gasoline) var a:= triangular(0.3,1,1.7) var b:= triangular(0.9,1,1.1) PM emission: 8-seat vehicle: 0.1a 4-seat vehice: 0.1a personal car: 0.047triangular(0,1,2) CO2 emission: 8-seat vehicle: 232b 4-seat vehicle: 153b personal car: 168triangular(0.3,1,1.7) PM unit lethality, deaths/kg [2] Primary fine particle emissions of 24290 kg/a caused 12.5 deaths in a risk assessment study in Helsinki (Tainio et al 2005). We use the distribution of deaths per emission derived from that study. fractiles([-722.3, 5.640, 42.28, 59.87, 80.13, 115.0, 203.7, 293.9, 359.8, 413.2, 464.0, 513.9, 566.2, 623.3, 685.4, 757.7, 844.1, 951.9, 1093, 1314, 2805])/1 M Emission unit cost, €/kg [1,2,25] The value of a statistical life is 0.98 – 2 M€ (Watkiss et al. 2005). CO2 emission trade started in the EU this year, and the market price is used. According to newspapers Helsingin Sanomat (May 7, 2005) and Taloussanomat (July 11, 2005), the price has varied between 10 and 30 €/ton. The standard road planning value for CO2 emission is 32 €/ton. PM emission cost: PM_unit_lethalityuniform(0.98 M,2 M) CO2 emission cost: uniform(5,40)/1000 Driver salary, €/h [29] Statistics Finland 2005 Monthly salary and social security costs (35%), and scaled to one hour assuming 160 hours of work per month. The salary is based on that of bus drivers in municipality-owned bus companies. var a:= 2313/1601.35; normal(a,a0.18) Fuel consumption, l/km [26] Fuel consumption of a vehicle. It is assumed that composite vehicles use diesel fuel and cars use gasoline. The values are based on standardised European consumption values of a new car. The same typical vehicles are used as in the Emission factor. 8-seat vehicle: (8.7/100)Triangular(0.75,1,1.25) 4-seat vehicle: (5.7/100)Triangular(0.75,1,1.25) personal car: (8/100)Triangular(0.5,1,1.5) Fuel price, €/l (AJ) Diesel fuel price for composite vehicles; gasoline price for cars. The values are based on a general follow-up of retail prices in Finland in fall 2004 – summer 2005. diesel: 0.95triangular(0.8,1,1.2) gasoline: 1.22triangular(0.8,1,1.2) Car maintenance, €/km [30] Maintenance costs (service, tyres, oil etc.). This is based on Autoliitto's report 'Costs of car 2004'. Insurance and use tax are excluded. Similar to capital costs, there may be other reasons to own the car, and then these would be sunken costs. Original values (assuming an old car with the original price 20000 e, 20000 km/a of driving) (€/a): Maintenance 844 Tyres 320. Thus, 1164/20000 = 0.0582 €/km Triangular(0.03, 0.058, 0.086) Ticket, €/trip (AJ) The income that the service provider wants to receive from composite traffic users in addition to the price of the direct costs (vehicle, fuel, driver, and parking costs). uniform(0.2,0.4) Rush delay h/trip, fraction (AJ) Delay that is caused by increased link intensity. The node contains two values. Delay is the average time of delay due to traffic jams during daytime. Reduction is the relative reduction to 'Link intensity' (average vehicle flow on the 30 most busy roads at 8.00–9.00 AM) that is needed to reduce the delay to 0 min. Delay: Uniform(0,10)/60 Reduction: 0.3 Time unit cost, €/h [25] The cost of time spent waiting for a composite vehicle or in traffic jam. This is based on the standard road planning values. Triangular(0,5.9,11.8) Drive variab, fraction (AJ) Willingness to drive. This is expressed as fraction of composite driver's salary. Each row represents one possibility for the distribution of individual valuations in the population. Probability distributions are used to represent this variation within the population. Three possible distributions of variation within the population: A: Uniform(-0.3,0) B: Triangular(-0.1,0,0.3) C: Uniform(-0.2,0.2) Drive uncert, – (AJ) The uncertainty between several valuation distributions about Drive_variab on the population level. A: 1/3 B: 1/3 C: 1/3 Car occupancy, fraction [31] Proportion of cars with different numbers of passengers. The original data is from streets entering downtown Helsinki during one weekday (from 6.00 to 21.00) in May. Passengers (incl driver): 1: 0.72 2: 0.233 3: 0.033 4: 0.01 5: 0.004 Results The trips and vehicle types used are shown in Figure 2. The need for less efficient vehicles (4-passenger-seat vehicles not fully occupied) remained almost constant in spite of the large variation in the trip volume occurring during the day. When the volume increased, more efficient vehicles were used, and therefore the number of vehicles increased in a sublinear fashion. In addition, the fraction of trips without transfer increased. The improved performance during rush hours is a major benefit compared with personal cars, for which the opposite is true. Figure 2 Composite traffic trips by vehicle type as a function of time. The fraction of composite trips (composite fraction) is 50% of the current 1.3 million personal car trips per day. Note that a trip with a transfer is calculated as two half-trips and may appear in two different vehicle types. The several outcomes studied clearly indicated that composite traffic would reduce most pressures attributable to private car (Table 1). For example, large reductions were seen in traffic volumes during rush hours (this is of benefit to car drivers as well) and CO2 emissions. (Note that the absolute numbers of vehicle flow are likely biased upwards because most smaller roads are excluded from the model.) Table 1 The pressures and costs from traffic (composite+car) in the Helsinki metropolitan area. Pressure Private cars only 25% composite traffic 50% composite traffic 75% composite traffic 100% composite traffic Fraction of composite trips without transfer (%) - 8.7 19.5 28.1 35.0 Vehicles needed (number) 68000 60700 49300 37000 19700 Parking spaces needed (number) 91900 81900 66800 49600 25200 Average vehicle flow on the 30 most busy roads (vehicles/h at 8.00–9.00 AM) 5150 4350 3440 2380 1220 Injuries due to accidents (cases per year) 565 (537–593) 532 (498–566) 483 (425–543) 429 (336–524) 367 (233–504) Deaths due to accidents (cases per year) 13.0 (9.0–17.5) 12.2 (8.4–16.3) 11.1 (7.52–15.2) 9.89 (6.35–14.0) 8.46 (4.66–12.9) Deaths due to fine particles (cases per year) 95.4 (0.3–292) 96.7 (0.6–284) 87.6 (0.6–253) 75.8 (0.5–215) 61.0 (0.4–179) Fine particle (<2.5 μm of diameter) emissions (kg per day) 500 (158–842) 507 (239–774) 459 (258–659) 397 (242–551) 320 (167–473) Carbon dioxide emissions (ton per day) 1790 (1660–1910) 1580 (1480–1670) 1280 (1210–1350) 953 (907–999) 574 (535–613) Driver salaries (thousand € per day) 0 599 (422–776) 947 (667–1230) 1260 (888–1630) 1560 (1100–2020) Vehicle costs (capital+operational) (thousand € per day) 2750 (1930–3930) 2340 (1710–3230) 1820 (1380–2430) 1270 (1030–1590) 667 (582–753) Time cost due to delay (thousand € per day) 365 (20.7–994) 308 (77.8–664) 233 (73.8–393) 328 (104–553) 424 (134–713) Average car trip cost to passenger (€ per trip) 2.88 (1.74–4.19) 2.76 (1.65–4.02) 2.66 (1.57–3.94) 2.76 (1.62–4.09) - Average composite trip cost to passenger (€ per trip) - 3.70 (3.00–4.42) 2.91 (2.38–3.43) 2.68 (2.19–3.15) 2.54 (2.08–2.99) Mean (90% confidence interval when applicable). The number of vehicles and parking places is theoretical and involves the modelled trips only; a car owner may need the car for trips outside Helsinki even if he/she uses composite traffic. The true number of cars in the area was 346 400 in 2001 [14]. Decreased congestion reduces the costs of private cars. This feedback phenomenon is partly taken into account in time costs. The current ticket prices for buses, subway, and trams are 1.70 € per trip in Helsinki and 2.90 € per trip between communities in the Helsinki metropolitan area. Note that the car trip and composite trip costs include time costs due to delay (rush, transfers). If a passenger requires a trip without a transfer, the additional price to him/her will be 3 – 6 € per trip during daytime. This cost is due to reduced efficiency in trip aggregation. Fine particle emissions did not decrease as extensively due to the putative shift from the current gasoline-dominated car fleet to diesel composite vehicles. Improved technology may change this result in the next few years [6]. For example mandatory particle traps in all new diesel vehicles would dramatically reduce the fine particle emissions from composite traffic. However, at the level of 50% composite traffic, there would be about ten fewer deaths due to accident and PM emission reductions, which is not negligible. The salary of the driver was a new, expensive cost item. It is the main reason why the composite traffic is not obviously superior to the private car (Figure 3). Figure 3 Costs of daytime trips separated by source (mean and 90% confidence intervals). Emission costs are not calculated for the passenger; ticket costs are not calculated for society. Driver costs with car may be negative, because some people prefer driving themselves. There are no time costs for cars in the figure, because it was estimated that there are no traffic jams in the default scenario (50% composite traffic). Of all trips in Figure 2, 20% were direct; but 80% did involve one transfer at one of the specific transfer points but with only a few minutes of extra waiting. The willingness to use public transport generally decreases if there is a need to change once or especially more than once. With composite traffic, this discomfort is likely to be less of a problem because even with one transfer, the shortest route is used and the impact on trip duration is minor. One relevant stakeholder that should be considered is the random passenger (i.e., not average) who could take a car but would consider composite traffic as an alternative. The overview of this modelled examination is shown in Figure 4. Almost half of the modelled daytime passengers were estimated to find composite traffic more attractive than the private car even without any subsidies. The percentage was 45% for trips longer than 5 km (these trips consist of 75% of all trips), and 20% for shorter trips. The single most important variable affecting the attractiveness was the number of persons travelling together, as this obviously reduced the cost per person with the private car. Figure 4 Individual variation and uncertainty in the cost of a composite trip. The cost of a composite trip is compared with a private car trip for an individual passenger. The estimates are for daytime trips with 50% composite fraction scenario. The trips are divided into two groups based on length (blue cross: <5 km; red plus: >= 5 km). The variation between individuals is shown on the X axis, with people most in favour of composite traffic on the left. The expected values across individuals are shown as lines, and the dots represent the uncertainty of the value. Another stakeholder is society: whether it should subsidise composite traffic to achieve a certain level of modal shift from private cars to composite traffic. Society is burdened also with different costs compared to those incurred by the individual passenger: health effects of fine particles, greenhouse gas emissions, and opportunity costs for land currently used as parking spots. Figure 5A and 5B show that composite traffic is clearly beneficial when it replaces 30% or more of the private car traffic. Although the societal cost (excluding subsidies) decrease until all car traffic is replaced (Figure 5A), the subsidies needed to attract more passengers to composite traffic increase progressively (Figure 5B). It seemed that at 70–80%, the subsidies start to have negative impacts on the societal benefits. Figure 5 Marginal societal costs of traffic (composite+car) as a function of composite fraction. A: Societal costs (excluding subsidies for composite traffic) during different periods of day. B: Subsidies to composite ticket prices needed to reach the composite fraction target (i.e., to make that fraction of current private car passengers to favour composite traffic according to the model) during different periods of the day. For comparison, the current subsidies to public transportation in the Helsinki metropolitan area are on the range of 380 000 € per day [23]. C: Societal costs (including subsidies) during daytime with extending areal coverage of composite traffic (starting from the most densely populated areas). The legend shows the number of inhabitants living in the covered area. The pink curve (538100 inhabitants) is the city of Helsinki). Another important practical question is whether it is possible to start up a system like this without prohibitive capital expenditure. As we have pointed out, composite traffic becomes more profitable when there is a high volume of trips. Thus, it seems that a good strategy would be to start the operation during heavy commuting hours in densely populated areas, and then to rapidly expand the service beyond the low-volume "inefficiency bump" seen in Figure 5A. A volume that would be already profitable to the service provider and still produce societal benefit is 8000 trips per day (25% composite fraction during daytime in downtown Helsinki, an area with 82000 inhabitants) (Figure 5C). This would require a fleet of about 250 vehicles, which is 10% of the number of taxis currently in the Helsinki metropolitan area [19]. If this critical volume is reached, the system becomes more and more profitable as it expands, at least until it covers around 50% of all daytime car trips in most of the area (420000 trips per day). We used several origin-destination trip matrices to test the sensitivity of the aggregation. One extreme case tested was a matrix where the trip rate from all areas to all other areas was the same number at each time point. The total volume at each time point varied in the same way as in the default matrix. This is a very flat trip matrix with no spatial trip aggregates, which is an unfavourable situation for the composite traffic. However, the major conclusions remained the same with all matrices tested. The largest differences between the matrices were that the fraction of trips without transfer decreased with flat matrices, and the number of vehicles needed varied, usually being higher than with the default matrix. The model did not directly assess the additional driving needed to pick up or drop off passengers. Therefore we performed a sensitivity analysis where the composite trips were on average 1.8 km longer than the private car trips (this was a rough upper-bound estimate based on the sizes of origin and destination areas). On average, this increased the cost by approximately 20 cents per trip. Although this is not negligible, it does not alter the overall conclusions (compare for example to Figure 4). Further studies are warranted on this issue. The separate treatment of uncertainty and variation made it possible to evaluate their importance in the model. The value of information (i.e. the price that is worth paying to reduce a particular uncertainty) was calculated for these decisions [20]. For the passenger, there is little value of information, as there is mainly individual variation and only marginal uncertainty (Figure 6A). For the societal question of whether to subsidise composite traffic at 50% composite fraction or not at all, the total value of resolving all uncertainty is only about 30 000 € per day, and the value for every single variable was zero. This means that the conclusion is robust and even if the certainty about any single variable was found out to be the most unfavourable to the composite traffic, the optimal decision would still remain unchanged. However, when the question was about the optimal level of subsidies, more uncertainty was involved with the decision (Figure 6B). The single most important variable was about the customers' willingness to drive him/herself rather than being a passenger in a public vehicle. This suggests that comparative studies on passenger attitudes are warranted. Figure 6 Value of information of the daytime trips for society and the passenger. Total VOI is the expected value of perfect information for all uncertainty; other rows are expected values of partial perfect information for each uncertain variable. Variables with zero value (i.e. variables that could not change the decision) are omitted. A part of the value of car price is actually due to variation that is not explicitly separated from the uncertainty. Figure 7 The Action module. Figure 8 The Composite traffic module. Discussion Compared with buses and the subway, the composite traffic is quick because it uses the shortest route and does not stop between the origin and the destination. Nonetheless, it is likely to be slightly slower than the private car, because a few minutes are lost during transfer; in addition, the vehicle picks and drops 4–8 passengers from and to the ends in a limited area. On the other hand, it is often possible to transport the passenger directly to the destination without the need of searching for a parking place nearby. We were careful not to unrealistically exaggerate the benefits of the composite traffic. On the contrary, we excluded several clear but not easily quantifiable benefits: Replacement of low-volume bus routes with composite traffic would improve service and reduce costs at the same time. Composite traffic would be an efficient feeder for high-volume public transport modes. Reduced road traffic volumes would save money currently used on road construction, maintenance, and infrastructure. Finally, we assumed that the trips are uncorrelated in time (given the total volume at each time point). However, trips are often directed to or from particular locations such as schools, offices, stadiums, and supermarkets at specific times, which improves trip aggregation. There are several kinds of approaches in use to solve some of the problems attributable to car traffic. Some involve improving the car to reduce its pressures per vehicle-km (airbags or catalysts [7]); some aim at improving traffic flow [9]. There have also been attempts to replace the car by developing personal rapid transit – a transportation mode that utilises new techniques to offer flexible transportation services [21]. Although this is novel approach, these systems usually require extensive (and expensive) new vehicle or road infrastructure, and this lessens the possibilities of their implementation and success. Composite traffic requires no new techniques, except a system for collecting, organising, and distributing the trip information. The interface to this database and optimising system could be based on text messages sent to mobile phones, in conjunction with an Internet site where passengers can create personalised profiles. The composite vehicles would probably need to have a real-time connection to a global positioning system and a centralized trip optimising system. This would enable a minute-to-minute planning of vehicle routes and destinations so that the number of vehicles driving empty or waiting could be minimised. We did not estimate the capital or operational costs of implementing the trip optimising system, but it is reasonable to assume that it could be centrally organised by the community without dramatically changing the level of societal benefits. For comparison, the current electronic travel card system in Helsinki area handles a million trips per day with operational costs ca. 0.02 € per trip [3]. We did not aim to describe in detail, how the system would look like in practice. The eight and four-passenger-seat vehicles were used as examples of possible vehicles because they are currently in mass production. The fuel and propulsion systems could well be better than the current standards, because the relatively small fleet, small geographical operational area, and probably centralised ownership would allow for specialised solutions. The aim of this work was simply to study whether it would be possible to develop a flexible transport system that would utilise trip aggregation and that would be more effective than the current private car. Conclusion In conclusion, we have shown that 1) with composite traffic, it is possible to aggregate trips in an urban area so that the level of service and flexibility are comparable to those achieved by private car; 2) many important pressures caused by car traffic could be reduced by 50–70% (Table 1), but at the cost of considerable driver salary costs; 3) almost 50% of day-time passengers would consider it more attractive than using their own car even without subsidies; 4) it is possible to start the service with a small volume and then expand it later, thus reducing the financial risks; and 5) the passenger and society have an important interplay as decision-makers, but the composite traffic shows robust benefits to both groups despite current variation and uncertainty. However, the most important impact (and the most difficult to model) of composite traffic may come from the increased degree of freedom in urban policy-making and planning, and public health policy, when the pressures associated with increased private car use and traffic are relieved without impairing the ability of people to travel rapidly about the city. Competing interests The author(s) declare that they have no competing interests. Authors' contributions JTT conceived the study, performed modelling and drafted the manuscript. MT participated in the design of the study and helped to draft the manuscript. Both 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 Composite traffic model for calculating health and other pressures. CompositeTraffic_1_0_1.ana (Analytica™ 3.1 file in XML format). Model identifier (unified resource name) is URN:NBN:fi-fe20051439. For a free browser, see . The model contains the full model including the code and the descriptions. It has the road and trip data algorithms as well as static (precalculated) results; the decision analysis can be performed in real-time using the default or user-defined assumptions. For more guidance and updates of the model, see . Click here for file Acknowledgements We wish to thank Prof. Matti Jantunen, Prof. Jouko Tuomisto, M.Sc. Olli Leino, M.Sc. Marjo Niittynen, M.Sc. Sanna Lensu, and Ms. Arja Tamminen for their helpful comments, evaluations, and work to collect data, and Ewen MacDonald for improving the grammar. This study was funded by the Academy of Finland, grants 53307 and 8042, and the National Technology Agency of Finland (Tekes), grants 40715/01 and 616/31/02. ==== Refs Watkiss P Pye S Holland M CAFE CBA: Baseline analysis 2000 to 2020 2005 AEAT/ED51014/Baseline Issue 5 Oxon, AEA Technology Environment 1 112 Tainio M Tuomisto JT Aarnio P Hänninen O Koistinen K Jantunen M Pekkanen J Health effect caused by primary fine particulate matter (PM2.5) emitted from buses in Helsinki Metropolitan Area, Finland Risk Anal 2005 25 149 158 10.1111/j.0272-4332.2005.00574.x YTV YTV:n toiminta- ja taloussuunnitelma 2003-2005 (Plan and budget of the Helsinki Metropolitan Area Council, 2003-2005) Pääkaupunkiseudun julkaisusarja (Publications of the Helsinki Metropolitan Area Council) 2002 C 2002:20 Helsinki, YTV (Helsinki Metropolitan Area Council) 1 61 Nieminen K Liikenneonnettomuudet Helsingissä vuonna 2003 (Traffic Accidents in Helsinki in 2003) Helsingin kaupunkisuunnitteluviraston julkaisuja (Publications of the Urban Planning Office of Helsinki) 2004 2004: 17 Helsinki, Helsingin kaupunkisuunnitteluvirasto (Helsinki City Planning Department) 1 93 Finland S 2005 Helsinki (Internet), Statistics Finland Ortmeyer TH Pillay P Trends in transportation sector technology energy use and greenhouse gas emissions Proceedings of the Ieee 2001 89 1837 1847 10.1109/5.975921 Mediavilla-Sahagun A ApSimon HM Urban scale integrated assessment of options to reduce PM10 in London towards attainment of air quality objectives Atmos Environ 2003 37 4651 4665 10.1016/j.atmosenv.2003.07.004 Vangi D Virga A Evaluation of energy-saving driving styles for bus drivers Proceedings of the Institution of Mechanical Engineers Part D-Journal of Automobile Engineering 2003 217 299 305 10.1243/09544070360613255 Hounsell NB McDonald M Urban network traffic control Proceedings of the Institution of Mechanical Engineers Part I-Journal of Systems and Control Engineering 2001 215 325 334 10.1243/0959651011541157 Elvik R Optimal speed limits - Limits of optimality models Transportation Research Record HIGHWAY SAFETY: WORK ZONES Transportation research record 2002 WASHINGTON, TRANSPORTATION RESEARCH BOARD NATL RESEARCH COUNCIL 32 38 Banister D Reducing the need to travel Environment and Planning B-Planning & Design 1997 24 437 449 Litman TA A.Hensher D and J.Button K Transportation demand management and "win-win" transportation solutions Handbook of transport and the environment Handbook in transport 2003 First edition The Netherlands, Elsevier Ltd. 805 814 Korpela J Nikkinen M Siltala S Yhteiskunnan korvaamien matkojen järjesteämisen hallinnointi ja rahoitus. (Management and financing of trips paid by society. Report of the travel dispatch working group) 2004 24/2004 Helsinki, Ministry of Transport and Communications 1 50 YTV Liikkumisen nykytila (The Present-day Traffic Situation) Pääkaupunkiseudun julkaisusarja (The Helsinki Metropolitan Area Publication Series) 2001 B 2001:10 Helsinki, Pääkaupunkiseudun yhteistyövaltuuskunta (Helsinki Metropolitan Area Council, YTV) 1 27 Tuomisto JT Tuomisto J Tainio M Niittynen M Verkasalo P Vartiainen T Kiviranta H Pekkanen J Risk-benefit analysis of eating farmed salmon Science 2004 305 476 476 15273377 10.1126/science.305.5683.476 Pearl J Causality Models, reasoning, and inference 2000 1 Cambridge, Cambridge University Press 1 384 YTV Seutu-CD '02 (Register-based CD-ROM of maps and data about the Helsinki metropolitan area) 2004 Helsinki, YTV (Helsinki Metropolitan Area Council) Kousa A Kukkonen J Karppinen A Aarnio P Koskentalo T A model for evaluating the population exposure to ambient air pollution in an urban area Atmos Environ 2002 36 2109 2119 10.1016/S1352-2310(02)00228-5 Jalasto P Ikonen T Yön ainoa valopilkku. Taksiliikennelakityöryhmän mietintö (Report of the working group on taxi transport legislation) Publications of the Ministry of Transport and Communications 2004 43/2004 Helsinki, the Ministry of Transport and Communications 1 132 Morgan MG Henrion M Uncertainty A guide to dealing with uncertainty in quantitative risk and policy analysis 1990 Cambridge, UK, Cambridge University Press 1 332 Lowson M New approach to effective and sustainable urban transport Transit: Rail Transit, Commuter Rail, Light Rail Transit, Major Activity Center Circulation Systems, New Technology and Maintenance 2003 42 49 WHO Environmental Health Indicators: Development of a methodology for the WHO European Region 2000 E71437 Copenhagen, World Health Organisation 1 111 Rosenberg M Räsänen J Joukkoliikenne keskisuurissa kaupungeissa. Vertailu ja yhteiskuntataloudellisia vaikutuksia (Public transport in medium-sized cities in Finland. Comparison and socio-economic impacts) Liikenne- ja viestintäministeriön julkaisuja (Publications of the Ministry of Transport and Communications) 2005 2/2005 Helsinki, Liikenne- ja viestintäministeriö (Ministry of Transport and Communications Finland) 1 61 Nieminen K Nieminen K Liikenneonnettomuudet Helsingissä vuonna 2004 (Traffic Accidents in Helsinki in 2004) Helsinki suunnittelee (Helsinki plan) 2005 4 Helsinki, Helsingin kaupunki, kaupunkisuunnitteluvirasto (City of Helsinki, Planning Department) 1 8 LVM Liikenneväylähankkeiden arvioinnin yleisohje (Guidelines for evaluating road planning) Liikenne- ja viestintäministeriön julkaisuja 2003 34 Helsinki, Liikenne- ja viestintäministeriö (Ministry of Transport and Communications Finland) 1 68 AKE Uuden auton kulutustiedot (The fuel consumption and emissions of a new car) 2003 Helsinki, Ajoneuvohallintokeskus (The Finnish Vehicle Administration AKE) 1 48 Rakennusvirasto Pysäköinti Helsingissä (Parking in Helsinki) 2005 Helsinki, Rakennusvirasto, Helsingin kaupunki (City of Helsinki, Public Works Department) Mäkelä K Laurikko J Kanner H Suomen tieliikenteen pakokaasupäästöt (The tailpipe emissions from road traffic in Finland Tutkimusraportti 2004 RTE 2814/04 Espoo, VTT Rakennus- ja yhdyskuntatekniikka (Technical Research Centre Finland, Building and Transport) 1 50 Kunnallinen työmarkkinalaitos (the Commission for Local Authority Employers) Kunta-alan palkkatasoselvitys (Salaries in public sector) 2001 Helsinki, Kunnallinen työmarkkinalaitos (The Commission for Local Authority Employers) Autoliitto Auton kustannukset 2004 (Costs of car 2004) 2005 Helsinki, Autoliitto (The Automobile and Touring Club of Finland) Hellman T Henkilöautojen kuormitus Helsingissä vuonna 2004 (The passenger number of personal cars in Helsinki in 2004) 2004 Helsinki, Helsingin kaupunkisuunnitteluvirasto (Helsinki City Planning Department) 1 13 14685084	16309549	PMC1325249	CC BY	2021-01-04 16:28:59	no	BMC Public Health. 2005 Nov 25; 5:123	utf-8	BMC Public Health	2,005	10.1186/1471-2458-5-123	oa_comm
==== Front RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-761635430710.1186/1742-4690-2-76ResearchUse of different but overlapping determinants in a retrovirus receptor accounts for non-reciprocal interference between xenotropic and polytropic murine leukemia viruses Van Hoeven Neal S [email protected] A Dusty [email protected] Division of Human Biology, Fred Hutchinson Cancer Research Center, Seattle, Washington 98109, USA2 Molecular and Cellular Biology Program, Fred Hutchinson Cancer Research Center, Seattle, Washington 98109, USA3 Current address: Centers for Disease Control, Atlanta, Georgia 30333, USA2005 15 12 2005 2 76 76 13 9 2005 15 12 2005 Copyright © 2005 Van Hoeven and Miller; licensee BioMed Central Ltd.2005Van Hoeven and Miller; 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 Retrovirus infection depends on binding of the retroviral envelope (Env) protein to specific cell-surface protein receptors. Interference, or superinfection resistance, is a frequent consequence of retroviral infection, and occurs when newly-synthesized Env binds to receptor proteins resulting in a block to entry by retroviruses that use the same receptors. Three groups of viruses demonstrate a non-reciprocal pattern of interference (NRI), which requires the existence of both a common receptor utilized by all viruses within the group, and a specific receptor that is used by a subset of viruses. In the case of amphotropic and 10A1 murine leukemia viruses (MLV), the common and specific receptors are the products of two related genes. In the case of avian sarcoma and leukosis virus types B, D, and E, the two receptors are distinct protein products of a single gene. NRI also occurs between xenotropic and polytropic MLV. The common receptor, Xpr1, has been identified, but a specific receptor has yet to be described. Results Using chimeric receptor proteins and interference studies, we have identified a region of Xpr1 that is uniquely utilized by xenotropic MLV and show that this receptor domain is required for non-reciprocal interference. Conclusion We propose a novel pattern of receptor usage by xenotropic and polytropic MLV to explain the NRI observed between these viruses. We propose that the specific and common receptor determinants for xenotropic and polytropic viruses are simultaneously present in discreet domains of a single Xpr1 protein. ==== Body Background Retroviral infection of a host cell is initiated by interaction of the retroviral Env protein surface (SU) subunit with a specific host cell receptor. This interaction triggers conformational changes within the Env protein that bring the virus and host cell membranes in close proximity, resulting in fusion and delivery of the viral capsids into the host cell cytoplasm (reviewed in [1,2]). In addition to promoting virus entry, the intracellular interaction of a viral Env and its cognate receptor can limit subsequent infection by subsequent viruses that bind the same receptor. This phenotype is referred to as interference or superinfection resistance because it prevents reinfection of a cell by the same virus strain, and has been used to classify viruses that utilize common cell surface receptors. Currently, mammalian retroviruses are divided into at least 10 different interference groups [3,4]. Within these groups, several retroviruses show a non-reciprocal interference pattern (NRI), where infection by one virus will block infection by a second virus, but infection by the second virus only slightly inhibits infection by the first virus. As the receptors for different retroviruses have been identified, it has become clear that NRI occurs in cases where related viruses within an interference group utilize a partially overlapping set of receptors for entry. In the case of amphotropic and 10A1 MLV [5] these receptors are Pit1 (Slc20a1) and Pit2 (Slc20a2), the products of two different genes with similar sequence and function. The phosphate transporter Pit2 serves as the receptor for both amphotropic MLV [6,7] and 10A1 [8]. However, 10A1 also binds to the closely related phosphate transporter Pit1, the receptor for gibbon ape leukemia virus (GALV) [9] and feline leukemia virus subtype B (FeLV-B) [10]. Because the amphotropic Env cannot bind to Pit1, it cannot block 10A1 infection of cells that express both receptors, while the 10A1 Env can block amphotropic MLV infection [8]. NRI also occurs among avian sarcoma and leukosis viruses (ASLV) types B, D, and E. Viruses of types B and D can interfere with each other as well as type E viruses, whereas ASLV-E can interfere with itself, but not with types B or D. This group of viruses have all been shown to utilize a common receptor, CAR1 [11,12]. Immunoprecipitation studies with different viral Env proteins have shown that this protein, encoded by the tv-b locus in chickens, produces two distinct protein products that differ in their disulfide bond pattern. One form, designated the type 1 receptor, can interact with ASLV-B and ASLV-E, whereas an additional form, the type 2 receptor, is specific for ASLV-B [13]. Another set of retroviruses that show NRI are xenotropic and polytropic MLV (X-MLV and P-MLV, respectively). Studies in cells derived from mink and the wild mouse Mus dunni demonstrated NRI between X-MLV and P-MLV [4,14], implying the existence of a common receptor. In both cases, initial infection of cells with X-MLV strains resulted in complete resistance to subsequent infection by P-MLV isolates. However, initial infection of cells with P-MLV strains did not block infection by X-MLV, although the X-MLV titers observed were decreased [4,14]. The hypothesis that these viruses share a common receptor was confirmed by the identification of a single cDNA from humans [15,16] and mice [17] that could mediate infection of both viruses when expressed in resistant cells. However, the identification of a single cell surface receptor is inconsistent with the interference patterns observed between these two viruses. Previously established mechanisms of NRI would suggest the existence of a specific X-MLV receptor that cannot be utilized by P-MLV. Screening of cDNA libraries by three groups independently failed to identify additional genes encoding a xenotropic specific receptor. Furthermore, genetic studies in mice have mapped susceptibility loci for xenotropic and polytropic viruses to the same region of mouse chromosome 1, and it is currently believed that these studies have identified alleles of the same gene [18,19]. These studies collectively argue against the existence of a separate locus encoding an X-MLV specific receptor, and suggest that the specific and the common receptor are encoded by the same gene. The common receptor, designated Xpr1, is a multiple-pass transmembrane protein of unknown function, although the gene displays a high homology to the Saccharomyces cerevisiae Syg1 gene. In yeast, Syg1 is involved in regulation of G-protein mediated signaling [20]. Current topology models predict that the receptor contains four extracellular loops (ECL), and intracellular amino and carboxy termini (Figure 1). Studies subsequent to the identification of the receptor have found residues within the putative third and fourth ECL, at amino acid positions 500 and 582 of the NIH Swiss mouse Xpr1 protein (mXpr1), that are critical for X-MLV receptor function [21]. Due to the ability of P-MLV isolates to utilize mXpr1, a similar set of residues required for P-MLV function were not identified. Our initial studies have focused on examining the determinants for both X-MLV and P-MLV in the same receptor. Making use of chimeras between the functional human and the nonfunctional hamster Xpr1 orthologs, we have identified regions of human Xpr1 that are sufficient to generate functional receptors for xenotropic and polytropic viruses. These studies suggest that two entry determinants are present on Xpr1. One determinant in the putative fourth ECL can be utilized by X-MLV and P-MLV, while a second determinant present in the third ECL can only be used by X-MLV. These results and additional interference studies support a novel model to explain NRI between these two virus types and have identified the xenotropic-specific receptor determinant as a particular domain of Xpr1. Figure 1 Analysis of human/hamster Xpr1 chimeras for receptor function. The predicted transmembrane domain structure of Xpr1 is shown at top and a corresponding block diagram is shown just below with the extracellular loops (ECL) shown in grey. A series of chimeras were constructed by exchange of the indicated fragments of hXpr1 and haXpr1. Restriction enzyme sites used in construction of the Xpr1 chimeras are shown above the block diagram. Chimeric receptors were subcloned into a retroviral expression vector and were transfected into CHO cells. The cells were then grown in medium containing G418 to select for expression of the Neo gene also carried by the expression plasmid. Cells were then exposed to LAPSN vectors bearing either the AKR6 or the 1E Env and the apparent titers of the vectors were determined. Results are means of at least two independent experiments with triplicate determinations in each experiment. Results Role of the putative third and fourth ECL of Xpr1 in xenotropic and polytropic virus entry To identify regions of human Xpr1 (hXpr1) that are required for xenotropic and polytropic virus receptor function, chimeric receptors combining coding sequences from hXpr1 and from the non-functional hamster receptor (haXpr1) were made and tested for receptor function following expression in Chinese hamster ovary (CHO) cells (Figure 1). Chimeric receptors were named based on the order of human (U) and hamster (A) sequences that include the putative extracellular domains of the receptor. Because CHO cells can be infected by some X-MLV strains, we used the Env from an X-MLV strain (AKR6) that was unable to mediate transduction of CHO cells even when haXpr1 was overexpressed in the cells (Figure 1, construct AAAA). We also tested the Env from a P-MLV strain (1E) of Friend mink cell focus-forming virus (FrMCF) that mediates only a low rate of transduction of CHO cells overexpressing haXpr1 (Figure 1, construct AAAA). Both Env proteins could mediate relatively efficient transduction of CHO cells expressing hXpr1 (Figure 1, construct UUUU). CHO cells expressing the Xpr1 chimeras were exposed to xenotropic [LAPSN(AKR6)] or polytropic [LAPSN(1E)] vectors and vector titers were determined (Figure 1). Cells expressing the UUAA chimera were poorly transduced by LAPSN(AKR6) or LAPSN(1E). Conversely, cells expressing the AAUU chimera were transduced at levels only slightly lower than those observed for hXpr1, indicating that the third and fourth loops of hXpr1 are important for both xenotropic and polytropic virus receptor function. Additional analysis of the determinants in this region shows that either the third or the fourth ECL is sufficient for xenotropic virus entry, but that only the fourth ECL can mediate polytropic virus entry. In particular, the AKR6 xenotropic vector could efficiently transduce cells expressing the AAAU or the AAUA chimeras, while the 1E polytropic vector could infect cells expressing the AAAU chimera but not the AAUA chimera. Xenotropic and polytropic Env show reciprocal interference on some chimeric receptors In previous interference studies, infection with a xenotropic virus blocks subsequent infection by viruses bearing either xenotropic or polytropic Env. In contrast, expression of a polytropic Env blocks subsequent infection by other polytropic viruses, but only slightly inhibits xenotropic infection [4,14]. Using our chimeric Xpr1 proteins, we examined the requirement for different regions of Xpr1 in interference between AKR6 and 1E pseudotype vectors. To establish CHO cell lines expressing both a chimeric Xpr1 receptor and a retroviral Env, CHO cells were transduced with retroviral vectors expressing the chimeric receptors and were then maintained in medium containing replication-competent AKR6 or 1E virus for a period of 6 weeks, as described in Materials and Methods. Cells expressing Xpr1 chimeras and viral Env proteins were challenged with LAPSN(AKR6) or LAPSN(1E) vectors. The level of interference was determined by comparing the titers of LAPSN(AKR6) and LASPN(1E) vectors on mock infected cells versus that on cells infected with a replication competent virus. In CHO cells expressing the AAUU chimera we observed a non-reciprocal pattern of interference between AKR6 and 1E viruses (Figure 2, left panels) similar to that reported previously. Specifically, CHO/AAUU cells infected with AKR6 virus were refractory to transduction by both LAPSN(AKR6) and LAPSN(1E), while CHO/AAUU cells infected with 1E virus were fully susceptible to transduction by LAPSN(AKR6) and were somewhat resistant to transduction by LAPSN(1E). The weak resistance of the 1E-infected CHO/LAAUUSN cells to transduction by LAPSN(1E) is somewhat surprising given that significant levels of interference have previously been described with this Env [4]. The titer we observed was only 10 fold lower than that observed in mock infected CHO/LAAUUSN cells, but was reproduced in multiple independent experiments. Taken together, these results demonstrate NRI for xenotropic and polytropic viruses in CHO cells expressing the AAUU chimeric receptor, similar to that observed previously for xenotropic and polytropic viruses. Figure 2 Analysis of AKR6 and 1E virus interference in CHO cells expressing the AAUU and AAAU chimeric receptors. CHO cells transduced by retroviral vectors expressing the chimeric receptors AAUU or AAAU were infected with AKR6 or 1E viruses by maintenance of the cells in virus-containing medium or in standard medium (mock infected) for six weeks. After infection the cells were seeded into 6-cm-diameter dishes, were exposed to vectors bearing the indicated Env, and vector titers were determined. Data from two independent infection/vector-titer-measurement experiments, one represented by grey boxes and the other by black boxes, are shown. Titer measurements in each experiment were performed in triplicate. The interference patterns on CHO/AAAU cells were markedly different from those described for CHO/AAUU cells. The AAAU receptor contains only a single entry determinant that can be utilized by both AKR6 and 1E pseudotyped viruses. In cells expressing this receptor, transduction by the LAPSN(AKR6) or LAPSN(1E) vectors was blocked by the presence of either AKR6 or 1E Env (Figure 2, right panels), thus showing a pattern of reciprocal interference. Although transduction by LAPSN(AKR6) was not completely blocked by 1E Env, a similar degree of interference was observed in two independent experiments, and the observed differences in titer were found to be statistically significant in both cases by using the Student's t-test (p < 0.05). In summary, these experiments demonstrate a non-reciprocal interference pattern between AKR6 and polytropic viruses on the AAUU chimera, and a reciprocal pattern of interference in the AAAU chimera, which contains only the putative fourth ECL of human Xpr1. These results support the hypothesis that xenotropic virus can utilize either the third or fourth ECL of hXpr1 for cell entry, but that polytropic virus can only use the fourth ECL. When the third ECL is replaced with the non-functional loop from haXpr1, both viruses can only use the fourth ECL for entry and therefore show reciprocal interference. SU domains of AKR6 and 1E Env show high sequence similarity to prototypical xenotropic and polytropic Env SU domains To characterize the interaction of AKR6 and 1E Env proteins with Xpr1 in more detail, we isolated and cloned the receptor-binding surface (SU) subunits from both proteins. The sequence of the SU region of each Env protein was determined by sequencing a PCR fragment isolated from Hirt DNA extracted from virus-infected dunni cells. Amino acid sequence alignments of AKR6 and 1E SU regions and the those of the prototypic NZB X-MLV [22,23] and FrMCF P-MLV [24] strains shows that the 1E sequence is most like that of the FrMCF virus and the AKR6 sequence is most like that of the NZB sequence (Figure 3). For example, the 1E Env sequence contains a four residue deletion relative to NZB and AKR6 xenotropic Env proteins that is also present in the FrMCF polytropic Env. Additional sequence differences between the Env proteins, many of which occur in two variable regions, are likely to account for differences in host range observed between these viruses. Figure 3 Amino acid sequence comparison of the Env SU domains of AKR6 X-MLV, 1E P-MLV, and prototypic X-MLV and P-MLV. Amino acid alignment of the Env SU domains of NZB X-MLV [GenBank:K02730], AKR6 X-MLV [GenBank:DQ199948], 1E P-MLV [GenBank:DQ199949], and FrMCF P-MLV [GenBank:X01679]. Sequences are shown starting with the initiator methionine and include endoplasmic reticulum signal sequences of unknown lengths. Variable regions A and B, believed to be responsible for receptor recognition [45], are indicated by brackets. Non-conservative amino acids differences are indicated by black boxes and conservative changes are indicated by grey boxes. Blue boxes indicate amino acids that are identical among the P-MLVs but dissimilar from one or more of those of the X-MLVs, identical among the X-MLVs but dissimilar from one or more of those of the P-MLVs, or both. Cyan boxes indicate amino acids that are identical among the P-MLVs and similar to those of the X-MLVs, identical among the X-MLVs and similar to those of the P-MLVs, or both. A full-length env gene containing the cloned AKR6 SU sequence and the transmembrane (TM) subunit sequence from NZB X-MLV was constructed and was transfected into LGPS/LAPSN cells to generate LAPSN(AKR6env) virus. The titer of this virus on dunni cells was 3 × 104 AP+ FFU/ml. To verify the identity of the cloned AKR6 Env, we measured the titer of the LAPSN(AKR6env) vector on dunni cells previously infected with replication competent AKR6 or 1E viruses (Figure 4A). LAPSN(AKR6env) transduction of dunni/AKR6 cells was almost completely blocked (<10 AP+ FFU/ml). In contrast, the titer of this vector on dunni/1E cells was reduced by only about 10-fold. As a control, the titer of LAPSN(10A1) vector on dunni and dunni/AKR6 cells was also measured. The 10A1 Env utilizes Pit1 and/or Pit2 for entry, and so should not be affected by the presence of AKR6 xenotropic Env in the cells. As expected, the LAPSN(10A1) titers were equivalent on these cell lines (Figure 4A). The block to LAPSN(AKR6env) transduction in cells chronically infected with AKR6 suggests that the cloned sequence encodes a protein that binds the same receptor as biological isolates of AKR6. Furthermore, the infection patterns observed on dunni/AKR6 and dunni/1E cells are consistent with the NRI previously observed for X-MLV and P-MLV. Figure 4 Binding and interference properties of cloned AKR6 SU and 1E SU. (A) LAPSN(AKR6env) and LAPSN(10A1env) vector titers were measured on dunni cells and dunni cells infected with replication-competent AKR6 or 1E viruses. Data shown are means ± SD of at least two independent experiments with duplicate determinations in each experiment. (B) Binding of 1E-SU-IgG to dunni cells and to dunni cells infected with replication-competent viruses. (C) Binding of Ampho-SU-IgG to dunni cells infected with 4070A amphotropic virus. Data in (B) and (C) are from a representative experiment and show data from ~18,000 live cells (cells that exclude propidium iodide) per histogram. A full-length env gene containing the cloned 1E SU sequence and the transmembrane (TM) subunit sequence from NZB X-MLV was constructed and was transfected into LGPS/LAPSN cells, but vector production from these cells was not detected. Examination of multiple 1E-SU PCR clones isolated from various Hirt preparations of 1E virus DNA indicated that the 1E-SU clone we used to construct the Env expression vector does not contain inactivating mutations. Attempts to clone the remaining TM sequence from 1E Env by PCR using primers to conserved regions of Env were unsuccessful, suggesting that 1E may have unique sequences present in the TM domain that are required for proper Env function. To verify that the cloned 1E SU sequence had the properties of a polytropic virus SU domain, we generated a human IgG tagged version of 1E-SU (1E-SU-IgG). Following production of the protein by transient transfection and purification by FPLC, we examined the binding of 1E-SU-IgG to dunni cells by flow cytometry (Figure 4B). To address the binding specificity of this reagent, and by extension of our cloned SU sequence, we also examined the binding to dunni cells infected with replication competent 1E or with 4070A amphotropic viruses. Similar binding of 1E-SU-IgG was observed in both control and dunni/4070A, whereas reduced binding was observed in dunni/1E cells. As a control, we found that Ampho-SU-IgG protein binding to dunni cells was inhibited in cells infected by an amphotropic virus (Figure 4C). The ability of replication competent 1E virus to inhibit binding of 1E-SU-IgG to cells demonstrates that the cloned SU recognizes a protein that is also bound by the 1E virus isolate. From this result, we conclude that the cloned SU sequence is representative of the Env present in the 1E virus. Analysis of xenotropic and polytropic Env binding to cells expressing human, hamster and chimeric receptors The ability of AKR6-pseudotype vector to utilize chimeric receptors that contain either of two non-overlapping regions of hXpr1 suggests that this virus can bind independently to either of the two regions of the cellular receptor. To test this prediction, we measured binding of AKR6 virus to CHO cells expressing various receptors by FACS analysis (Figure 5) using a rat antibody (83A25) that recognizes epitopes in the C-terminus of Env but does not interfere with virus binding to cells [25]. We found a clear increase in AKR6 virus binding to cells expressing hXpr1 in comparison to cells expressing haXpr1. AKR6 virus binding to cells expressing the AAAU chimeric receptor was similar to that of cells expressing hXpr1, consistent with the ability of the AAAU chimera to mediate entry of vectors bearing the AKR6 Env. Interestingly, AKR6 virus binding to cells expressing the AAUA chimera was much higher than that of cells expressing hXpr1. It is important to note that we have not determined the relative cell surface expression levels of the receptors and receptor chimera, and it is possible that differences in binding reflect varied protein levels as opposed to differences in binding affinities. However, binding of the AKR6 virus to cells expressing the AAUA and AAAU chimeras at levels at least as high as to cells expressing hXpr1 is consistent with the hypothesis that the AKR6 Env can independently bind the third or the fourth ECL of hXpr1. Figure 5 Measurement of AKR6 virus binding to cells expressing chimeric receptors. CHO cells transduced with retroviral vectors expressing hamster, human or chimeric Xpr1 receptor proteins were incubated with or without LAPSN(AKR6) virus and virus binding was detected by flow cytometry using the 83A25 anti-Env primary and a fluorescent secondary antibody. Each histogram represents 14,000 to 18,000 live cells (cells that exclude propidium iodide). The experiments were repeated twice with similar results. The 1E-pseudotype vector could only utilize chimeric receptors that contained the fourth ECL of hXpr1, suggesting that only chimeric receptors containing the fourth ECL of hXpr1 would bind the 1E Env. In this case we could not measure 1E virus binding to cells because the 83A25 rat antibody did not bind to the 1E Env (data not shown), in agreement with previous data showing that 83A25 does not recognize Env from some strains of FrMCF [25]. Instead, to measure 1E Env binding we measured binding of the 1E-SU-IgG protein to cells expressing the chimeric receptors (Figure 6). 1E-SU-IgG binding to hXpr1 was higher than that to haXpr1, consistent with the difference in receptor activities of these proteins. 1E-SU-IgG binding to cells expressing the AAUA chimeric receptor was similar to that for cells expressing hXpr1 while binding to cells expressing the AAAU chimera was higher than that to AAUA- or hXpr1-expressing cells. These results indicate that the 1E Env can bind most efficiently to a receptor containing the fourth ECL (AAAU), but equal binding of 1E Env to AAUA and human Xpr1 was not expected based on the 1E vector transduction data. As with the AKR6 virus binding studies above, it is possible that differences in receptor expression may have influenced these results. In addition, there is relatively high binding of 1E-SU-IgG to haXpr1, a poor receptor for 1E-pseudotype vectors. Figure 6 Measurement of 1E-SU-IgG binding to cells expressing chimeric receptors. CHO cells transduced with retroviral vectors expressing hamster (AAAA, green), human (UUUU, red) or chimeric (AAUA, orange; AAAU, blue) Xpr1 receptor proteins were incubated with (solid lines) or without (dashed lines) purified 1E-SU-IgG, with fluorescent anti-IgG secondary antibody, and were analyzed by flow cytometry. All analyses were performed in the same experiment with the same FACS settings. Each histogram represents ~13,000 live cells (cells that exclude propidium iodide). The experiment was repeated once with similar results. Discussion Results obtained here with the hamster/human receptor chimeras are consistent with previous studies demonstrating the importance of residues within the putative third and fourth ECL of Mus dunni Xpr1 in xenotropic receptor function [21]. In that study, mutations in both the third and fourth ECL of Mus dunni Xpr1 were required to abolish xenotropic receptor function while mutations in either ECL alone did not limit virus entry. In the current study, the ability of AKR6 pseudotyped vectors to utilize either the AAUA or the AAAU chimera as a receptor demonstrates that either the third or fourth human ECL is sufficient to support X-MLV entry. Taken together, our experiments with chimeric receptors suggest a model for entry of X-MLV and P-MLV that is consistent with the NRI observed previously, given that no X-MLV specific receptor has been identified. We propose that two receptor functions are present simultaneously in different domains of Xpr1. One domain, which resides in the fourth ECL functions as a recognition site for both xenotropic and polytropic viruses, while the second receptor domain in the third ECL can only interact efficiently with xenotropic Env. Our model for NRI predicts that the xenotropic and polytropic viruses should show a reciprocal pattern of interference in a receptor lacking the X-MLV specific receptor domain. The interference experiments described here using the AAAU and AAUU chimeras confirm this prediction. The interference pattern on the AAUU chimera, which contains both entry domains, is non-reciprocal due to the presence of the third extracellular loop. If the xenotropic specific determinant is removed, as in the AAAU chimera, X-MLV entry is markedly inhibited in cells expressing the 1E Env. This finding demonstrates that the third ECL is required for NRI, and that a chimeric receptor lacking this region serves as a common receptor for both P-MLV and X-MLV. In the interference experiments described here, 1E Env was sometimes unable to completely block infection by a 1E-pseudotype challenge vector (Table 2). Previous work suggests that such incomplete interference may reflect an inherent inability of P-MLV to completely block their cellular receptor. In vitro studies specifically examining the mechanism of P-MLV pathogenesis have shown that infection of cells by polytropic/MCF viruses results in accumulation of unintegrated extrachromosomal viral DNA, suggesting that P-MLV are capable of superinfecting cells in culture [26]. This finding is consistent with studies from other oncoretroviral systems showing that pathogenic viral stains can often superinfect cells [27-29]. Given that receptor mediated interference is the primary mechanism by which viruses prevent superinfection, the demonstrated ability of P-MLV to initiate multiple rounds of infection suggests that some polytropic Env proteins are inherently incapable of blocking certain receptors. However, it should be noted that strong interference by polytropic Env proteins can be observed in some cases (Table 2) [4]. It is tempting to speculate that the regions we have identified through our chimera analyses represent the motifs within Xpr1 that are responsible for binding to the viral Env. The critical portions of the molecule are believed to lie outside of the cell, and therefore represent candidates for SU binding domains. However, it is difficult to accurately predict the topology of transmembrane receptors, as was shown in the case of Pit1 and Pit2. Initial predictions of receptor topology were used to design a number of chimeras similar to those described here. Regions within those chimeras were identified that enhanced infection by GALV or amphotropic MLV respectively, and it was suggested that these regions were responsible for virus binding [30-33]. However, recent experiments have provided a new, experimentally verified topology for Pit2 [34], and several of the previously identified critical regions were found to lie on the inner surface of the cell membrane. Therefore, before a specific role can be firmly assigned to the third and fourth ECL of Xpr1, the topology of the protein must be established. Conclusion Results presented here indicate that the non-reciprocal interference between polytropic and xenotropic retroviruses can be explained by a common receptor domain in the putative fourth ECL of Xpr1 and a specific receptor domain for xenotropic virus in the third ECL of the same Xpr1 protein. Methods Virus and cell line nomenclature Cell lines containing integrated retroviral vectors are indicated by the name of the cell line, followed by a slash, followed by the name of the integrated vector (e.g. dunni/LAPSN, or CHO/LN). Retroviral vectors in the viral form are described by the vector name followed, in parentheses, by the name of the replication-competent virus or packaging cell line used to produce the vector [e.g. LAPSN(AKR6), LAPSN(PA317)]. Where packaging cell lines have been used, the Gag and Pol proteins are from Moloney murine leukemia virus. Cell culture Chinese hamster ovary (CHO) cells (CHO-K1, ATCC CCL 61) were grown in minimum essential medium-alpha (α-MEM) (Gibco) supplemented with 10% fetal bovine serum (FBS) (Hyclone). All other cell lines were grown in Dulbecco's minimal essential medium (DMEM) (Gibco) supplemented with 10% FBS. CHO cells expressing chimeric receptors were generated by calcium phosphate-mediated transfection of receptor expression constructs. One day post-transfection, cells were trypsinized and seeded at 1:10 dilution into medium containing G418 (750 μg active compound per ml) and were maintained in selection medium for 7 to 10 days. Surviving cells were pooled and utilized in subsequent transduction assays. Mus dunni tail fibroblasts (dunni cells), the generation of dunni/LN, dunni/LAPSN, and helper virus-infected derivatives have been described [4]. LGPS/LAPSN cells [35] are a clone of NIH 3T3 cells that express Moloney MLV Gag and Pol proteins and contain the retroviral vector LAPSN [6]. Retrovirus packaging cell lines used included PA317 [36], PD223 [37] and FlyRD [38]. All cells were grown in a 37°C incubator at 10% CO2 and 100% relative humidity. Chimeric receptor plasmids and retroviral vectors Receptor chimeras are named to indicate the origin of the sequence in each putative extracellular loop, based on the receptor topology model provided in Figure 1. This model has been suggested in previous studies [21], and was confirmed for this study by using a number of topology prediction algorithms located on the ExPASy proteomics server [39]. For the human/hamster Xpr1 receptor chimeras (Figure 1), "A" indicates sequence from the Cricetulus griseus hamster receptor [GenBank:AF198106], while a "U" is used for the human sequence derived from a HeLa cell cDNA library [GenBank:AF099082]. Chimeric Xpr1 proteins were constructed by exchanging restriction fragments as indicated in Figure 1. The 2 kb DNA fragments containing the hXpr1 or haXpr1 coding regions were blunt ended with Klenow and was cloned into SmaI digested pBluescript II (Stratagene, La Jolla CA). Following the exchange of fragments required to generate chimeric receptors, all constructs were confirmed by sequencing using primers internal to the receptor sequence. Retroviral vectors expressing the chimeric receptors were made by insertion of 2 kb XhoI-BamHI fragments containing the receptor coding regions from pBluescript into the retroviral expression plasmid LXSN [40] after digestion of pLXSN with HpaI and BamHI. Additional retroviral vectors used here included LAPSN [6], which encodes AP and Neo, and LN [40], which encodes Neo. Viruses and infection assays The AKR6 xenotropic and 1E polytropic virus isolates were a kind gift from Bruce Chesebro [14]. LAPSN(AKR6) and LAPSN(1E) retroviral vectors were generated by infecting dunni/LAPSN cells with AKR6 or 1E helper virus, as described previously [4]. LAPSN(AKR6env) and LAPSN(1Eenv) vectors were generated by transfection of pSX2-AKR6env and pSX2-1Eenv into LGPS/LAPSN cells using standard calcium phosphate protocols. Briefly, LGPS/LAPSN cells were plated into 6-cm-diameter culture dishes at 5 × 105 cells per dish approximately16 h prior to transfection. The following day, 9 μg of the Env expression plasmid was transfected into the cells with 1 μg of pCMV-βgal as a control for transfection efficiency. The following day cells were rinsed with PBS, and incubated with 4 ml culture medium per plate overnight. The conditioned medium was collected, filtered through a 0.45 μm pore-size filter, and was frozen at -80°C. Vector titers were determined by limiting dilution assay on dunni cells. Additional viral vectors, including LAPSN (PA317), LAPSN (PD223), and LAPSN(FlyRD), were obtained by collecting conditioned medium from established producer lines. Transduction assays in cell lines expressing chimeric receptors were carried out as follows. Approximately 16 h before infection, cell lines were plated at 7 × 104 cells/well into 6-well (d = 3.4 cm) tissue culture dishes. Immediately prior to infection, medium was changed to include 4 μg/ml Polybrene. Virus was added at appropriate dilutions, and the cells incubated for 48 h to allow expression of the alkaline phosphatase protein from the integrated LAPSN vector. Cells were then fixed in 3.7% formaldehyde in phosphate-buffered saline for 8 min at room temperature. Fixed cells were washed three times with phosphate-buffered saline. Endogenous alkaline phosphatase was inactivated by incubating the cells at 68°C for 1 h. Cells were then stained for alkaline phosphatase activity by incubating the cells over night in AP staining buffer (100 mM Tris pH 8.5, 100 mM NaCl, 50 mM MgCl2, 1mg/ml Nitro Blue tetrazolium, 100 μg/ml X-Phos). Transduction events were measured by counting AP+ foci. Env cloning Env SU sequences from the AKR6 [GenBank:DQ199948] and 1E [GenBank:DQ199949] viruses were obtained by PCR from low molecular weight DNA obtained from infected cells. Specifically, dunni cells were plated at 105 cells in 6-cm-diameter tissue culture dishes. Following overnight incubation, the cells were infected at high multiplicity of infection (~100) with helper virus-containing stocks of LAPSN(AKR6) and LAPSN(1E) in the presence of 4 μg/ml Polybrene (Sigma). 16 h post-infection, low molecular weight DNA was isolated using the method of Hirt [41]. Env sequences corresponding to the SU portion of Env were isolated by PCR using primers Xeno5'env (5'-ATGGAAGGTTCAGCGTTCTCAAAACCCC-3') and Xeno3'Env (5'-TGCCGCCCATAGTAAGTCCTCC-3'). Following gel purification using a Qiaquick gel purification kit (Qiagen), fragments were cloned into pCR2.1 using a Topo-TA cloning strategy (Invitrogen, Carlsbad CA). Full length Env coding regions were generated by ligation of a SacI-XhoI fragments into pBS-TM, a pBluescript-based vector containing a C-terminal fragment from the NZB env gene [GenBank:K02730]. The pBS-TM plasmid was made by insertion of a SacI-NotI fragment from pCSI-ENZB [16] into pBluescript II. Expression plasmids were generated by subcloning of XhoI-NotI fragments into pCR3.1 (Invitrogen) to generate pCR3.1-AKR6env and pCR3.1-1Eenv. To improve expression in murine and CHO cells, a BamHI-HincII fragment containing the human cytomegalovirus immediate early promoter was replaced with a BamHI-NheI fragment containing the Moloney MLV LTR promoter and enhancer from pSX2 [42], to generate pSX2-AKRenv and pSX2-1Eenv. These plasmids were sequenced to confirm the presence of complete Env open reading frames. The 1E-SU-IgG plasmid was generated by ligation of a SacI-XhoI fragment from pCR2.1-1E-Env into pCI-NSU?9-hFc [16]. To confirm the identity and integrity of the resulting fusion protein, the construct was sequenced using primers internal to the 1E-SU. Virus and Env SU binding assays Production and purification of 1E-SU-IgG for binding assays was carried out as described for other similar proteins [43,44]. For flow cytometry assays, 106 cells were incubated with 1–2 μg of purified fusion protein in a final volume of 100 μl for 2 h. Following washing, cells were incubated with a fluorescent anti human-IgG secondary antibody (DAKO F0315) for 1 h. Cell fluorescence was determined by flow cytometry on a FACSCalibur (BD Biosciences), and data was analyzed using CellQuest software. For virus binding assays, 106 cells expressing the indicated receptor chimeras were incubated with LAPSN(AKR6) virus at 4°C for two h. Cells were washed three times with phosphate-buffered saline containing 2% FBS and were incubated with 1 ml hybridoma supernatant containing 83A25 antibody for 2 h. Following two additional washes and incubation with a FITC-conjugated anti-Rat-IgG secondary antibody, cell fluorescence was determined by flow cytometry using a FACSCalibur. Interference assays To establish CHO cell lines expressing high levels of AKR6 and 1E Env, cells were maintained in conditioned medium from dunni/LN cells (mock), or dunni/LN cells productively infected with AKR6 or 1E helper viruses. Conditioned medium (α-MEM with 10% FBS) was collected, centrifuged at 1,000 × g for 10 min to remove cells and debris, and frozen at -80°C for 24 h. Prior to addition to CHO cells, a 1:1 mixture of dunni conditioned medium and fresh α-MEM with 10% FBS was supplemented with 4 μg/ml Polybrene to facilitate infection. The conditioned medium mix was added to cells every 24 h. As CHO cultures reached confluence (approximately every 3 days) cells were removed from the culture dish with trypsin/EDTA and split 1:10 into new 6-cm-diameter dishes. After 6 weeks, cells were trypsinized, counted on a hemacytometer and plated at 105 cells/well in 6 well dishes. Cells were then transduced with LAPSN(AKR) or LAPSN(1E) viral vectors. The titer of each vector was determined by limiting dilution. The degree of interference can be determined by comparing the vector titer on mock infected cells to that obtained on cells infected with AKR6 or 1E viruses. Competing interests The author(s) declare that they have no competing interests. Authors' contributions NSVH helped design the study, carried out the experiments, analyzed the data, and drafted the manuscript. ADM helped design the study and write the manuscript. Acknowledgements This study was supported by grants HL54881, DK47754, and HL36444 from the NIH. ==== Refs Weissenhorn W Dessen A Calder LJ Harrison SC Skehel JJ Wiley DC Structural basis for membrane fusion by enveloped viruses Mol Membr Biol 1999 16 3 9 10332732 10.1080/096876899294706 Colman PM Lawrence MC The structural biology of type I viral membrane fusion Nat Rev Mol Cell Biol 2003 4 309 319 12671653 10.1038/nrm1076 Sommerfelt MA Weiss RA Receptor interference groups of 20 retroviruses plating on human cells Virology 1990 176 58 69 1691887 10.1016/0042-6822(90)90230-O Miller AD Wolgamot G Murine retroviruses use at least six different receptors for entry into Mus dunni cells J Virol 1997 71 4531 4535 9151846 Ott D Friedrich R Rein A Sequence analysis of amphotropic and 10A1 murine leukemia viruses: close relationship to mink cell focus-inducing viruses J Virol 1990 64 757 766 2153240 Miller DG Edwards RH Miller AD Cloning of the cellular receptor for amphotropic murine retroviruses reveals homology to that for gibbon ape leukemia virus Proc Natl Acad Sci USA 1994 91 78 82 8278411 van Zeijl M Johann SV Closs E Cunningham J Eddy R Shows TB O'Hara B A human amphotropic retrovirus receptor is a second member of the gibbon ape leukemia virus receptor family Proc Natl Acad Sci U S A 1994 91 1168 1172 8302848 Miller DG Miller AD A family of retroviruses that utilize related phosphate transporters for cell entry J Virol 1994 68 8270 8276 7966619 O'Hara B Johann SV Klinger HP Blair DG Rubinson H Dunn KJ Sass P Vitek SM Robins T Characterization of a human gene conferring sensitivity to infection by gibbon ape leukemia virus Cell Growth Differ 1990 1 119 127 2078500 Takeuchi Y Vile RG Simpson G O'Hara B Collins MK Weiss RA Feline leukemia virus subgroup B uses the same cell surface receptor as gibbon ape leukemia virus J Virol 1992 66 1219 1222 1309898 Brojatsch J Naughton J Rolls MM Zingler K Young JA CAR1, a TNFR-related protein, is a cellular receptor for cytopathic avian leukosis-sarcoma viruses and mediates apoptosis Cell 1996 87 845 855 8945512 10.1016/S0092-8674(00)81992-3 Adkins HB Brojatsch J Naughton J Rolls MM Pesola JM Young JA Identification of a cellular receptor for subgroup E avian leukosis virus Proc Natl Acad Sci U S A 1997 94 11617 11622 9326659 10.1073/pnas.94.21.11617 Adkins HB Blacklow SC Young JA Two functionally distinct forms of a retroviral receptor explain the nonreciprocal receptor interference among subgroups B, D, and E avian leukosis viruses J Virol 2001 75 3520 3526 11264341 10.1128/JVI.75.8.3520-3526.2001 Chesebro B Wehrly K Different murine cell lines manifest unique patterns of interference to superinfection by murine leukemia viruses Virology 1985 141 119 129 2983494 10.1016/0042-6822(85)90188-6 Tailor CS Nouri A Lee CG Kozak C Kabat D Cloning and characterization of a cell surface receptor for xenotropic and polytropic murine leukemia viruses Proc Natl Acad Sci U S A 1999 96 927 932 9927670 10.1073/pnas.96.3.927 Battini JL Rasko JE Miller AD A human cell-surface receptor for xenotropic and polytropic murine leukemia viruses: possible role in G protein-coupled signal transduction Proc Natl Acad Sci USA 1999 96 1385 1390 9990033 10.1073/pnas.96.4.1385 Yang YL Guo L Xu S Holland CA Kitamura T Hunter K Cunningham JM Receptors for polytropic and xenotropic mouse leukaemia viruses encoded by a single gene at Rmc1 Nat Genet 1999 21 216 219 9988277 10.1038/6005 Kozak CA Genetic mapping of a mouse chromosomal locus required for mink cell focus-forming virus replication J Virol 1983 48 300 303 6310150 Kozak CA Susceptibility of wild mouse cells to exogenous infection with xenotropic leukemia viruses: control by a single dominant locus on chromosome 1 J Virol 1985 55 690 695 2991590 Spain BH Koo D Ramakrishnan M Dzudzor B Colicelli J Truncated forms of a novel yeast protein suppress the lethality of a G protein alpha subunit deficiency by interacting with the beta subunit J Biol Chem 1995 270 25435 25444 7592711 10.1074/jbc.270.43.25435 Marin M Tailor CS Nouri A Kozak SL Kabat D Polymorphisms of the cell surface receptor control mouse susceptibilities to xenotropic and polytropic leukemia viruses J Virol 1999 73 9362 9368 10516044 Levy JA Pincus T Demonstration of Biological Activity of a Murine Leukemia Virus of New Zealand Black Mice Science 1970 170 326 327 4318971 Levy JA Xenotropic viruses: murine leukemia viruses associated with NIH Swiss, NZB, and other mouse strains Science 1973 182 1151 1153 4356281 Hartley JW Wolford NK Old LJ Rowe WP A new class of murine leukemia virus associated with development of spontaneous lymphomas Proc Natl Acad Sci U S A 1977 74 789 792 191826 Evans LH Morrison RP Malik FG Portis J Britt WJ A neutralizable epitope common to the envelope glycoproteins of ecotropic, polytropic, xenotropic, and amphotropic murine leukemia viruses J Virol 1990 64 6176 6183 1700832 Yoshimura FK Wang T Nanua S Mink cell focus-forming murine leukemia virus killing of mink cells involves apoptosis and superinfection J Virol 2001 75 6007 6015 11390602 10.1128/JVI.75.13.6007-6015.2001 Reinhart TA Ghosh AK Hoover EA Mullins JI Distinct superinfection interference properties yet similar receptor utilization by cytopathic and noncytopathic feline leukemia viruses J Virol 1993 67 5153 5162 8394443 Chen IS Temin HM Establishment of infection by spleen necrosis virus: inhibition in stationary cells and the role of secondary infection J Virol 1982 41 183 191 6283112 Weller SK Joy AE Temin HM Correlation between cell killing and massive second-round superinfection by members of some subgroups of avian leukosis virus J Virol 1980 33 494 506 6245246 Johann SV van Zeijl M Cekleniak J O'Hara B Definition of a domain of GLVR1 which is necessary for infection by gibbon ape leukemia virus and which is highly polymorphic between species J Virol 1993 67 6733 6736 8411375 Pedersen L Johann SV van Zeijl M Pedersen FS O'Hara B Chimeras of receptors for gibbon ape leukemia virus/feline leukemia virus B and amphotropic murine leukemia virus reveal different modes of receptor recognition by retrovirus J Virol 1995 69 2401 2405 7884886 Pedersen L van Zeijl M Johann SV O'Hara B Fungal phosphate transporter serves as a receptor backbone for gibbon ape leukemia virus J Virol 1997 71 7619 7622 9311843 Lundorf MD Pedersen FS O'Hara B Pedersen L Amphotropic murine leukemia virus entry is determined by specific combinations of residues from receptor loops 2 and 4 J Virol 1999 73 3169 3175 10074169 Salaun C Rodrigues P Heard JM Transmembrane topology of PiT-2, a phosphate transporter-retrovirus receptor J Virol 2001 75 5584 5592 11356966 10.1128/JVI.75.12.5584-5592.2001 Miller AD Garcia JV von Suhr N Lynch CM Wilson C Eiden MV Construction and properties of retrovirus packaging cells based on gibbon ape leukemia virus J Virol 1991 65 2220 2224 1850008 Miller AD Buttimore C Redesign of retrovirus packaging cell lines to avoid recombination leading to helper virus production Mol Cell Biol 1986 6 2895 2902 3785217 Wolgamot G Rasko JE Miller AD Retrovirus packaging cells expressing the Mus dunni endogenous virus envelope facilitate transduction of CHO and primary hematopoietic cells J Virol 1998 72 10242 10245 9811768 Cosset FL Takeuchi Y Battini JL Weiss RA Collins MK High-titer packaging cells producing recombinant retroviruses resistant to human serum J Virol 1995 69 7430 7436 7494248 The ExPASy proteomics server Miller AD Rosman GJ Improved retroviral vectors for gene transfer and expression Biotechniques 1989 7 980 990 2631796 Hirt B Selective extraction of polyoma DNA from infected mouse cell cultures J Mol Biol 1967 26 365 369 4291934 10.1016/0022-2836(67)90307-5 Miller AD Chen F Retrovirus packaging cells based on 10A1 murine leukemia virus for production of vectors that use multiple receptors for cell entry J Virol 1996 70 5564 5571 8764070 Kurre P Kiem HP Morris J Heyward S Battini JL Miller AD Efficient transduction by an amphotropic retrovirus vector is dependent on high-level expression of the cell surface virus receptor J Virol 1999 73 495 500 9847355 Van Hoeven NS Miller AD Improved enzootic nasal tumor virus pseudotype packaging cell lines reveal virus entry requirements in addition to the primary receptor Hyal2 J Virol 2005 79 87 94 15596804 10.1128/JVI.79.1.87-94.2005 Battini JL Heard JM Danos O Receptor choice determinants in the envelope glycoproteins of amphotropic, xenotropic, and polytropic murine leukemia viruses J Virol 1992 66 1468 1475 1310758	16354307	PMC1325250	CC BY	2021-01-04 16:24:29	no	Retrovirology. 2005 Dec 15; 2:76	utf-8	Retrovirology	2,005	10.1186/1742-4690-2-76	oa_comm
==== Front Part Fibre ToxicolParticle and Fibre Toxicology1743-8977BioMed Central London 1743-8977-2-111633225410.1186/1743-8977-2-11ResearchCarbon black nanoparticles induce type II epithelial cells to release chemotaxins for alveolar macrophages Barlow Peter G [email protected] Anna [email protected] Ken [email protected] Janis [email protected] Vicki [email protected] M.R.C/University of Edinburgh Centre for Inflammation Research, Queen's Medical Research Institute, 47 Little France Crescent, Edinburgh EH16 4TJ, UK2 Biomedicine Research Group, Napier University, 10 Colinton Road, Edinburgh EH10 5DT, UK3 ELEGI/Colt Laboratories, M.R.C/University of Edinburgh Centre for Inflammation Research, Queen's Medical Research Institute, 47 Little France Crescent, Edinburgh EH16 4TJ, UK2005 6 12 2005 2 11 11 9 8 2005 6 12 2005 Copyright © 2005 Barlow et al; licensee BioMed Central Ltd.2005Barlow 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 Alveolar macrophages are a key cell in dealing with particles deposited in the lungs and in determining the subsequent response to that particle exposure. Nanoparticles are considered a potential threat to the lungs and the mechanism of pulmonary response to nanoparticles is currently under intense scrutiny. The type II alveolar epithelial cell has previously been shown to release chemoattractants which can recruit alveolar macrophages to sites of particle deposition. The aim of this study was to assess the responses of a type II epithelial cell line (L-2) to both fine and nanoparticle exposure in terms of secretion of chemotactic substances capable of inducing macrophage migration. Results Exposure of type II cells to carbon black nanoparticles resulted in significant release of macrophage chemoattractant compared to the negative control and to other dusts tested (fine carbon black and TiO2 and nanoparticle TiO2) as measured by macrophage migration towards type II cell conditioned medium. SDS-PAGE analysis of the conditioned medium from particle treated type II cells revealed that a higher number of protein bands were present in the conditioned medium obtained from type II cells treated with nanoparticle carbon black compared to other dusts tested. Size-fractionation of the chemotaxin-rich supernatant determined that the chemoattractants released from the epithelial cells were between 5 and 30 kDa in size. Conclusion The highly toxic nature and reactive surface chemistry of the carbon black nanoparticles has very likely induced the type II cell line to release pro-inflammatory mediators that can potentially induce migration of macrophages. This could aid in the rapid recruitment of inflammatory cells to sites of particle deposition and the subsequent removal of the particles by phagocytic cells such as macrophages and neutrophils. Future studies in this area could focus on the exact identity of the substance(s) released by the type II cells in response to particle exposure. ==== Body Background Exposure to the combustion-derived nanoparticulate component of particulate matter (PM) has been implicated in adverse effects in toxicology [1-3] and epidemiology [4-9] studies. Nanoparticles have at least one dimension less than 100 nm, similar to ultrafine particles which have a diameter of less than 100 nm. Nanoparticles have been reported to display increased toxicity compared to an equivalent mass of fine particles of the same material when inhaled into the rat lung [10]. This has been attributed to the oxidative stress emanating from the greater surface area of the nanoparticles [11-13] and their potential to interact with other components of PM such as transition metals [12]. Macrophages are the primary defence against inhaled particulates on the airspace surfaces. Following particle deposition, macrophages phagocytose particles and transport them out of the lung via the mucociliary escalator (MCE). It is hypothesised that the type II epithelial cells in the alveolar space play an important role in the modulation of inflammatory processes within the lung. Studies have shown that type II alveolar epithelial cells can release inflammatory cytokines such as rantes [14], monocyte chemoattractant protein-1 (MCP-1) [14], interleuklin-8 (IL-8) [16], tumour necrosis factor alpha (TNFα) [15] and the complement protein C5a [17] which can recruit leukocytes to sites of inflammation by acting as chemoattractants [16,17]. Driscoll et al., [18] demonstrated that α-quartz can induce an increase in MIP-2 mRNA in type II cells which could contribute to the accumulation of neutrophils seen in quartz-exposed lungs. Driscoll et al., [15] has also demonstrated that TNFα, produced by macrophages exposed to particles, can induce type II cells to release chemokines such as MIP-2. Jimenez et al., [19] showed the same effect in macrophages exposed to PM10. Barrett et al., [20] reported that the production of MIP-2 and MCP-1 by a murine type II cell line following silica exposure was mediated by TNFα which, together with silica exposure, could also induce ROS production by the type II cells. Becher et al., [21] also demonstrated that release of type II cell IL-6, TNFα and MIP-2 differed depending on the type of stone particle that the cells were exposed to. O'Brien et al., [14] showed that macrophages migrated towards conditioned medium obtained from type II cells treated with IL-1α. Chemokines in the conditioned medium such as rantes, MCP-1 and Granulocyte Monocyte Colony Stimulating Factor (GM-CSF) were the agents responsible for the chemotaxis. Stimulation of type II cells with bradykinin has also been shown to induce the production of proteins that are chemotactic for neutrophils and monocytes [22]. Paine III et al., [23] observed increased MCP-1 production by type II cells stimulated with IL-1 and TNFα. We have previously shown that carbon black nanoparticles can activate serum to produce factors that induce macrophage migration [24]. This manuscript describes a study in which two cell lines are used as a model to examine the effects of fine and nanoparticle forms of both carbon black and titanium dioxide on potential of type II cells to generate substances that induce macrophage migration. We hypothesised that exposure of a type II cell line to fine and nanoparticles results in the release of chemotactic proteins from the type II cells. We further hypothesise that, in the mammalian lung, these proteins may play an important role in the recruitment of macrophages towards sites of particle deposition and inflammation. Results Lactate dehydrogenase (LDH) release from L-2 cells following treatment with fine & nanoparticle carbon black and titanium dioxide Figures 1(a) and 1(b) show that LDH release from the L-2 cells increased in a dose dependant manner upon treatment with all particle types until the highest release induced by nanoparticle TiO2 and carbon black was reached at a particle concentration of 1 mg/ml. Maximal release of LDH following treatment with fine TiO2 and carbon black occurred at a particle concentration of 2 mg/ml. The 1 mg/ml treatments of nanoparticle carbon black and TiO2 induced a significant increase in LDH release from L-2 cells compared to the negative control (p < 0.05). Nanoparticle carbon black also resulted in a significant increase in LDH release (p < 0.05) at a particle concentration of 2 mg/ml. High doses of both types of fine particles were also observed to induce similar increases in LDH release but this was not statistically significant. However, maximal LDH release induced by particle treatments was noted to be lower than that observed following cell lysis with Triton X-100 (positive control) possibly due to high particle concentrations interfering with the assay. The negative control in these experiments (untreated cells) showed minimal LDH release. Low doses of particulates (31–500 μg) did not appear to show any significant size, composition or surface area-related differences in response. Figure 1 LDH release from L-2 cells following exposure to varying doses of fine and nanoparticle TiO2 (a) and carbon black (b) for 24 hours. The positive control (cells lysed with Triton X-100) and the negative control (untreated cells) are also shown. All values are the mean of three experiments conducted in triplicate ± SEM. The asterisk denotes a significant difference of the effect of both types of nanoparticles from the negative control (* p < 0.05). Concentration and time course studies with zymosan activated serum (ZAS) and J774.2 macrophages Figure 2(a) shows the effects of varying concentrations of ZAS on macrophage migration following a 6 hour incubation period. The graph shows that the absorbance increases as the concentration of ZAS increases from 2 to 20% indicating increased numbers of macrophages migrating into or through the polycarbonate filter. Significant increases in macrophage migration were observed at both the 10% and 20% dilutions of ZAS. This was compared to the negative control of foetal bovine serum (FBS) incubated with sterile saline, which showed no significant increases in macrophage migration. Figure 2 (a) The effect of varying concentrations of ZAS on macrophage migration compared to a negative control of non-activated serum. (6 hour incubation time). Asterisks denote a significant difference from the control (* p < 0.05). (b) A time course examination of macrophage migration towards 10% ZAS compared to a negative control of non-activated serum. Asterisks denote a significant difference from the control (*** p < 0.001). Figure 2(b) shows a time course study of macrophage migration towards 10% ZAS. The results indicate that the ZAS induced significant increases (p < 0.001) in macrophage migration at both the 3 hour and 6 hour time points. The highest level of macrophage migration was found at the 6 hour time point where migration was found to be 8-fold greater than that of the negative control. The negative control (FBS incubated with sterile saline) did not display any potential for macrophage migration and this may serve as an indicator that random macrophage migration, chemokinesis, was minimal. Macrophage migration stimulated by conditioned medium from L-2 Cells exposed to TNFα L-2 cells were exposed to TNFα in order to induce the secretion of chemotactic molecules (Figure 3). Supernatants from cells treated with TNFα at a concentration of 1 ng/ml, induced significant macrophage migration (p < 0.05) compared to the negative control (conditioned medium obtained from untreated cells). Macrophage migration was also noted at both the lower and higher TNFα concentrations although this was not significant. The other negative control, TNFα alone, at comparable concentrations to those used for cell treatments, induced modest macrophage migration. Figure 3 The chemotactic effect of conditioned medium from L-2 cells exposed to varying doses of TNFα compared with TNFα alone. Asterisks denote a significant change from the cell free TNFα treatment (* p < 0.05). Graph shows the result of one experiment with three replicates. Macrophage migration stimulated by conditioned medium from L-2 Cells exposed to fine and nanoparticle carbon black and TiO2 When compared to the negative control (conditioned medium obtained from untreated cells), nanoparticle carbon black treatment of L-2 cells generated a conditioned medium that induced a significant increase (p < 0.01) in macrophage migration (Figure 4). In comparison, conditioned medium from L-2 cells treated with fine TiO2, CB and nanoparticle TiO2 did not induce any significant increases in macrophage migration. It should also be noted that both fine and nanoparticle carbon black treatment of L-2 cells did induce significant increases in macrophage migration when compared to another negative control; medium incubated with the particles alone (p < 0.01 and p < 0.001 for fine and nanoparticle carbon black respectively). Figure 4 Macrophage migration induced by L-2 cell supernatant following treatment with 125 μg/ml fine and nanoparticle TiO2 and carbon black (striped bars indicate cell free treatments). ZAS and conditioned media from L-2 cells treated with TNFα (1 ng/ml) represent positive controls. Asterisks denote a significant increase compared to the negative control (* p < 0.05). Significant increases from the appropriate cell-free treatments are denoted by a dollar ($ p < 0.01). SDS-PAGE electrophoresis of conditioned medium from L-2 cells treated with a sub-toxic dose of nanoparticle carbon black A non-reducing SDS-PAGE gel was used to separate the protein components of the conditioned medium obtained from L-2 cells treated for 24 hours with either fine or nanoparticle carbon black (125 μg/ml). Cells were also incubated with serum-free medium for 24 hours and this conditioned medium was run together with the two other treatments as a negative control. As can be seen from the gel (Figure 5a), only protein bands corresponding to very high molecular weights were identifiable in all treatments used. Figure 5 SDS-PAGE gels of conditioned medium from L-2 cells treated with 125 μg fine or nanoparticle carbon black for 24 hours. The negative control is represented by cells treated with medium only (M). Gel A was run under non-reducing conditions and Gel B was run under reducing conditions via the addition of DTT. Marker bands and corresponding molecular weights (kDa) are indicated at the left hand side of the gels. M = Medium Only, F = Fine carbon black treatment, N = Nanoparticle carbon black treatment Figures 5b shows conditioned medium obtained from similar particle treatments run under reducing conditions. As can be seen from the gel, an increased number and density of protein bands are observed in the lane containing the conditioned medium obtained from cells that have been treated with nanoparticle carbon black. Bands are clearly noted around the 14 kDa weight and the density of this band is increased in the lane containing the conditioned medium from the nanoparticle treated cells. The negative control in both of these gels, conditioned medium obtained from untreated cells, showed minimal protein banding only under reducing conditions. Chemotactic activity of size fractions of conditioned medium obtained from L-2 cells treated with a sub-toxic dose of nanoparticle carbon black The conditioned medium generated by nanoparticle carbon black treatment (125 μg/ml) of L-2 cells was size-fractionated by centrifugation through size selective filters. It was found that the fraction sizes of 5–10 kDa and 10–30 kDa both induced a significant increase in macrophage migration (p < 0.01) when compared to the negative control (conditioned medium obtained from untreated cells). The other two fractions of 0–5 and >30 kDa did not induce a significant increase in macrophage migration (figure 6). Figure 6 Macrophage migration induced by centrifuge fractions of conditioned medium from L-2 cells treated with nanoparticle carbon black (125 μg/ml). The positive control is indicated by the yellow bar representing migration towards ZAS. The negative control was conditioned medium obtained from untreated cells. Asterisks denote a significant increase compared to the negative control (* p < 0.05). Discussion Combustion-derived nanoparticles have been implicated in the adverse effects of ambient particulate air pollution (reviewed in [1]) In addition, the toxicological impact of nanoparticles has now come under increased scrutiny due to the emergence of a relatively new technological discipline; nanotoxicology. Studies have shown that nanoparticles can be more toxic than an equivalent mass dose of fine particles comprised of the same material [10,11,25,26] The increased toxicity of nanoparticles can be attributed to several factors that depend on the nanoparticle under consideration, but for low toxicity, low solubility particles such as those used in the present study. the larger surface area of the particles and resulting oxidative stress appears to be the dominant property [1,27]. In this study, the effects of particle treatments on type II cells and their potential to induce macrophage migration is investigated. When treating L-2 cells with the particles used in this study, the concentration of LDH detected in the conditioned medium increased in a dose dependant manner. Maximal LDH release was detected at doses of 1–2 mg/ml of all particles tested. These high doses would never be attained in the lungs under plausible exposure conditions but were used in the study for completeness. The nanoparticles demonstrated a non-significant trend of greater toxicity towards the cells compared to the fine particles. This was indicated by significant increases in LDH concentrations following treatment with both types of nanoparticles. While these results may provide an accurate representation of particle cytotoxicity by measuring LDH concentrations, it is possible that LDH may have been adsorbed on to the surface of the particles following release from the cells. Desai and Richards [28] and Jones et al., [29] both noted how different types of biological molecules could be adsorbed on to the surface of inorganic and mineral dusts. Brown et al., [30] also showed extensive binding of Bovine Serum Albumin (BSA) to the surface of nanoparticle carbon black. This may occur when measuring LDH concentrations, especially when taking into account the large surface area of the nanoparticles and the potential for protein adsorption. Zymosan activated serum has been used as a positive stimulatory agent to induce macrophage chemotaxis in numerous studies with particles [31-33]. Zymosan A is a polysaccharide extracted from yeast cell walls when added to serum, activates the complement system generating large quantities of the macrophage chemoattractant C5a In this study, the J774 macrophages migrated towards ZAS in a dose-dependant manner with significant increases in macrophage migration being observed at ZAS concentrations of 10% and 20% (p < 0.001). Therefore, it was decided, based upon a review of the relevant literature and these results, that 10% ZAS would be an adequate concentration to use as a positive control in future studies using the chemotaxis chamber. Other macrophage chemotaxis studies have routinely allowed at least 3 hours incubation time to allow macrophages to complete shape change and migration [31-34]. While 3 hours incubation appeared to allow for a significant level of macrophage migration to take place, our results indicated that 6 hours would provide more opportunity for the macrophages to migrate through the filter. This decision took into consideration the fact that future migration studies would be testing a very dilute conditioned medium and as such, macrophages responses could possibly be slower than those observed with Zymosan activated serum. Prior to conducting particle treatments on L-2 cells, a preliminary experiment was conducted to determine if stimulation by a pro-inflammatory cytokine such as TNFα could induce the type II cells to release macrophage chemoattractants. A significant increase in macrophage migration was observed when the L-2 cells were treated with TNFα at a concentration of 1 ng/ml for 6 hours. A similar macrophage response was not observed in the cell-free supernatants i.e. TNFα at equivalent concentrations that had been incubated for 6 hours in cell-free wells. Although increased migration was observed at TNFα concentrations of 0.1 and 10 ng/ml this was not statistically significant. The 10 ng/ml TNFα concentration may be expected to induce a dose-dependant increase in migration but it is possible that the TNFα concentration could be approaching toxic levels and that this may have actually inhibited migration. The results from this experiment indicate that the L-2 cells were capable of releasing macrophage chemoattractants following pro-inflammatory stimulation. The fact that similar increases in macrophage migration were not observed with the TNFα alone indicated that residual TNFα itself in the medium could not explain the chemotactic response and that chemoattractants released by the L-2 cells into the conditioned medium were responsible. These results correlate well with those published by Standiford et al., [35], Paine III et al., [23] and O'Brien et al., [14] who reported that TNFα is able to induce the secretion of chemoattractants from type II epithelial cells. Studies have also shown that IL-9 also induces the release of chemoattractants from bronchial epithelial cells [36]. However, we chose TNFα as the stimulant since it has been shown to be present in the BAL fluid of particle treated animals in many in vivo studies [21,37,38] as well as being released by cells in vitro after particle treatment [19,39]. The results obtained from the LDH experiments indicated that 125 ug/ml of particles would act as a suitable sub-toxic dose for further investigation. L-2 cells were treated at that dose with four different types of particles for 24 hours and the conditioned medium that was generated was tested for potential to induce macrophage migration. The nanoparticle carbon black treatment of the L-2 cells generated a conditioned medium that induced a significant increase in macrophage migration compared to the negative control, whereas none of the other particle treatments had a significant effect on macrophage migration. This would indicate that the carbon black nanoparticles, as well as causing increased cytotoxicity at high dose, are able to stimulate the release of macrophage chemoattractants when exposed to L-2 cells at sub-toxic doses. Cell-free treatments were also conducted to determine whether any soluble substances present on the surface of the particles had the potential to directly induce macrophage migration but no such migration was stimulated. This took into consideration studies by Milanowski et al., [34] which showed macrophage chemotaxis towards bacterial and fungal products extracted from organic dust samples. In fact, the migration towards media treated with particles was observed to be slightly lower (although not significantly lower) than the control. This may suggest that the particles may bind or inactivate factors in the media that stimulate cell migration or that they release factors that inhibit migration. Two nanoparticles were used in these studies, TiO2 and carbon black, but only the carbon black showed any effect in stimulating the release of chemotaxins. This is in accord with previous findings that fine TiO2 was least inflammogenic of a panel of four nanoparticle types, including nanoparticle carbon black, and had least surface free radical activity as measured in a plasmid DNA scission assay [40]. Although both are classified as low toxicity low solubility particles, nanoparticle carbon black has a much greater surface area than nanoparticle TiO2 and this may be a major factor in generation of free radicals in vivo [41]. Following the experiments using four different particle types, it was decided that further experiments should focus upon the effects of the nanoparticle carbon black on the L-2 cells and the elucidation of the products that were released from the type II cells as a result of particle exposure. SDS-PAGE gels of conditioned medium obtained from L-2 cells treated with 125 μg of either fine or nanoparticle carbon black were run in non-reducing and reducing conditions. In non-reducing conditions, no bands were visible below the 45 kDa molecular weight. However, with the addition of the reducing agent DTT, a higher number of protein bands were noted in the nanoparticle carbon black conditioned medium compared to the medium obtained from the fine carbon black and untreated cells. This indicates that an increased number and concentration of proteins were released from the type II cells into the conditioned medium as a result of the nanoparticle carbon black exposure. One or more of these proteins may account for the increased macrophage chemotactic activity in conditioned medium obtained from such cells. The increased bands were present in the molecular weight range of approximately 14 – 35 kDa. A subsequent experiment where the conditioned medium was centrifuged through molecular weight filters and then tested for chemotactic activity indicated that the substance(s) that induced macrophage migration was located in the 5 – 30 kDa weight range. This is based on the observation that the 5–10 kDa and 10–30 kDa fractions of the conditioned medium induced a significant increase in macrophage migration compared to the negative control. A number of chemotactic proteins such as MCP-1, MIP-1, MIP-2 and rantes are known to be located within these size ranges [42]. There are important points to take note of when drawing comparisons between in vitro studies and the dosimetry of environmental particle exposure in individuals. Obviously, exposure to a given particulate can vary depending upon a person's occupation, location, method of travel, behaviour etc. Using a miners group as a model, however, Kuempel et al., [43] estimated that a coal miner will inhale >350 g of coal dust over a lifetime, and that around 40 g of this will be deposited in the alveolar region of the lung. This figure is not applicable to most people, but draws attention to the potential particle exposure which can occur in a given occupation or workplace. This is a chronic exposure versus the acute exposure used in our study but is important nonetheless. Impaired clearance and protracted exposure of epithelial cells to particles are key factors when attempting to identify the reasons behind the increased toxicity induced by nanoparticles. Although some of the doses used in this study may not be physiologically relevant, they can provide some idea of how type II cells may react in response to acute and, depending on clearance efficacy, sub-chronic exposures to particles. The phenomenon of rat lung overload in response to high exposures to low toxicity low solubility particles includes, amongst other important pathological sequelae, failure of clearance with subsequent retention of particles [44]. Under conditions of overload there is greatly enhanced contact between particles and epithelium, due to the high rate of particle deposition, exceeding the macrophage's capacity to phagocytose them [45]. Overload is driven by surface area [46] and so the findings described here provide a mechanism that may help to explain the retention of particle-loaded macrophages in the lung periphery during overload. By this mechanism, the extensive interaction between particle surfaces and epithelial cells characteristic of overload would cause release of chemotaxins that would 'attract' the particle-loaded macrophages to the alveolar region. This gradient would directly compete with the normal chemotactic gradient that would draw such macrophages up the respiratory tract for muco-ciliary clearance. Renwick et al., [26] demonstrated that the same types of nanoparticles used in this study, induced an increase in the potential for macrophages to migrate towards a positive chemoattractant, ZAS. In this manscript we have shown that these particles also possess the ability to induce type II epithelial cells to release elevated levels of macrophage chemoattractants. The chemoattractants can stimulate two events, the first being the migration of macrophages towards epithelial cells exposed to particles, as in the case of epithelial cell exposure to nanoparticle carbon black. This event will recruit cells and promote inflammation, but may potentially decrease subsequent clearance if the signal does not subside. In the second instance, a chemotactic gradient may stimulate removal of phagocytic cells out of the lung via the mucociliary escalator, promoting particle clearance. The data shown here demonstrates that nanoparticle carbon black was the most potent tested at eliciting chemotaxin release. This may suggest that such nanoparticles could reverse the normal chemotactic gradient and prevent clearance leading to overload at a lower mass burden than larger particles and this is in fact, amply supported by data (reviewed in [47]). These are conflicting interpretations of particle effects and require further work to determine the exact pathways governing particle clearance. Conclusion The regulation and stimulation of leukocyte recruitment in the lung is an important factor in the regulation of inflammation and several regulatory mechanisms exist. The response of macrophages to paracrine signals appears to be an important driver for inflammation and macrophages respond to chemotactic signals from neutrophils [31], type II cells [14] and from other macrophages [48]. Other studies have focussed on the inhibition of macrophage chemotaxis induced by molecules such as the macrophage migration inhibitory factor (MIF), described by Hermanowski-Vosatka et al., [49] which inhibits macrophage movement. However, as mentioned previously, several studies have demonstrated that type II cells can secrete a wide range of pro-inflammatory mediators capable of inducing macrophage migration to sites of inflammation. It is likely that, due to the large surface area of carbon black nanoparticles and their subsequent ability to induce oxidative stress [11], there is signalling for expression of genes for chemotaxins by the type II cells. This may indicate an adaptive response of lung epithelial cells in contact with deposited particles which would aid in the rapid recruitment of inflammatory cells to the sites of particle deposition, and the subsequent removal of the particles by phagocytic cells such as macrophages and neutrophils. Obvious future studies in this area could pursue the identity of the substance(s) released by the type II cells in response to particle exposure. This study indicates that the factors inducing macrophage migration are in the molecular weight range between 5 and 30 kDa. However, due to the wide range of substances released by the type II cells, it is likely that no one single substance is responsible for inducing macrophage recruitment and, as such, further experiments in this area may involve deciphering the complex mixtures of chemotactic molecules released by the type II cells to ascertain their specific cellular targets. Synergistic or potentiative interactions between the chemotactic molecules should also be taken into consideration. Other investigations may attempt to examine the effects of other forms of particles as well as various components of particulate air pollution. Materials and methods Chemicals and reagents L-Glutamine (200 mM); penicillin-streptomycin (1000 μg/ml) and foetal bovine serum (FBS) were obtained from Invitrogen, UK. RPMI 1640 (without L-glutamine), Zymosan A and Bovine Serum Albumin (BSA) were obtained from Sigma Chemicals Company, Dorset, UK. The Rapi Diff 2 (Romanowsky) stain set was obtained from Raymond A Lamb, London. All other chemicals and reagents were purchased from Sigma Chemicals Company, Dorset, UK unless otherwise stated. Culture of L-2 type II epithelial cell line The rat alveolar type II cell line L-2 was obtained from the European Collection of Animal Cell Cultures (Salisbury, England). The cells were grown in 25 cm2 tissue culture flasks in RPMI 1640 medium supplemented with 1% L-glutamine, 1% penicillin/streptomycin and 10% heat inactivated FBS. Cultures were incubated in a humidified incubator at 37°C/5% CO2. Cell counts were performed using 0.3% trypan blue exclusion (1:1 dilution with cells) and an improved Neubauer haemocytometer. Cells were removed from the flask by adding 1 × trypsin/EDTA in Hanks balanced salt solution (HBSS) and incubating for 5 minutes, then centrifuged at 900 g for 2 minutes and resuspended in RPMI 1640 supplemented with 10% FBS. Culture of J774.2 macrophage cell line The adherent murine monocytic-macrophage cell line J774.2 was obtained from the European Collection of Animal Cell Cultures (Salisbury, England). The cells were grown in 25 cm2 tissue culture flasks in RPMI 1640 medium supplemented with 1% L-glutamine, 1% penicillin/streptomycin and 10% heat-inactivated FBS. Culture flasks were stored in a humidified incubator at 37°C and 5% CO2. Cell counts and viability were assessed using an improved Neubauer haemocytometer and trypan blue exclusion. All cells that were found to be non-adherent in the culture flasks were discarded by washing prior to use. Particles The particles used in these experiments were fine carbon black (H. Haeffner & Co Ltd., Chepstow, UK), fine titanium dioxide (Tioxide Ltd.) and nanoparticle carbon black and titanium dioxide (Degussa, UK). The particle composition is identical for both types of carbon black and both types of TiO2 respectively. However, the varying diameter and surface area of the particles are described in Table 1. Table 1 Mean diameter and surface area of particles used Particle Type Mean Diameter (nm) Surface Area (m2/g) Fine TiO2 250.0 6.6 Nanoparticle TiO2 29.0 49.78 Fine Carbon Black 260.2 7.9 Nanoparticle Carbon Black 14.3 253.9 Particle data taken from Renwick et al., [26] Cell treatments Cells were seeded at 40,000 cells per well in a 96 well plate and incubated for 24 hours in RPMI 1640 supplemented with 10% FBS. Serial dilutions (62.5 – 2000 μg/ml by mass dose) of each type of particle were prepared in serum-free RPMI and sonicated in a water bath sonicator for 5 minutes before use. After 24 hours, the medium was removed from the cells and replaced with appropriate concentrations of particle before incubation for a further 24 hours. Following particle treatments, the medium was removed from the cells and centrifuged for 30 minutes at 15000 g to remove the particles. The supernatants were removed and stored in separate tubes at -80°C for use in the LDH assay. LDH assay Positive control samples for the LDH assay were prepared by lysing cells with 50 μl of 0.1% Triton X-100 dissolved in phosphate buffered saline (PBS). The solutions were then centrifuged at 11,000 g for three minutes and 10 μl of the supernatant was used as a positive control to indicate total releasable LDH. A negative control was obtained by incubating the cells with serum-free medium. To create a standard curve, pyruvate standards were prepared (using distilled H2O and 1 mg/ml NADH dissolved in 0.75 mM sodium pyruvate), ranging in concentrations equivalent to 0 – 2000 Units/ml of LDH. The standards were transferred (60 μl) in triplicate, to a 96-well plate. Into all of the remaining wells of the plate, 50 μl of NADH was added. The samples were incubated at 37°C/5% CO2 for 5 minutes. The test and control samples (10 μl) were added to the relevant wells on the plate containing the NADH solution. The plate was incubated at 37°C, 5% CO2 for 30 minutes. Following incubation, 50 μl of 2, 4-dinitrophenylhydrazine (0.2 mg/ml dissolved in 1 M HCl) was added to all of the wells and the plate was left for 20 minutes at room temperature. Sodium hydroxide (50 μl of 4 M concentration) was then added to each well and the plate was left at room temperature for 5 minutes. The plate was shaken for 5 seconds and the absorbance of all the wells was measured at 540 nm with a Dynex MRX microplate reader. Macrophage chemotaxis assay The chemotaxis apparatus utilised in these studies was a re-usable 96-well Neuroprobe chemotaxis chamber (Receptor Technologies, UK). Each sample (30 μl) was loaded, in triplicate, into the bottom wells of the chamber. A Neuroprobe polycarbonate filter (pore size 5 μM) was inserted between the layers. J774.2 macrophages (2 × 105) in 200 μl of serum-free RPMI 1640 were added to the top of each well. The chamber was incubated at 37°C in 5% CO2 for 6 hours and the filter was removed and washed three times with PBS on the upper side to remove non-migrated macrophages. The filter was stained with a Romanowsky (Diff-Quick) stain. The optical density of each well on the filter was read at 540 nm in a Dynex multiwell plate reader. Increasing absorbance correlates with the increasing number of macrophages moving through the filter. Zymosan activated serum (ZAS) ZAS is a known activator of C5a in blood serum and, as such, has been utilised in numerous studies as a positive control in chemotaxis assays. A stock solution of Zymosan A in saline (10 mg/ml) was prepared and sonicated for 5 minutes. The stock Zymosan solution (100 μl) was added to 900 μl of FBS. To prepare a control solution, 100 μl of sterile saline was added to 900 μl serum. Both solutions were incubated in a shaking water bath at 37°C for 2 hours and then at 56°C for 30 minutes. The solutions were centrifuged for 5 minutes at 2000 g then aliquoted and stored at -80°C until use. For use in the chemotaxis assay, both solutions were diluted to the appropriate concentration using serum-free RPMI 1640. To determine the optimum incubation times and ZAS concentrations for use in the chemotaxis assay, a solution of ZAS (1 mg/ml) and a negative control were prepared as detailed above. For the concentration response experiments, both solutions were diluted to 2%, 5%, 10% and 20% with RPMI 1640 medium. Solutions were then tested using the chemotaxis protocol as described above. For the time course experiments, 10% ZAS solution was prepared and tested in the chemotaxis chamber as detailed above. The cells were incubated in the chamber for 45 minutes, 90 minutes, 3 hours or 6 hours before the filter was removed and stained. Exposure of L-2 cells to TNFα L-2 cells were seeded in a 96-well plate at a cell count of 40,000 cells per well and incubated overnight. Rat TNFα (Biosource, UK) was prepared at three different concentrations; 0.1, 1 (positive control) and 10 ng/ml using sterile PBS containing 0.1% bovine serum albumin. The medium from the L-2 cells was removed and replaced with 200 μl of TNFα at the required concentration. TNFα was also added at an equivalent concentration to wells that did not contain any L-2 cells in order to provide a cell-free comparison. Negative controls were prepared by adding serum-free RPMI to both the L-2 cells and the cell-free wells. The plate was incubated in a humidified incubator at 37°C and 5% CO2 for 6 hours. Following incubation, the supernatants were removed and tested for potential to induce macrophage migration using the chemotaxis assay protocol as described previously. Exposure of L-2 cells to particles A 96-well plate was seeded with L-2 cells (40,000 cells per well) and incubated overnight in a humidified incubator at 37°C and 5% CO2. Suspensions of fine and nanoparticle carbon black and titanium dioxide were prepared in serum-free RPMI at a concentration of 125 μg/ml. The suspensions were sonicated for 10 minutes prior to use. A 1 ng/ml TNFα solution was also prepared in 0.1% BSA in PBS to be used as a positive control. The serum-free particle suspensions and TNFα were added to the relevant wells in the 96-well plate with L-2 cells. A duplicate cell-free 96-well plate was also prepared using identical solutions of TNFα and serum-free particle suspensions. Both 96-well plates were subsequently incubated in a humidified incubator at 37°C and 5% CO2. After 24 hours incubation, the two plates were removed and the supernatants were extracted and stored in eppendorf tubes. The eppendorf tubes were centrifuged at 15,000 g for 15 minutes and the supernatants transferred to fresh tubes. The samples were stored at -80°C prior to analysis. Preparation of L-2 conditioned medium Trypsin/EDTA in HBSS (1×) was added to a flask of L-2 cells at approximately 80% confluence and incubated for 5 minutes at 37°C/5% CO2. Serum-free RPMI culture medium (10 mls) was added to the cells and the cells were centrifuged at 900 g for 2 minutes. The supernatant was decanted and the cells were resuspended in normal culture medium. Cell counts were assessed and adjusted to 1 million cells per ml. The cells were seeded in a 6 well plate at 1 million cells per well and incubated overnight. To prepare particle suspension, fine and nanoparticle carbon black was suspended in serum-free RPMI at a concentration of 1 mg/ml and sonicated for 10 minutes before diluting to 125 μg/ml. The serum-free particle suspensions (200 μl) were added to the 6-well plate and incubated for 24 hours. Cells were also incubated with serum-free RPMI for 24 hours. Following incubation, the conditioned medium was removed and centrifuged at 15,000 g for 15 minutes and the supernatants were pooled and stored at -80°C until use. SDS-PAGE gel electrophoresis of L-2 conditioned medium A 15% separating gel was prepared by mixing 23% distilled H2O, 50% acrylamide mix, 25% Tris HCl (1.5 M, pH 8.8), 1% sodium dodecyl sulphate (SDS, 10%) and 1% ammonium persulphate (APS, 10%). TEMED was added at a concentration of 0.04% immediately prior to pipetting the gel into the casting apparatus. A stacking gel was prepared by mixing 68% H2O, 17% acrylamide mix, 12.5% Tris HCl (0.5 M, pH 6.8), 1% SDS (10%) and 1% APS (10%). TEMED was added at a concentration of 0.1% immediately prior to adding the gel to the separating gel in the casting apparatus. Immediately following addition of the stacking gel to the casting apparatus, 1 μl of bromophenol blue was added to the stacking gel. To prepare the samples, 20 μl of L-2 conditioned medium was pipetted into an eppendorf. For the non-reducing gel (Figure 5a), 30 μl of sample buffer was added to the conditioned medium. For the reducing gel (Figure 5b), 27 μl of sample buffer and 3 μl of dithiothreitol (DTT) was added to each of the samples. Both sets of samples were then incubated on a heating block; 1 minute incubation for non-reducing samples and 2 minutes for reducing samples. After heating, the samples were immediately stored on ice. Once cast, the gels were transferred to an electrophoresis tank and immersed in 1 × buffer. Samples (20 μl) as well as coloured molecular weight markers were added to the gels and the electrophoresis was run at 40 mA for 45 minutes. Both gels were stained with the Coomassie blue stain overnight and destained by microwaving (3 × 1 minute) in fresh distilled water. Differential centrifugation of L-2 conditioned medium Conditioned medium, obtained from L-2 cells treated with nanoparticle carbon black according to the protocol above, was separated according to molecular weight using Vivaspin 2 centrifuge filters operated according to manufacturers' instructions. A 30 kDa filter was used first, followed by a 10 kDa filter and finally a 5 kDa filter. Supernatants were stored at -80°C until tested for potential to induce macrophage migration using the chemotaxis protocol outlined above. The negative control in this experiment was conditioned medium obtained from cells that had not been treated with any particles Statistical analysis Results are expressed as the mean ± SEM. All figures are the results of three separate experiments with three observations in each unless otherwise stated. Statistical analysis was accomplished by a one-way ANOVA with Tukey's multiple comparison. Competing interests The author(s) declare that they have no competing interests. Authors' contributions PB aided in the conception of the study, drafted the manuscript and performed all assays. ACB participated in the conception and design of the study. KD participated in the design of the study and scientific contribution to the manuscript. JM participated in critical assessment and scientific contribution to the manuscript. VS supervised the study and participated in the data analysis and drafting of the manuscript. All authors have read and approved the final manuscript. Acknowledgements This study was funded by Napier University ==== Refs Donaldson K Tran CL Jimenez LA Duffin R Newby DE Mills N Combustion-derived nanoparticles: A review of their toxicology following inhalation exposure Part Fibre Toxicol 2005 2 10 16242040 10.1186/1743-8977-2-10 Morawska L Zhang JJ Combustion sources of particles. 1. Health relevance and source signatures Chemosphere 2002 49 1045 1058 12492164 10.1016/S0045-6535(02)00241-2 Neumann HG Health risk of combustion products: toxicological considerations Chemosphere 2001 42 473 479 11219671 10.1016/S0045-6535(00)00219-8 Pope CA IIIHill RW Villegas GM Particulate air pollution and daily mortality on Utah's Wasatch Front Environ Health Perspect 1999 107 567 573 10379003 Laden F Neas LM Dockery DW Schwartz J Association of fine particulate matter from different sources with daily mortality in six U.S. cities Environ Health Perspect 2000 108 941 947 11049813 Schwartz J What are people dying of on high air-pollution days Environ Res 1994 64 26 35 8287840 10.1006/enrs.1994.1004 Ibald-Mulli A Wichmann HE Kreyling W Peters A Epidemiological evidence on health effects of ultrafine particles J Aerosol Med 2002 15 189 201 12184869 10.1089/089426802320282310 Wichmann HE Peters A Epidemiological evidence of the effects of ultrafine particle exposure Phil Trans R Soc Lond A 2000 358 2563 2565 10.1098/rsta.2000.0668 Clancy L Goodman P Sinclair H Dockery DW Effect of air-pollution control on death rates in Dublin, Ireland: an intervention study Lancet 2002 360 1210 1214 12401247 10.1016/S0140-6736(02)11281-5 Ferin J Oberdörster G Penney DP Pulmonary retention of ultrafine and fine particles in rats Am J Respir Cell Mol Biol 1992 6 535 542 1581076 Beck-Speier I Dayal N Karg E Maier KL Schumann G Schulz H Oxidative stress and lipid mediators induced in alveolar macrophages by ultrafine particles Free Radic Biol Med 2005 38 1080 1092 15780766 10.1016/j.freeradbiomed.2005.01.004 Wilson MR Lightbody JH Donaldson K Sales J Stone V Interactions between ultrafine particles and transition metals in vivo and in vitro Toxicol Appl Pharmacol 2002 184 172 179 12460745 10.1006/taap.2002.9501 Duffin R Tran CL Clouter A Brown DM MacNee W Stone V Donaldson K The importance of surface area and specific reactivity in the acute pulmonary inflammatory response to particles Ann Occup Hyg 2002 46 242 245 O'Brien AD Standiford TJ Christensen PJ Wilcoxen SE Paine R III Chemotaxis of alveolar macrophages in response to signals derived from alveolar epithelial cells J Lab Clin Med 1998 131 417 424 9605106 10.1016/S0022-2143(98)90142-1 Driscoll KE Carter JM Hassenbein DG Howard B Cytokines and particle-induced inflammatory cell recruitment Environ Health Perspect 1997 105 1159 1164 9400717 Simon RH Paine R III Participation of pulmonary alveolar epithelial cells in lung inflammation J Lab Clin Med 1995 126 108 18 7636384 Strunk RC Eidlen DM Mason RJ Pulmonary alveolar type II epithelial cells synthesize and secrete proteins of the classical and alternative complement pathways J Clin Invest 1988 81 1419 1426 2966814 Driscoll KE Howard BW Carter JM Asquith T Johnston C Detilleux P Kunkel SL Isfort RJ Alpha-quartz-induced chemokine expression by rat lung epithelial cells: effects of in vivo and in vitro particle exposure Am J Pathol 1996 149 1627 1637 8909252 Jimenez LA Drost EM Gilmour PS Rahman I Antonicelli F Ritchie H MacNee W Donaldson K PM (10)-exposed macrophages stimulate a pro-inflammatory response in lung cells via TNF-alpha Am J Physiol Lung Cell Mol Physiol 2002 282 L237 L248 11792628 Barrett EG Johnston C Oberdorster G Finkelstein JN Silica-induced chemokine expression in alveolar type II cells is mediated by TNFα-induced oxidative stress Am J Physiol Lung Cell Mol Physiol 1999 276 L979 L988 Becher R Hetland RB Refsnes M Dahl JE Dahlman HJ Schwarze PE Rat lung inflammatory responses after in vivo and in vitro exposure to various stone particles Inhal Toxicol 2001 13 789 805 11498806 10.1080/089583701316941311 Koyama S Sato E Nomura H Kubo K Miura M Yamashita T Nagai S Izumi T Bradykinin stimulates type II alveolar cells to release neutrophil and monocyte chemotactic activity and inflammatory cytokines Am J Pathol 1998 153 1885 1893 9846978 Paine R IIIWolfe MW Standiford TJ Burdick MD Rollins BJ Strieter RM MCP-1 expression by rat type II alveolar epithelial cells in primary culture J Immunol 1993 150 4561 4570 8482848 Barlow PG Donaldson K MacCallum J Clouter A Stone V Serum exposed to nanoparticle carbon black displays increased potential to induce macrophage migration Toxicol Lett 2005 155 397 401 15649623 10.1016/j.toxlet.2004.11.006 Stone V Shaw J Brown DM MacNee W Faux SP Donaldson K The role of oxidative stress in the prolonged inhibitory effect of ultrafine carbon black on epithelial cell function Toxicol In Vitro 1998 12 649 659 10.1016/S0887-2333(98)00050-2 Renwick LC Brown D Clouter A Donaldson K Increased inflammation and altered macrophage chemotactic responses caused by two ultrafine particle types Occup Environ Med 2004 61 442 447 15090666 10.1136/oem.2003.008227 Donaldson K Stone V Current hypotheses on the mechanisms of toxicity of ultrafine particles Ann Ist Super Sanita 2003 39 405 410 15098562 Desai R Richards RJ The adsorption of biological macromolecules by mineral dusts Environ Res 1978 16 449 464 210013 10.1016/0013-9351(78)90178-0 Jones BM Edwards JH Wagner JC Absorption of serum proteins by inorganic dusts Br J Ind Med 1972 29 287 292 4339804 Brown DM Stone V Findlay P MacNee W Donaldson K Increased inflammation and intracellular calcium caused by ultrafine carbon black is independent of transition metals or other soluble components Occup Environ Med 2000 57 685 691 10984341 10.1136/oem.57.10.685 Donaldson K Slight J Brown DM Effects of products from inflammatory pulmonary neutrophils on alveolar macrophage chemotaxis, spreading, and thymidine incorporation Inflammation 1989 13 443 453 2759686 10.1007/BF00914927 Donaldson K Brown GM Brown DM Slight J Robertson MD Davis JM Impaired chemotactic responses of bronchoalveolar leukocytes in experimental pneumoconiosis J Pathol 1990 160 63 69 2156037 10.1002/path.1711600113 Warheit DB George G Hill LH Snyderman R Brody AR Inhaled asbestos activates a complement-dependent chemoattractant for macrophages Lab Invest 1985 52 505 514 3990243 Milanowski J Sorenson WG Dutkiewicz J Lewis DM Chemotaxis of alveolar macrophages in response to microbial products derived from organic dust Environ Res 1995 69 59 66 7588495 10.1006/enrs.1995.1025 Standiford TJ Kunkel SL Phan SH Rollins BJ Strieter RM Alveolar macrophage-derived cytokines induce monocyte chemoattractant protein-1 expression from human pulmonary type II-like epithelial cells J Biol Chem 1991 26 9912 9918 2033076 Little FF Cruikshank WW Center DM IL-9 stimulates release of chemotactic factors from human bronchial epithelial cells Am J Respir Cell Mol Biol 2001 25 347 352 11588013 Schins RP Lightbody JH Borm PJ Shi T Doanldson K Stone V Inflammatory effects of coarse and fine particulate matter in relation to chemical and biological constituents Toxicol Appl Pharmacol 2004 195 1 11 14962500 10.1016/j.taap.2003.10.002 Li XY Brown D Smith S MacNee W Donaldson K Short-term inflammatory response following intratracheal instillation of fine and ultrafine carbon black in rats Inhal Toxicol 1999 11 709 731 10477444 10.1080/089583799196826 Brown DM Donaldson K Borm PJ Schins RP Dehnhardt M Gilmour P Jimenez LA Stone V Calcium and ROS-mediated activation of transcription factors and TNF-alpha cytokine gene expression in macrophages exposed to ultrafine particles Am J Physiol Lung Cell Mol Physiol 2004 286 L344 L353 14555462 10.1152/ajplung.00139.2003 Dick CA Brown DM Donaldson K Stone V The role of free radicals in the toxic and inflammatory effects of four different ultrafine particle types Inhal Toxicol 2003 15 39 52 12476359 10.1080/08958370304454 Brown DM Wilson MR MacNee W Stone V Donaldson K Size-dependent proinflammatory effects of ultrafine polystyrene particles: a role for surface area and oxidative stress in the enhanced activity of ultrafines Toxicol Appl Pharmacol 2001 175 191 199 11559017 10.1006/taap.2001.9240 The RCSB Protein Data Bank Kuempel ED O'Flaherty EJ Stayner LT Smith RJ Green FHY Vallyathan V A biomathematical model of particle clearance and retention in the lungs of coal miners Regul Toxicol Pharmacol 2001 34 69 87 11502158 10.1006/rtph.2001.1479 Mauderly JL Relevance of particle-induced rat lung tumors for assessing lung carcinogenic hazard and human lung cancer risk Environ Health Perspect 1997 105 1337 1346 9400748 Tran CL Jones AD Cullen RT Donaldson K Exploration of the Mechanisms of Retention and Clearance of Low-Toxicity Particles in the Rat Lung Using a Mathematical Model Inhal Toxicol 1999 11 1077 1108 10562698 10.1080/089583799196600 Tran CL Buchanan D Cullen RT Searl A Jones AD Donaldson K Inhalation of poorly soluble particles. II. Influence Of particle surface area on inflammation and clearance Inhal Toxicol 2000 12 1113 1126 11114784 10.1080/08958370050166796 Oberdorster G Significance of particle parameters in the evaluation of exposure-dose-response relationships of inhaled particles Inhal Toxicol 1996 8 73 89 11542496 Brieland JK Flory CM Jones ML Miller GR Remick DG Warren JS Fantone JC Regulation of monocyte chemoattractant protein-1 gene expression and secretion in rat pulmonary alveolar macrophages by lipopolysaccharide, tumor necrosis factor-alpha, and interleukin-1 beta Am J Respir Cell Mol Biol 1995 12 104 109 7811465 Hermanowski-Vosatka A Mundt SS Ayala JM Goyal S Hanlon WA Czerwinski RM Wright SD Whitman CP Enzymatically inactive macrophage migration inhibitory factor inhibits monocyte chemotaxis and random migration Biochemistry 1999 38 12841 12849 10504254 10.1021/bi991352p	16332254	PMC1325251	CC BY	2021-01-04 16:24:21	no	Part Fibre Toxicol. 2005 Dec 6; 2:11	utf-8	Part Fibre Toxicol	2,005	10.1186/1743-8977-2-11	oa_comm
==== Front BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1771633666910.1186/1471-2164-6-177Research ArticleComparative analysis of programmed cell death pathways in filamentous fungi Fedorova Natalie D [email protected] Jonathan H [email protected] Geoff D [email protected] Jennifer R [email protected] William C [email protected] The Institute for Genomic Research, 9712 Medical Center Drive, Rockville, MD 20850, USA2 Faculty of Life Sciences, 1.800 Stopford Building, University of Manchester, Manchester M13 9PT, UK3 The George Washington University School of Medicine, Department of Biochemistry and Molecular Biology, 2300 Eye Street, NW Washington, DC 20837, USA2005 8 12 2005 6 177 177 20 9 2005 8 12 2005 Copyright © 2005 Fedorova et al; licensee BioMed Central Ltd.2005Fedorova 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 Fungi can undergo autophagic- or apoptotic-type programmed cell death (PCD) on exposure to antifungal agents, developmental signals, and stress factors. Filamentous fungi can also exhibit a form of cell death called heterokaryon incompatibility (HI) triggered by fusion between two genetically incompatible individuals. With the availability of recently sequenced genomes of Aspergillus fumigatus and several related species, we were able to define putative components of fungi-specific death pathways and the ancestral core apoptotic machinery shared by all fungi and metazoa. Results Phylogenetic profiling of HI-associated proteins from four Aspergilli and seven other fungal species revealed lineage-specific protein families, orphan genes, and core genes conserved across all fungi and metazoa. The Aspergilli-specific domain architectures include NACHT family NTPases, which may function as key integrators of stress and nutrient availability signals. They are often found fused to putative effector domains such as Pfs, SesB/LipA, and a newly identified domain, HET-s/LopB. Many putative HI inducers and mediators are specific to filamentous fungi and not found in unicellular yeasts. In addition to their role in HI, several of them appear to be involved in regulation of cell cycle, development and sexual differentiation. Finally, the Aspergilli possess many putative downstream components of the mammalian apoptotic machinery including several proteins not found in the model yeast, Saccharomyces cerevisiae. Conclusion Our analysis identified more than 100 putative PCD associated genes in the Aspergilli, which may help expand the range of currently available treatments for aspergillosis and other invasive fungal diseases. The list includes species-specific protein families as well as conserved core components of the ancestral PCD machinery shared by fungi and metazoa. ==== Body Background Aspergillus fumigatus is the most prevalent causative agent of invasive aspergillosis in immunocompromised patients and it can also cause asthma, allergies, and mycotoxicosis [1]. Other species of this genus including Neosartorya fischeri (teleomorph of A. fischerianus), A. flavus, A. terreus, A. niger, and A. nidulans can also cause diseases in humans, animals and plants all over the world [2]. Despite the medical and agricultural importance of this genus, limited antifungal therapies are available due to high toxicity, low efficacy rates, and growing drug resistance [3]. Activation of endogenous programmed cell death (PCD) reactions is a promising approach to combat invasive aspergillosis and other fungal diseases. Several antifungal agents including amphotericin B and rapamycin have been shown to induce cell death cascades in filamentous fungi [4-7]. Better understanding of these pathways might provide the basis for the development of novel anti-fungal therapeutics against aspergillosis and give further insights into evolution of programmed cell death in the eukaryotic cell. Several different cell death programs seem to exist in parallel in fungi and often resemble mammalian apoptosis and yeast autophagy [8-11]. The best studied form of programmed cell death in filamentous fungi is heterokaryon incompatibility (HI) that can be triggered by cellular fusions between hyphae of incompatible individuals during vegetative growth or between incompatible germlings during the establishment of fungal colonies [12]. These fusion between two individuals with incompatible het (heterokaryon incompatibility) loci triggers the HI reaction characterized by growth inhibition, repression of asexual sporulation, hyphal compartmentation and death in the heterokaryotic cell [13,14]. Although HI is ubiquitous in filamentous fungi, its biological significance and evolutionary origin is still unknown. It has been proposed to serve as a self/nonself recognition system responsible for limiting genetic exchange and horizontal transfer of cytoplasmic infectious elements [15-17]. Coincidentally, one of the HI inducers in P. anserina is a prion capable of infectious propagation [18]. Much of what is known about the underlying mechanisms of programmed cell death in fungi comes from only three species, Saccharomyces cerevisiae, P. anserina and N. crassa. With the availability of the newly sequenced genomes of A. nidulans (J. Galagan et al., in press), A. oryzae (M. Machida et al., in press), A. fumigatus Af293 (W. Nierman et al., in press), and N. fischeri (W. Nierman, unpublished) we were able to identify their putative PCD effectors and mediators. In this study, we applied a BLASTp-plus-phylogeny reconstruction approach to survey the Aspergillus spp. genomes for homologs of characterized programmed cell death proteins from fungi and animals. Several fungi-specific families as well as components of the core cell death machinery shared by filamentous fungi and metazoan were identified. Results and discussion Inducers of HI-triggered PCD Several components of the HI reaction encoded by the het genes, the HI suppressor genes, and the HI-induced genes have been characterized at the molecular level in P. anserina and N. crassa. In natural populations of P. anserina, N. crassa and Cryphonectria parasitica, the functional het genes exist as two or more polymorphic alleles conferring alternative specificities [19-22]. Different loci may be responsible for triggering the HI reaction in these two and other fungal species. Heterocomplex formation between alternative het gene products is thought be a common theme in nonself recognition during allelic and non-allelic HI [23,24]. One of the genetically (and potentially physically) interacting partners is often a HET domain protein [15] that may interact with well-conserved proteins playing important roles in development and differentiation (Dementhon and L. Glass, unpublished). Although at least eight loci have been implicated in HI in A. nidulans [25], none of them has been characterized at the molecular level. To identify putative HI-associated proteins in the Aspergilli, we first searched completely sequenced fungal genomes using known inducers and mediators of heterokaryon incompatibility as BLASTp queries. We examined the domain composition and phyletic distribution of the BLASTp hits and built phylogenetic trees for several protein families. We also applied domain fusion analysis to several so called Rosetta Stone proteins with unusual domain composition to infer protein domain interactions and functional linkage between putative HI-associated proteins identified in the Aspergilli proteomes. HET-C2 proteins Our database BLASTp searches identified orthologs of P. anserina HET-C2 [26], in all filamentous fungal genomes sequenced to date (Table 1). The family also has a high level of sequence conservation and wide phyletic distribution in other taxa. HET-C2 orthologs are found in several Saccharomycetes, including Debaryomyces hansenii and Kluyveromyces lactis, (Fig. 1). Yet no homologs are detected in S. cerevisiae and S. pombe, suggesting a gene loss in some yeast lineages. HET-C2 homologs are also present in most animals and plants. Table 1 Putative HI inducers Protein/Domain ASP SOR Scer Spom BAS Biological function/process HET-C 1 1 0 0 1 Unknown RNR1 1 1 1 1 1 Cell cycle control HET-C2 1 1 0 0 1 Cell cycle controla, sphingolipid sensing MAT-1/2 1–2 1 2 2 0 Regulation of sexual differentiation HET 7–38 38–94 0 0 0 Signaling, regulation of sexual differentiation HET-s/LopB 0–1 1–7 0 0 0 Unknown NACHT 12–19 2–18 0 0 0 Signaling, NTP binding, oligomerization Pfs+NACHT 5–7 1 0 0 0 Signaling, nucleoside metabolism ASP, A. fumigatus, N. fischeri, A. oryzae, A. nidulans; SOR, N. crassa, G. zeae, and M. grisea; BAS: Cryptococcus neoformans and Ustilago maydis; Scer, S. cerevisiae; Spom, Schizosaccharomyces pombe; a, putative function. Figure 1 Phylogenetic tree of the HET-C2/GLTP/ACD11 family of proteins. Tree reconstruction was performed as described in the Methods section. Experimentally characterized proteins are shown in yellow. The numbers indicate percent bootstrap values for selected internal branches. Hemiascomycete members including Debaryomyces hansenii, Ashbya gossypii, Yarrowia lipolytica, and Kluyveromyces lactis clustered together with the basidiomycete U. maydis. The high level of conservation among the HET-C2 family members is consistent with the important role these proteins may play in the glycosphingolipid and sphingosine metabolism and possibly in regulation of cellular stress responses. HET-C2 shows significant similarity to human GLTP [27] and Arabidopsis thaliana ACD11 [28] proteins, which catalyze the intermembrane transfer of glycosphingolipids and sphingosines, respectively. ACD11 has also been shown to function in PCD and pathogen defense in plants. In Aspergilli, sphingosines have been also shown to induce an apoptosis-like PCD [4] and to affect cell cycle progression [29]. P. anserina HET-C2 was proposed to act as a glycolipid metabolite sensor in addition to its role in glycolipid transfer, regulation of ascospore maturation, and triggering HI [26,30]. The high level of sequence conservation in this family, suggests that the role of HET-C2 orthologs in the Aspergilli PCD pathway is likely to be similar. HET-C proteins Further analysis showed that all Aspergillus species have direct orthologs of N. crassa HET-C [31] and its close homolog of unknown function (Table 1). The phylogenetic tree has a bipartite division resulting from an early gene duplication event predating the separation of Eurotiomycetes and Sordariomycetes (Fig. 2). HET-C is orthologous to A. nidulans HetC and homologous to TinC [32]. The HET-C domain is found in many ascomycetes and basidiomycete species, but surprisingly in only one yeast species, Yarrowia lipolytica, which clusters together with HET-C homologs from basidiomycete species. Unexpectedly, a partial HET-C domain is also present in several epiphytic and symbiotic bacteria including two gamma-proteobacteria, Pseudomonas syringae [GenBank: AAY39263 and GenBank: AAO58004], a cyanobacterium, Nostoc punctiforme [GenBank: ZP_00106220], and an actinomycete of the genus Frankia [GenBank: ZP_00567177]. The phylogenetic tree of the conserved N-terminal portion of HET-C shows that the bacterial proteins form a coherent clade with a long branch connecting it to the rest of the tree. We had to exclude the Nostoc punctiforme protein from the phylogenetic analysis because it was too divergent, but it also showed more similarity to the bacterial proteins. Figure 2 Phylogenetic tree of the HET-C family of proteins. Tree reconstruction was performed as described in the Methods section. Experimentally characterized proteins are shown in yellow. The numbers indicate percent bootstrap values for selected internal branches. Based on the current tree topology, the origin of the bacterial homologs is not clear. It can be attributed to vertical inheritance from the last common ancestor between bacteria and fungi followed by massive gene loss in most bacterial and yeast lineages. Alternatively, it can be explained by horizontal transfer of the ancestral het-C gene from epiphytic fungi followed by rapid divergence in bacteria. In the latter case, the gene must have persisted in bacterial populations by conferring a selective advantage to the recipients. Since heterologous expression of an N. crassa het-C allele was also shown to trigger an HI-like growth defect in P. anserina [21], the het-C homologs in P. syringae or related species may induce growth inhibition in epiphytic filamentous fungi and thus facilitate substrate defense. N. crassa polypeptides encoded by the het-C alleles of alternative specificity were shown to form a heterocomplex localized to the plasma membrane during the HI reaction [23]. It has a putative signal peptide, a conserved HET-C domain and a divergent C-terminal glycine-rich region, often found in extracellular glycoproteins. The biological role of the HetC proteins in the Aspergilli is unknown. A. nidulans TinC has been shown to stabilize the NimA mitotic kinase required for mitotic entry [32]. A. nidulans strains lacking tinC displayed cold and osmotic sensitivity and overexpression of its truncated form produced growth inhibition, defects in nuclear envelope fission and cell cycle [32]. It is unlikely that either protein triggers the HI reaction in the A. nidulans. Moreover, het-C may not act as a bona fide het gene in other fungal species, since no het-C polymorphism was observed in A. flavus (K. Ehrlich and P. Cotty, unpublished), A. nidulans [32], and P. anserina [21] isolates. Nonetheless, expression of the N. crassa het-c(PA) allele triggers a growth inhibition similar to the HI reaction in P. anserina. If N. crassa HET-C has a similar role to TinC, it may explain the growth inhibition effects caused by expression of N. crassa (and possibly bacterial) het-c in P. anserina [21]. HET domain proteins HET domain [15,33] proteins were identified using the HMMer package as described in Methods. Unlike the ubiquitous HET-C2 family, the HET domain appears to be limited to filamentous ascomycetes and is not detected in yeasts or basidiomycete species (Table 1). In the Aspergilli, the number of HET domain proteins varies from seven in N. fischeri to 38 in A. oryzae. The tree topology delineates multiple duplication events in filamentous ascomycetes species followed by rapid diversification and gene loss in several Aspergillus spp. (data not shown). Orthologous relationships within this Aspergillus family are difficult to establish, except for a subfamily of HET and Ankyrin domain proteins, which appear to be related by direct vertical descent (data not shown). The HET domain expansion in filamentous ascomycetes may represent a niche adaptation strategy to process a large number of similar stimuli associated with defense against pathogens, self/nonself recognition, differentiation, or analogous roles. It is found in N. crassa HET-6 and TOL and in P. anserina HET-D and HET-E, and so appears to be critical to the HI reaction in both species (for review see [15]). In P. anserina HET-D and HET-E, HET domains are followed by a NACHT domain and multiple WD repeats, while N. crassa proteins contain a coiled-coil domain and LRR repeats, instead (see Figure 3). In addition to HET-6 and TOL, N. crassa has about 50 other HET domain proteins, whose role in the HI reaction if any is as yet unknown. Identification of the HET-s/LopB domain Initial BLASTp searches using the P. anserina HET-s sequence [34] as a query revealed homologs in A. nidulans, P. chrysogenum, M. grisea, N. crassa and G. zeae (Table 1). Iterative PSI-BLAST searches identified a new domain that includes more proteins from the same species plus a pathogenicity protein, LopB, from the Dothideomycete fungus Leptosphaeria maculans [35]. For LopB and most other members of this family, sequence similarity is limited to the N-terminal globular domain of HET-s (Fig. 4) [36]. Two members from A. nidulans and N. crassa have an adjacent NACHT domain (described below) at the N- and C-terminus, respectively. Figure 4 Multiple alignment of the HET-s/LopB protein family. The first line in the alignment shows the prediction of secondary structure content: h for helical, e for extended, c for coiled. Residues conserved among several proteins are marked with gray shading. The polymorphic positions in HET-s and HET-S proteins are shown with red shading. Proteins are listed under their unique GenBank identifiers (first left column). Species are indicated in the second from left column: Pans, Podospora anserina; Lmac, Leptosphaeria maculans; Mgri, Magnaporthe grisea; Ncra, Neurospora crassa; Anid, Aspergillus nidulans, Gzea, Gibberella zeae. Yellow shading indicates uncharged amino acids (A, I, L, V, M, F, Y, or W). Conserved small residues (G, A, or S) are shown in green. Charged residues (D, E, K, R, N, or Q) are shown in blue. The residues corresponding to the proteinase K-resistant amyloid core in P. anserina Het-s are highlighted in purple and underscored. As mentioned earlier, HI was proposed to act as a self/nonself recognition system responsible for limiting the spread of numerous infectious elements in natural populations [15-17]. Coincidentally, HET-s prion behaves as a non-conventional infectious element capable of propagation during anastomosis and sexual reproduction in P. anserina [37]. HET-s can exist in two forms: as a normal protein [18] and as an infectious prion [38], capable of propagating as a self-perpetuating amyloid aggregate [18,36]. Its rather unexpected similarity to LopB implies that members of the family may have another function unrelated to HI. Although its specific role in L. maculans is unknown, LopB- mutants showed impaired ability to form lesions on oilseed rape [35]. LopB contains a predicted signal peptide suggesting that it is secreted and might contribute to the L. maculans pathogenicity by compromising host membranes. The fusions between HET-s/LopB and NACHT domain in N. crassa and A. nidulans suggests that, in other species, proteins containing one of these two domains may physically interact. STAND domain proteins Using P. anserina HET-E as a BLASTp query, we identified several proteins containing NACHT domain in the Aspergilli and other filamentous ascomycetes. Further HMMer searches detected two Aspergillus-specific expansions of the STAND domain [39]: NACHT NTPases and NB-ARC ATPases (Table 1). These NTP-binding proteins are often linked to various protein-binding modules such as WD40, Ankyrin or TPR at the C-terminus and a highly divergent nucleoside phosphorylase (Pfs) domain at their N-terminus (Fig. 3). A different type of domain composition is found in several other STAND NTPases. NB-ARC can fused to a LipA domain found in putative serine esterases and in the SesB protein from Nectria haematococca [40]. Some NACHT NTPases are linked to the HET-s/LopB domain described above. The orthologous relationships within the two STAND domain expansions are difficult to establish since the proteins are highly divergent and exhibit uneven phyletic distribution. They also appear to have undergone multiple domain shuffling events as well as lineage-specific gene loss and expansions during the evolution of the Aspergilli as well as other filamentous fungi. Figure 3 Domain organization of NACHT, HET-s/LopB, and HET domain proteins. Each shape indicates a specific conserved domain. Fused domains that form a single polypeptide chain are connected by a horizontal line. Aspergilli proteins are located in the area with the yellow background. Ank, Ankyrin domain; CARD, caspase recruitment domain; CC, coiled coil domain; HET, HET domain; HET-s/LopB, new domain found in HET-s and LopB proteins; LRR, leucine-rich repeat; NACHT, NACHT domain; NB-ARC, NB-ARC domain; Pfs, nucleoside phosphorylase domain; SesB/LipA, SesB/LipA domain, found in putative serine esterases and in signaling protein SesB from Nectria haematococca; TIR, toll-interleukin receptor domain, TPR, tetratricopeptide repeat; WD, WD40 domain, found in eukaryotic proteins with various functions including adaptor/regulatory modules in signal transduction; typically contains the WD dipeptide at its C-terminus and is 40 residues long. Figure is not drawn to scale. As mentioned earlier, one of N. crassa NACHT domain protein is linked to the HET-s/LopB domain. P. anserina HET-D and HET-E are fused to the HET domain and 11 WD40 repeats, which determine their allelic specificity [41]. HET-E has been shown to genetically (and potentially physically) interact with HET-C2 to trigger incompatibility in P. anserina suggesting that the interaction may activate the ceramide stress response pathway [24]. Similar to the HET domain expansion, the two STAND domain expansions may represent a niche adaptation strategy in filamentous ascomycetes. The multiple fusions involving STAND domains in filamentous fungi may be responsible for the enhancement of their repertoire of signal-transducing interactions, linking preexisting signaling pathways, or integrating multiple signals. Although their specific biological role is unknown, several functional inferences can be made regarding the role of the STAND domain proteins in the Aspergilli. Despite the variability of their domain architecture, the regulatory/signaling function of STAND NTPases seems to be conserved from fungi to man to possibly bacteria. The domain has been implicated in hetero-oligomerization and signal transduction during apoptosis, inflammatory and pathogen responses in animals and plants, and in transcriptional regulation of secondary metabolism in bacteria [42-44]. Similar to their fungal counterparts, animal and plant members of the superfamily tend to be fused to death effector domains at the N-terminus and to repetitive protein binding/regulatory modules at the C-terminus [39]. Other observations suggest that Pfs and STAND domain fusion proteins in the Aspergilli may play a regulatory or signaling role. In plants, the Pfs domain was found in several stress-inducible enzymes [42]. The Pfs domain is also found in bacterial methylthioadenosine/S-adenosylhomocysteine (MTA/SAH) nucleosidases and phosphoribosyltransferases [45]. The bacterial nucleosidases function in methionine salvage pathway and control intracellular levels of MTA/SAH and, in several cases, production of a quorum-sensing signaling molecule [46]. In animals, MTA has been shown to affect many critical responses including regulation of gene expression, proliferation, differentiation and apoptosis [47]. The NB-ARC and LipA as well as NACHT and HET-s/LopB domain fusion proteins may function as bistable switches in signaling cascades. A putative serine esterase domain, LipA, is found in the SesB protein implicated in bistability of developmental signaling cascades in Nectria haematococca [40]. Incidentally, sesB is adjacent to a gene, which encodes a NACHT and Ankyrin domain protein. HET-s is also associated with bistability of the HI reaction resulting from the spread of the infectious prion in P. anserina [38]. Another line of evidence implicating this family in integration of developmental, stress, and nutrient availability signals, comes from expression data. At least five NACHT domain proteins appear to be regulated by LaeA in A. fumigatus (N. Keller, S. Kim and W. Nierman, unpublished), a putative chromatin-dependent regulator of secondary metabolism, virulence and conidiation [48,49]. A few other putative PCD-associated proteins seem to be affected by LaeA in A. fumigatus, including both metacaspases, bZIP transcription factor JlbA, BAX Inhibitor family protein, Cu2+/Zn2+ superoxide dismutase SOD1, histone chaperone ASF1, and AMID-like mitochondrial oxidoreductase, some of which are described below. Mediators of HI-triggered PCD In addition to the putative HI inducers, the Aspergilli possess homologs of HI suppressors from P. anserina and N. crassa (Table 2) [15]. Thus, sequence similarity searches detected orthologs of N. crassa VIB-1 [50] and P. anserina MOD-A, MOD-D and MOD-E [51-53]. VIB-1, a putative regulator of conidiation and HI in N. crassa, is orthologous to Penicillium chrysogenum PhoG and A. nidulans PacG, which were annotated as putative non-repressible acid phosphatases via transformation experiments [50,54,55]. There is no apparent orthologs in yeasts, although, VIB-1 is a distant homolog of the Ndt80p transcription factor from S. cerevisiae [56]. The distribution of MOD-A orthologs also appears to be limited to filamentous ascomycetes. Only A. oryzae, but not other Aspergilli, contains an ortholog of MOD-A, implicated in ascomycete-specific functions such as regulation of growth arrest during HI and female organ formation in P. anserina [51]. On the contrary, MOD-D and MOD-E display a very high level of sequence conservation and have a much wider phyletic distribution with homologs in all fungi and higher eukaryotes. MOD-D, a G protein alpha subunit, is orthologous to GpaB and GanB from A. fumigatus and A. nidulans, respectively [57,58]. Table 2 Putative HI mediators Protein ASP SOR Scer Spom BAS Biological function IDI-1 0 0 0 0 0 Unknown IDI-2 0–1 0–1 0 0 0 Unknown IDI-3 0 0 0 0 0 Unknown IDI-4, JlbA 1 1 0 0 0 Regulation of transcription in response to nutritional stress VIB-1 1–2 1 0 0 0 Regulation of sporulation MOD-D, GpaB, GanB 1 1 1 0 1 Regulation of asexual sporulation, pathogenicity, G-protein signaling MOD-A 0–1 1–3 0 0 0 Regulation of sexual differentiation MOD-E, HSP90 1 1 2 1 2 Regulation of sexual differentiation, life span, protein folding IDI-7, AUT7 1 1 1 1 1 Regulation of sexual differentiation, autophagy IDI-6, Alp2p 1 1 2 2 1 Regulation of sexual differentiation, autophagy ASP, A. fumigatus, N. fischeri, A. oryzae, A. nidulans; SOR, N. crassa, G. zeae, and M. grisea; BAS: C. neoformans and U. maydis; Scer, S. cerevisiae; Spom, S. pombe. The degree of sequence conservation seems to be linked to the relative importance of the biological function and more conserved proteins appear to be functional orthologs. The similarity between VIB-1 and Ndt80p, a transcription factor involved in regulation of sporulation and meiosis in S. cerevisiae [56], suggests that phoG and pacG may encode a transcriptional regulators of the acid phosphatase, rather than the enzyme itself [59]. MOD-D and MOD-E have been shown to function as an alpha subunit of heterotrimeric G protein, and an HSP90 family molecular chaperone, respectively [52,53]. P. anserina MOD-E, in addition to suppressing HI, is involved in regulation of development and the sexual cycle [53,60]. The MOD-E/HSP90 function during the sexual cycle appears to be conserved from fungi to mammals [61]. Mammalian HSP90 family chaperones also mediate the unfolded protein response to endoplasmic reticulum stress through regulation of the secretory pathway, cell cycle and programmed cell death [61,62]. Besides HI suppressors, Aspergilli also possess orthologs of P. anserina idi-2, idi-4, idi-6 and idi-7 genes (Table 2) induced by heterokaryon incompatibility and implicated in autophagy in response to starvation and sporulation and in regulation of sexual differentiation [8,63-65]. IDI-6 and IDI-7 proteins seem to be highly conserved across fungal species; while IDI-2 and IDI-4 are poorly conserved and their distribution is limited to filamentous fungi. No detectable homologs of IDI-1 and IDI-3 are found in Aspergilli and an IDI-2 ortholog is only present in A. oryzae. Autophagic serine protease IDI-6 is orthologous to A. fumigatus Alp2 that has been shown to function in regulation of sporulation as well as pathogenesis in A. fumigatus [66]. Likewise, IDI-4 an ortholog of the A. fumigatus JlbA, a putative bZIP transcription factor induced by amino acid starvation [67]. Many HI-associated genes have wide phyletic distribution and are well-conserved across filamentous fungi. Some are involved in autophagy, suggesting that the incompatibility function might have evolved by recruiting components of the cellular system controlling adaptation to starvation [11]. In addition, cytological alterations during HI in P. anserina are similar to those observed during starvation and treatment with rapamycin, an inhibitor of the TOR (target of rapamycin) signaling pathway that controls autophagic degradation in S. cerevisiae [68]. Yet, many seem to perform unrelated functions such as regulation of development or sexual differentiation, implying that heterokaryon incompatibility may have utilized components of these cell programs as well. Downstream PCD machinery In addition to HI, filamentous fungi appear to possess a wide range of PCD reactions triggered by various death stimuli. In Aspergilli, the apoptotic-like phenotypes and are observed during entry into stationary phase and sporulation and on exposure to certain antifungal agents, peptides, and sphingosines [4,6,7,69,70]. Similarly, in N. crassa, morphological changes during the HI reaction, starvation, and DNA damage response resemble apoptosis [9,10,71]. The apoptotic machinery of filamentous fungi may share some key components with yeast and mammalian systems. First, we looked for homologs of apoptotic proteins in S. cerevisiae, which can undergo PCD in response to nutritional and oxidative stresses, plant antifungal peptides, hydrogen peroxide, or during aging and mating [72-75]. To identify candidate apoptosis-associated proteins in the Aspergilli, BLASTp similarity searches were performed with yeast apoptotic proteins as queries. Homologs of more than 30 yeast proteins were detected including metacaspases and caspase-regulating serine protease HtrA2 (Table 3). Phylogenetic analysis shows that many Aspergillus proteins are in one-on-one orthologous relationships with S. cerevisiae, S. pombe, and basidiomycete proteins (data not shown). Table 3 Putative apoptotic mediators (fungal protein homologs) Protein ASP SOR Scer Spom BAS Biological function/process Aif1p 0 0 1 0 0 Caspase independent apoptosis Stm1p 1 1 1 1 1 Caspase independent apoptosis Cdc6p 1 1 1 1 1 Cell cycle control, DNA replication Asf1p 1 0–1 1 1 1 Cell cycle, chromatin assembly, mating Cdc13p 0 0 1 0 0 Cell cycle control, telomere-binding Cyc1p 1 1 1 1 1 Electron transport, cytochrome c Mre11p 1 1 1 1 1 Maintenance of genome integrity Rad50p 1 1 1 1 1 Maintenance of genome integrity Yca1p 2–3 1–4 1 1 1–2 Metacaspase Sfk1p 1 1 1 1 1 Mitochondrial death pathway Atp4p 1 1 1 1 1 Mitochondrial F1F0 ATP synthase DAP3 1 1 1 1 1 Mitochondrial fragmentation HtrA2 1 1–2 1 2 0 Mitochondrial homeostasis Lsm4p 1 1 1 1 1 mRNA processing Nsr1p 1 1 1 1 1 rRNA processing Hel10p 1 0–1 1 0 0 Apoptosis Uth1p 1 1 4 2 0 Response to stress Sod1p 1 1 1 1 1 Response to stress BI-1 1 1 1 1 1 Response to stress Mras1 1 1 1 1 1 Regulation of development, signaling FadA/GpaA 1 1 1 1 1 Regulation of sexual differentiation, sporulation, G-protein signaling Ste4p/CGB1 1 1 1 1 1 Regulation of sexual differentiation, sporulation, G-protein signaling Ste18p 1 1 1 1 1 Regulation of sexual differentiation Ste20p 1 0–1 1 1 1 Regulation of sexual differentiation Sip3p 1 1 2 1 0 Regulation of sexual differentiation Sst2p, FlbA 1 1–2 1 1 1 Regulation of sexual differentiation Oxa1 1 1 1 1 1 Regulation of life span, respiratory complex assembly RMP1 1 1 1 1 1 Regulation of development and life span, respiratory complex assembly Lag1p 1 1 1 2 1 Sphingolipid-mediated signaling Sar1p, SarA 1 1 1 1 1 Ubiquitin-proteosome system Cdc48p 1 1 1 1 1 Ubiquitin-proteosome system Ubp10p 0 0–1 1 0 0 Ubiquitin-proteosome system Ppa1p 1 1 1 1 0 Vacuolar ATPase subunit ASP, A. fumigatus, N. fischeri, A. oryzae, A. nidulans; SOR, N. crassa, G. zeae, and M. grisea; BAS: C. neoformans and U. maydis; Scer, S. cerevisiae; Spom, S. pombe. BLASTp and HMM searches identified homologs of ~50 human and mouse apoptotic proteins in the Aspergilli [see Additional file 1 ]. Similar to yeasts, filamentous ascomycetes lack the upstream metazoan apoptotic regulators including members of the bax/bcl-2 family and p53, while downstream components of the apoptotic machinery appear to be shared. Interestingly, many Aspergillus proteins are more similar to their human counterparts such as AMID, BIR1, HtrA, and CulA, than to yeast homologs. Thus, the tree topology of the AIF family confirmed that Aspergilli proteins are more closely related to human AMID than to S. cerevisiae Aif1p, which clustered together with plant homologs [see Additional file 2 ]. Moreover, homologs of several key components of the mammalian apoptotic machinery, including AmsH and Poly(ADP-ribose) polymerase (PARP), are not detected in S. cerevisiae (Table 4). Table 4 Putative apoptotic mediators absent in S. cerevisiae Protein/Domain ASP SOR Scer Spom BAS Biological function/process TRAF-3 1 1 0 0 1 Caspase dependent apoptosis Mst3/STK24 1 0–1 0 2 0–1 Caspase dependent apoptosis PARP 1 1 0 0 0 Caspase independent apoptosis AMID 1–2 0–1 0 0 1 Caspase independent apoptosis GRIM-19 1 1 0 0 1–2 Electron transport, NADH ubiquinone oxidoreductase NDUFS1 1 1 0 0 1 Electron transport, NADH ubiquinone oxidoreductase APAF 3–8 0–2 0 0 0 NTP binding, hetero-oligomerization 12R-LO 1 0–1 0 0 0 Peroxidation of arachidonic acid 15-LO 1 0–1 0 0 0 Peroxidation of arachidonic acid PTDSR/PSR 1 1 0 0 1 Recognition of apoptotic cells ASP, A. fumigatus, N. fischeri, A. oryzae, A. nidulans; SOR, N. crassa, G. zeae, and M. grisea; BAS: C. neoformans and U. maydis; Scer, S. cerevisiae; Spom, S. pombe. At least two Aspergilli proteins appear to be functional homologs to their mammalian counterparts. The enhanced PARP and caspase-like activity reported in A. nidulans during sporulation-induced PCD is consistent with the presence of both metacaspase-dependent and -independent apoptotic pathways [70,76]. In addition, the metacaspase-independent apoptosis pathway was shown to operate in A. fumigatus during stationary phase and treatment with fungicidal sphingoid bases and antifungal agents [4,69]. For the rest of the fungal proteins, further experimental characterization is required before any conclusions can be drawn regarding their involvement in PCD. Many yeast and mammalian apoptotic proteins appear to be involved in regulation of cell programs monitoring the cell status such as maintenance of genome integrity, cell cycle control, glycolipid metabolism, and ubiquitin-dependent proteolysis [see Additional file 2] [77]. It is likely that at least non-apoptotic function is conserved in both fungi and metazoa. The results also support the idea that complex development and differentiation in filamentous fungi may require additional PCD pathways or their components not found in unicellular yeasts [78]. Conclusion Our analysis identified more than 100 putative PCD-associated genes in this genus, suggesting a complex uncharacterized regulatory network. Their further characterization may help expand the range of currently available treatments for invasive aspergillosis. The list includes lineage-specific protein families as well as conserved core components of the ancestral PCD machinery shared by all fungi and metazoa. The most divergent group is comprised of putative HI inducers such as STAND, HET-s/LopB and HET domain proteins that show extreme variability in sequence, copy number and domain composition. The STAND NTPases are predicted to interact with different types of effector/signaling components and function as key integrators of stress and nutrient availability signals. On the other end of the spectrum are HI-associated proteins that show broad phyletic distribution and high sequence conservation. They tend to be involved in regulation of development, sexual differentiation, and stress reactions, suggesting that the HI function in filamentous fungi may have evolved by recruiting components from these preexisting pathways [11,52,53]. Further analysis revealed homologs of the yeast PCD proteins in ascomycete and basidiomycete species, further supporting the view that genes encoding the ancestral apoptotic machinery evolved with early eukaryotes [77-80]. Phylogenetic relationships among the putative PCD associated proteins appear to be complex and many Aspergillus proteins show a greater similarity to mammalian than to yeast proteins. In addition, homologs of several mammalian apoptotic proteins including PARP and AMID are found in filamentous fungi, but not in the unicellular yeast such as S. cerevisiae, suggesting that the Aspergilli may serve as an alternative model to study mechanisms of cell death. Methods Sequence similarity To identify human and mouse proteins implicated in PCD, we searched the Gene Ontology (GO) database [81] and the Apoptosis database [82]. Then, sequence similarity searches were performed using PSI-BLAST and Gapped BLAST against selected fungal genomes downloaded from GenBank. The searches were also performed against an in-house database composed of whole-genome sequences of several fungal species from finished and ongoing sequencing projects. The N. fischeri genome sequence has been generated in the course of the genome sequencing projects at TIGR, The Institute for Genomic Research (Rockville, MD). Conserved protein domains were identified using the HMMer package [83]. Phylogenetic Analysis Protein sequences were re-aligned using MUSCLE [84] and columns of low conservation removed manually. The alignments were then used to infer bootstrapped neighbor-joining and maximum-likelihood trees. The neighbor-joining trees were constructed using QuickTree [85] and the maximum-likelihood trees were constructed using the PHYLIP package [86], applying the JTT substitution model with a gamma distribution (alpha = 0.5) of rates over four categories of variable sites. In general, the maximum-likelihood and neighbor-joining trees were congruent. Authors' contributions NDF performed the comparative analysis, interpreted the results and prepared the biological aspects of the manuscripts. JHB performed the phylogenetic analysis and allowed use of his Automated Phylogenetic Inference System (APIS). GDR initiated the project and contributed to the comparative analysis. JRW and WCN contributed to the bioinformatics discussion and planning stage of this project and helped drafting the manuscript. Supplementary Material Additional File 1 Putative apoptotic mediators (mammalian protein homologs). Table indicating numbers of fungal homologs of mammalian apoptotic effectors and mediators. Click here for file Additional File 2 Phylogenetic tree of the AMID family of proteins. Tree reconstruction was performed as described in the Methods section. The numbers indicate percent bootstrap values for internal branches. Click here for file Acknowledgements Funding for the study for each author and for manuscript preparation was provided by the National Institute of Allergy and Infectious Diseases (U01A 148830 and R21 AI052236), USA, the Wellcome Trust, UK and the Fondo de Investigaciones Sanitarias, Spain. We also thank Michael Galperin and Vinita Joardar for insightful discussions and Catherine Ronning, Brian Haas, Paolo Amedeo and Joshua Orvis for computer support. Special thanks go to Louise Glass and Karine Dementhon for valuable comments. We would also like to thank the reviewers for their helpful comments. ==== Refs Latge JP Aspergillus fumigatus and aspergillosis Clin Microbiol Rev 1999 12 310 350 10194462 Wilson DM Mubatanhema W Jurjevic Z Biology and ecology of mycotoxigenic Aspergillus species as related to economic and health concerns Adv Exp Med Biol 2002 504 3 17 11922097 Gupta AK Tomas E New antifungal agents Dermatol Clin 2003 21 565 576 12956208 10.1016/j.clindermatol.2003.08.003 Krishnan S Manavathu EK Chandrasekar PH A comparative study of fungicidal activities of voriconazole and amphotericin B against hyphae of Aspergillus fumigatus J Antimicrob Chemother 2005 15824093 Cheng J Park TS Chio LC Fischl AS Ye XS Induction of apoptosis by sphingoid long-chain bases in Aspergillus nidulans Mol Cell Biol 2003 23 163 177 12482970 10.1128/MCB.23.1.163-177.2003 Krishnan S Manavathu EK Chandrasekar PH A comparative study of fungicidal activities of voriconazole and amphotericin B against hyphae of Aspergillus fumigatus J Antimicrob Chemother 2005 15824093 Mousavi SA Robson GD Oxidative and amphotericin B-mediated cell death in the opportunistic pathogen Aspergillus fumigatus is associated with an apoptotic-like phenotype Microbiology 2004 150 1937 1945 15184579 10.1099/mic.0.26830-0 Pinan-Lucarre B Paoletti M Dementhon K Coulary-Salin B Clave C Autophagy is induced during cell death by incompatibility and is essential for differentiation in the filamentous fungus Podospora anserina Mol Microbiol 2003 47 321 333 12519185 10.1046/j.1365-2958.2003.03208.x Marek SM Wu J Louise Glass N Gilchrist DG Bostock RM Nuclear DNA degradation during heterokaryon incompatibility in Neurospora crassa Fungal Genet Biol 2003 40 126 137 14516765 10.1016/S1087-1845(03)00086-0 Jacobson DJ Beurkens K Klomparens KL Microscopic and Ultrastructural Examination of Vegetative Incompatibility in Partial Diploids Heterozygous at het Loci in Neurospora crassa Fungal Genet Biol 1998 23 45 56 9514694 10.1006/fgbi.1997.1020 Dementhon K Paoletti M Pinan-Lucarre B Loubradou-Bourges N Sabourin M Saupe SJ Clave C Rapamycin mimics the incompatibility reaction in the fungus Podospora anserina Eukaryot Cell 2003 2 238 246 12684373 10.1128/EC.2.2.238-246.2003 Glass NL Rasmussen C Roca MG Read ND Hyphal homing, fusion and mycelial interconnectedness Trends Microbiol 2004 12 135 141 15001190 10.1016/j.tim.2004.01.007 Glass NL Jacobson DJ Shiu PK The genetics of hyphal fusion and vegetative incompatibility in filamentous ascomycete fungi Annu Rev Genet 2000 34 165 186 11092825 10.1146/annurev.genet.34.1.165 Saupe SJ Clave C Begueret J Vegetative incompatibility in filamentous fungi: Podospora and Neurospora provide some clues Curr Opin Microbiol 2000 3 608 612 11121781 10.1016/S1369-5274(00)00148-X Saupe SJ Molecular genetics of heterokaryon incompatibility in filamentous ascomycetes Microbiol Mol Biol Rev 2000 64 489 502 10974123 10.1128/MMBR.64.3.489-502.2000 Saenz GS Stam JG Jacobson DJ Natvig DO Heteroallelism at the het-c locus contributes to sexual dysfunction in outcrossed strains of Neurospora tetrasperma Fungal Genet Biol 2001 34 123 129 11686677 10.1006/fgbi.2001.1294 Biella S Smith ML Aist JR Cortesi P Milgroom MG Programmed cell death correlates with virus transmission in a filamentous fungus Proc Biol Sci 2002 269 2269 2276 12455515 10.1098/rspb.2002.2148 Coustou V Deleu C Saupe S Begueret J The protein product of the het-s heterokaryon incompatibility gene of the fungus Podospora anserina behaves as a prion analog Proc Natl Acad Sci U S A 1997 94 9773 9778 9275200 10.1073/pnas.94.18.9773 Mir-Rashed N Jacobson DJ Dehghany MR Micali OC Smith ML Molecular and functional analyses of incompatibility genes at het-6 in a population of Neurospora crassa Fungal Genet Biol 2000 30 197 205 11035941 10.1006/fgbi.2000.1218 Milgroom MG Cortesi P Analysis of population structure of the chestnut blight fungus based on vegetative incompatibility genotypes Proc Natl Acad Sci U S A 1999 96 10518 10523 10468641 10.1073/pnas.96.18.10518 Saupe SJ Clave C Sabourin M Begueret J Characterization of hch, the Podospora anserina homolog of the het-c heterokaryon incompatibility gene of Neurospora crassa Curr Genet 2000 38 39 47 10953880 10.1007/s002940000130 Wu J Saupe SJ Glass NL Evidence for balancing selection operating at the het-c heterokaryon incompatibility locus in a group of filamentous fungi Proc Natl Acad Sci U S A 1998 95 12398 12403 9770498 10.1073/pnas.95.21.12398 Sarkar S Iyer G Wu J Glass NL Nonself recognition is mediated by HET-C heterocomplex formation during vegetative incompatibility Embo J 2002 21 4841 4850 12234924 10.1093/emboj/cdf479 Saupe S Turcq B Begueret J A gene responsible for vegetative incompatibility in the fungus Podospora anserina encodes a protein with a GTP-binding motif and G beta homologous domain Gene 1995 162 135 139 7557402 10.1016/0378-1119(95)00272-8 Anwar MM Croft JH Dales RB Analysis of heterokaryon incompatibility between heterokaryon-compatibility (h-c) groups R and GL provides evidence that at least eight het loci control somatic incompatibility in Aspergillus nidulans J Gen Microbiol 1993 139 1599 1603 8371120 Saupe S Descamps C Turcq B Begueret J Inactivation of the Podospora anserina vegetative incompatibility locus het-c, whose product resembles a glycolipid transfer protein, drastically impairs ascospore production Proc Natl Acad Sci U S A 1994 91 5927 5931 8016091 Rao CS Lin X Pike HM Molotkovsky JG Brown RE Glycolipid transfer protein mediated transfer of glycosphingolipids between membranes: a model for action based on kinetic and thermodynamic analyses Biochemistry 2004 43 13805 13815 15504043 10.1021/bi0492197 Brodersen P Petersen M Pike HM Olszak B Skov S Odum N Jorgensen LB Brown RE Mundy J Knockout of Arabidopsis accelerated-cell-death11 encoding a sphingosine transfer protein causes activation of programmed cell death and defense Genes Dev 2002 16 490 502 11850411 10.1101/gad.218202 Cheng J Park TS Fischl AS Ye XS Cell cycle progression and cell polarity require sphingolipid biosynthesis in Aspergillus nidulans Mol Cell Biol 2001 21 6198 6209 11509663 10.1128/MCB.21.18.6198-6209.2001 Mattjus P Turcq B Pike HM Molotkovsky JG Brown RE Glycolipid intermembrane transfer is accelerated by HET-C2, a filamentous fungus gene product involved in the cell-cell incompatibility response Biochemistry 2003 42 535 542 12525182 10.1021/bi026896x Saupe SJ Kuldau GA Smith ML Glass NL The product of the het-C heterokaryon incompatibility gene of Neurospora crassa has characteristics of a glycine-rich cell wall protein Genetics 1996 143 1589 1600 8844148 Davies JR Osmani AH De Souza CP Bachewich C Osmani SA Potential link between the NIMA mitotic kinase and nuclear membrane fission during mitotic exit in Aspergillus nidulans Eukaryot Cell 2004 3 1433 1444 15590818 10.1128/EC.3.6.1433-1444.2004 Smith ML Micali OC Hubbard SP Mir-Rashed N Jacobson DJ Glass NL Vegetative incompatibility in the het-6 region of Neurospora crassa is mediated by two linked genes Genetics 2000 155 1095 1104 10880472 Deleu C Clave C Begueret J A single amino acid difference is sufficient to elicit vegetative incompatibility in the fungus Podospora anserina Genetics 1993 135 45 52 8224826 Anantharaman V Aravind L New connections in the prokaryotic toxin-antitoxin network: relationship with the eukaryotic nonsense-mediated RNA decay system Genome Biol 2003 4 R81 14659018 10.1186/gb-2003-4-12-r81 Balguerie A Dos Reis S Ritter C Chaignepain S Coulary-Salin B Forge V Bathany K Lascu I Schmitter JM Riek R Saupe SJ Domain organization and structure-function relationship of the HET-s prion protein of Podospora anserina Embo J 2003 22 2071 2081 12727874 10.1093/emboj/cdg213 Dalstra HJ Swart K Debets AJ Saupe SJ Hoekstra RF Sexual transmission of the [Het-S] prion leads to meiotic drive in Podospora anserina Proc Natl Acad Sci U S A 2003 100 6616 6621 12719532 10.1073/pnas.1030058100 Balguerie A Dos Reis S Coulary-Salin B Chaignepain S Sabourin M Schmitter JM Saupe SJ The sequences appended to the amyloid core region of the HET-s prion protein determine higher-order aggregate organization in vivo J Cell Sci 2004 117 2599 2610 15159455 10.1242/jcs.01116 Leipe DD Koonin EV Aravind L STAND, a class of P-loop NTPases including animal and plant regulators of programmed cell death: multiple, complex domain architectures, unusual phyletic patterns, and evolution by horizontal gene transfer J Mol Biol 2004 343 1 28 15381417 10.1016/j.jmb.2004.08.023 Graziani S Silar P Daboussi MJ Bistability and hysteresis of the 'Secteur' differentiation are controlled by a two-gene locus in Nectria haematococca BMC Biol 2004 2 18 15312233 10.1186/1741-7007-2-18 Espagne E Balhadere P Penin ML Barreau C Turcq B HET-E and HET-D belong to a new subfamily of WD40 proteins involved in vegetative incompatibility specificity in the fungus Podospora anserina Genetics 2002 161 71 81 12019224 Girardin SE Sansonetti PJ Philpott DJ Intracellular vs extracellular recognition of pathogens--common concepts in mammals and flies Trends Microbiol 2002 10 193 199 11912027 10.1016/S0966-842X(02)02334-X Belkhadir Y Subramaniam R Dangl JL Plant disease resistance protein signaling: NBS-LRR proteins and their partners Curr Opin Plant Biol 2004 7 391 399 15231261 10.1016/j.pbi.2004.05.009 van der Biezen EA Jones JD The NB-ARC domain: a novel signalling motif shared by plant resistance gene products and regulators of cell death in animals Curr Biol 1998 8 R226 7 9545207 10.1016/S0960-9822(98)70145-9 Mushegian AR Koonin EV Unexpected sequence similarity between nucleosidases and phosphoribosyltransferases of different specificity Protein Sci 1994 3 1081 1088 7920254 Beeston AL Surette MG pfs-dependent regulation of autoinducer 2 production in Salmonella enterica serovar Typhimurium J Bacteriol 2002 184 3450 3456 12057938 10.1128/JB.184.13.3450-3456.2002 Avila MA Garcia-Trevijano ER Lu SC Corrales FJ Mato JM Methylthioadenosine Int J Biochem Cell Biol 2004 36 2125 2130 15313459 10.1016/j.biocel.2003.11.016 Bok JW Keller NP LaeA, a regulator of secondary metabolism in Aspergillus spp Eukaryot Cell 2004 3 527 535 15075281 10.1128/EC.3.2.527-535.2004 Bok JW Balajee SA Marr KA Andes D Nielsen KF Frisvad JC Keller NP LaeA, a Regulator of Morphogenetic Fungal Virulence Factors Eukaryot Cell 2005 4 1574 1582 16151250 10.1128/EC.4.9.1574-1582.2005 Xiang Q Glass NL The control of mating type heterokaryon incompatibility by vib-1, a locus involved in het-c heterokaryon incompatibility in Neurospora crassa Fungal Genet Biol 2004 41 1063 1076 15531211 10.1016/j.fgb.2004.07.006 Barreau C Iskandar M Loubradou G Levallois V Begueret J The mod-A suppressor of nonallelic heterokaryon incompatibility in Podospora anserina encodes a proline-rich polypeptide involved in female organ formation Genetics 1998 149 915 926 9611202 Loubradou G Begueret J Turcq B MOD-D, a Galpha subunit of the fungus Podospora anserina, is involved in both regulation of development and vegetative incompatibility Genetics 1999 152 519 528 10353896 Loubradou G Begueret J Turcq B A mutation in an HSP90 gene affects the sexual cycle and suppresses vegetative incompatibility in the fungus Podospora anserina Genetics 1997 147 581 588 9335595 Marx F Haas H Hofer S Stoffler G Redl B Sequence and structure of Penicillium chrysogenum phoG, homologous to an acid phosphatase-encoding gene of Aspergillus nidulans Gene 1995 160 137 138 7628710 10.1016/0378-1119(95)00245-2 MacRae WD Buxton FP Sibley S Garven S Gwynne DI Davies RW Arst HNJ A phosphate-repressible acid phosphatase gene from Aspergillus niger: its cloning, sequencing and transcriptional analysis Gene 1988 71 339 348 3224828 10.1016/0378-1119(88)90051-0 Hepworth SR Friesen H Segall J NDT80 and the meiotic recombination checkpoint regulate expression of middle sporulation-specific genes in Saccharomyces cerevisiae Mol Cell Biol 1998 18 5750 5761 9742092 Han KH Seo JA Yu JH Regulators of G-protein signalling in Aspergillus nidulans: RgsA downregulates stress response and stimulates asexual sporulation through attenuation of GanB (Galpha) signalling Mol Microbiol 2004 53 529 540 15228532 10.1111/j.1365-2958.2004.04163.x Liebmann B Muller M Braun A Brakhage AA The cyclic AMP-dependent protein kinase a network regulates development and virulence in Aspergillus fumigatus Infect Immun 2004 72 5193 5203 15322014 10.1128/IAI.72.9.5193-5203.2004 Xiang Q Glass NL Identification of vib-1, a locus involved in vegetative incompatibility mediated by het-c in Neurospora crassa Genetics 2002 162 89 101 12242225 Loubradou G Begueret J Turcq B An additional copy of the adenylate cyclase-encoding gene relieves developmental defects produced by a mutation in a vegetative incompatibility-controlling gene in Podospora anserina Gene 1996 170 119 123 8621071 10.1016/0378-1119(95)00847-0 Szent-Gyorgyi C A bipartite operator interacts with a heat shock element to mediate early meiotic induction of Saccharomyces cerevisiae HSP82 Mol Cell Biol 1995 15 6754 6769 8524241 Patil C Walter P Intracellular signaling from the endoplasmic reticulum to the nucleus: the unfolded protein response in yeast and mammals Curr Opin Cell Biol 2001 13 349 355 11343907 10.1016/S0955-0674(00)00219-2 Dementhon K Saupe SJ Clave C Characterization of IDI-4, a bZIP transcription factor inducing autophagy and cell death in the fungus Podospora anserina Mol Microbiol 2004 53 1625 1640 15341644 10.1111/j.1365-2958.2004.04235.x Dementhon K Saupe SJ DNA-binding specificity of the IDI-4 basic leucine zipper factor of Podospora anserina defined by systematic evolution of ligands by exponential enrichment (SELEX) Eukaryot Cell 2005 4 476 483 15701809 10.1128/EC.4.2.476-483.2005 Bourges N Groppi A Barreau C Clave C Begueret J Regulation of gene expression during the vegetative incompatibility reaction in Podospora anserina. Characterization of three induced genes Genetics 1998 150 633 641 9755195 Reichard U Cole GT Hill TW Ruchel R Monod M Molecular characterization and influence on fungal development of ALP2, a novel serine proteinase from Aspergillus fumigatus Int J Med Microbiol 2000 290 549 558 11100830 Strittmatter AW Irniger S Braus GH Induction of jlbA mRNA synthesis for a putative bZIP protein of Aspergillus nidulans by amino acid starvation Curr Genet 2001 39 327 334 11525406 10.1007/s002940100215 Kamada Y Sekito T Ohsumi Y Autophagy in yeast: a TOR-mediated response to nutrient starvation Curr Top Microbiol Immunol 2004 279 73 84 14560952 Mousavi SA Robson GD Entry into the stationary phase is associated with a rapid loss of viability and an apoptotic-like phenotype in the opportunistic pathogen Aspergillus fumigatus Fungal Genet Biol 2003 39 221 229 12892635 10.1016/S1087-1845(03)00047-1 Thrane C Kaufmann U Stummann BM Olsson S Activation of caspase-like activity and poly (ADP-ribose) polymerase degradation during sporulation in Aspergillus nidulans Fungal Genet Biol 2004 41 361 368 14761796 10.1016/j.fgb.2003.11.003 Goldman GH McGuire SL Harris SD The DNA damage response in filamentous fungi Fungal Genet Biol 2002 35 183 195 11929209 10.1006/fgbi.2002.1344 Burhans WC Weinberger M Marchetti MA Ramachandran L D'Urso G Huberman JA Apoptosis-like yeast cell death in response to DNA damage and replication defects Mutat Res 2003 532 227 243 14643439 Eisler H Frohlich KU Heidenreich E Starvation for an essential amino acid induces apoptosis and oxidative stress in yeast Exp Cell Res 2004 300 345 353 15474999 10.1016/j.yexcr.2004.07.025 Ligr M Velten I Frohlich E Madeo F Ledig M Frohlich KU Wolf DH Hilt W The proteasomal substrate Stm1 participates in apoptosis-like cell death in yeast Mol Biol Cell 2001 12 2422 2432 11514626 Madeo F Herker E Wissing S Jungwirth H Eisenberg T Frohlich KU Apoptosis in yeast Curr Opin Microbiol 2004 7 655 660 15556039 10.1016/j.mib.2004.10.012 Hong SJ Dawson TM Dawson VL Nuclear and mitochondrial conversations in cell death: PARP-1 and AIF signaling Trends Pharmacol Sci 2004 25 259 264 15120492 10.1016/j.tips.2004.03.005 Aravind L Dixit VM Koonin EV Apoptotic molecular machinery: vastly increased complexity in vertebrates revealed by genome comparisons Science 2001 291 1279 1284 11181990 10.1126/science.291.5507.1279 Koonin EV Aravind L Origin and evolution of eukaryotic apoptosis: the bacterial connection Cell Death Differ 2002 9 394 404 11965492 10.1038/sj.cdd.4400991 Koonin EV Galperin MY Comparative Genomics and New Evolutionary Biology Sequence - Evolution - Function 2002 First Boston, Dordrecht, London, Kluwer Academic Press 227 294 Aravind L Dixit VM Koonin EV The domains of death: evolution of the apoptosis machinery Trends Biochem Sci 1999 24 47 53 10098397 10.1016/S0968-0004(98)01341-3 Gene Ontology [http://www.geneontology.org] Apoptosis database [http://www.apoptosis-db.org] HMMer package [http://hmmer.wustl.edu] Edgar RC MUSCLE: a multiple sequence alignment method with reduced time and space complexity BMC Bioinformatics 2004 5 113 15318951 10.1186/1471-2105-5-113 Howe K Bateman A Durbin R QuickTree: building huge Neighbour-Joining trees of protein sequences Bioinformatics 2002 18 1546 1547 12424131 10.1093/bioinformatics/18.11.1546 Retief JD Phylogenetic analysis using PHYLIP Methods Mol Biol 2000 132 243 258 10547839	16336669	PMC1325252	CC BY	2021-01-04 16:32:47	no	BMC Genomics. 2005 Dec 8; 6:177	utf-8	BMC Genomics	2,005	10.1186/1471-2164-6-177	oa_comm
==== Front BMC CancerBMC Cancer1471-2407BioMed Central London 1471-2407-5-1581633667510.1186/1471-2407-5-158Research ArticleCaffeic acid phenethyl ester decreases acute pneumonitis after irradiation in vitro and in vivo Chen Miao-Fen [email protected] Peter C [email protected] Paul-Yang [email protected] Cheng-Ta [email protected] Shuen-Kuei [email protected] Wen-Cheng [email protected] Department of Radiation Oncology, Chang Gung Memorial Hospital, Chia-Yi, Taiwan2 Graduate Institute of Clinical Medical Science, Chang Gung University, Toyuan, Taiwan3 Department of Radiation Oncology, Biochemistry and Biophysics, University of Rochester School of Medicine and Dentistry, Rochester, New York, USA4 Department of Pathology, Chang Gung Memorial Hospital, Chia-Yi, Taiwan5 Department of Internal medicine, Chang Gung Memorial Hospital, Chia-Yi, Taiwan2005 9 12 2005 5 158 158 8 9 2005 9 12 2005 Copyright © 2005 Chen 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 Lung cancer is relatively resistant to radiation treatment and radiation pneumonitis is a major obstacle to increasing the radiation dose. We previously showed that Caffeic acid phenethyl ester (CAPE) induces apoptosis and increases radiosensitivity in lung cancer. To determine whether CAPE, an antioxidant and an inhibitor of NF-kappa B, could be a useful adjuvant agent for lung cancer treatment, we examine the effects of CAPE on irradiated normal lung tissue in this study. Methods We compared the effects of CAPE on cytotoxicity and intracellular oxidative stress in normal lung fibroblast and a lung cancer cell line. For in vivo analysis, whole thorax radiation (single dose 10 Gy and 20 Gy) was delivered to BALB/c male mice with or without CAPE pretreatment. NF- kappaB activation and the expression levels of acute inflammatory cytokines were evaluated in mice after irradiation. Results The in vitro studies showed that CAPE cause no significant cytotoxicity in normal lung as compared to lung cancer cells. This is probably due to the differential effect on the expression of NF-kappa B between normal and malignant lung cells. The results from in vivo study showed that CAPE treatment decreased the expression of inflammatory cytokines including IL-1 alpha and beta, IL-6, TNF-alpha and TGF- beta, after irradiation. Moreover, histological and immunochemical data revealed that CAPE decreased radiation- induced interstitial pneumonitis and TGF-beta expression. Conclusion This study suggests that CAPE decreases the cascade of inflammatory responses induced by thoracic irradiation without causing toxicity in normal lung tissue. This provides a rationale for combining CAPE and thoracic radiotherapy for lung cancer treatment in further clinical studies. ==== Body Background Lung cancer is the leading cause of cancer death worldwide. Radiotherapy is an important modality of cancer treatment. Because radiation pneumonitis is a major obstacle to increasing the radiation dose, it is important to determine how the incidence of radiation- induced complication might be decreased and how the dose that normal lung can tolerate might be increased. Caffeic acid phenethyl ester (CAPE) is a phenolic antioxidant, an active anti-inflammatory component of propolis [1-3]. Several studies and our previous study have shown that the compound elicits several interesting biological functions including inducing apoptosis in various tumor cell types and anti-inflammatory properties [4-6]. Since reactive oxygen species (ROS) is the major mediators for radiation induced damage [7], a treatment combining radiation with an antioxidant might provide a strategy for preventing radiation injury to normal tissues [8]. Various investigators have demonstrated that radiation-induced proinflammatory cytokines contributed significant complications associated with radiotherapy [9,10]. Early manipulation of inflammatory responses could be useful in modifying subsequent late effect [14]. CAPE is an active anti-inflammatory compound [11,12], and a specific inhibitor of the transcription factor nuclear factor-κB (NF-κB) [13,14]. It might play a role in protecting normal tissue against damage from radiation treatment. However, the actual effects of CAPE on irradiated normal lung and the underlying mechanisms of protection are still unclear. An initial goal of our study was to assess the capacity of CAPE to decrease radiation pneumonitis. We utilized two cell lines, normal lung fibroblast (WI-38) and lung cancer cells (A549), to compare the effects of CAPE on normal and malignant lung cells in the presence and absence of radiation treatments in vitro. We were specifically interested in evaluating the effects of CAPE on the intracellular ROS and NF-κB activation in these cells. Another important goal of this study was to evaluate the efficiency of CAPE in an animal model of radiation- induced pneumonitis. Methods Human lung cancer A549 cell and human normal lung fibroblast WI-38 cells were obtained from ATCC. They were routinely cultured in complete MEM medium and maintained in a 37°C incubator with 5% CO2 and 95% air. Cell growth curve analysis with radiation Exponentially growing cells were treated with or without CAPE 6 μg/ml for 1 h prior irradiation. Cells were irradiated (9 Gy in a single fraction) with a 6 MeV electron beam generated by a linear accelerator at a dose rate of 300 cGy/min. After irradiation, the cells were allowed to grow in the incubator. Viable cells were counted on day 2, 4, 6 and 8 after treatment. Intracellular H2O2 and glutathione (GSH) analysis Intracellular H2O2 was assayed with a fluoresence dye, DCFH-DA, using a FACS caliber flow cytometery, as described previously [15]. To determine the intracellular GSH, we performed a colorimetric assay using ApoAlert glutathione detection kit (Clontech, CA). Cells were lysed and monochlorobimane was added to the lysate at 37°C for at least 15 min. The fluorescence intensity was measured in a plate reader at 395 nm. Electrophoretic mobility gel shift assays to analyze the binding activity of NF-κB The cells were treated with nuclear extraction reagent 4 hour after 9 Gy irradiation (Pierce, Rockford, IL). The nuclear proteins from BALB/c mice lung tissues were collected 12 hours after a 20 Gy irradiation. Four murine lung tissues from each group were checked. The protein content was measured by the Bradford method. The DNA oligonucleotides used for NF-κB binding were 5'-TTGTTACAAGGGACT TTCCG TGGGGACTTTCC AGGGAGGC GTGG-3' for human and 5'- AGTTGAGGGACTT TCCCAGGC-3' for mouse. Nuclear extracts were incubated with the biotin labeled DNA probe for 20 min at room temperature. The DNA- protein complex was separated from free oligonucleotide on a 5% polyacrylamide gel. After electrophoresis, the DNA-protein complex was transferred to a nylon membrane and cross-linked with UV. The membrane was incubated with streptavidin- horseradish peroxidase-conjugate and detected by ECL (Pierce, Rockford, IL). Mice, radiation and CAPE treatment Male BABL/c mice between 6 and 8 weeks old were purchased from the Animal Center of the National Science Council, Taipei, Taiwan. The protocol of animal experimentation was approved by the Chang Gung Memorial Hospital Experimental Animal Committee. For lung irradiation, anesthetized mice were restrained in modified Perspex tubes. The whole thorax was irradiated by 6 MV X-ray from a linear accelerator and a 1.5 cm bolus on the surface. Control mice were subjected to sham-irradiation. The mice were divided into four groups: control, CAPE alone, irradiation, irradiation with CAPE treatment. To analyze NF-κB activation and the expression of inflammatory cytokines, 24 mice were irradiated with 20 Gy and sacrificed at the indicated times. For histological examination, 12 mice received 10 Gy irradiation. The mice were injected intraperitoneally with CAPE (10 mg/kg, solubilized in saline containing 20% Tween 20) 30 min before irradiation and once a day for 10 days after irradiation. The expressions of inflammatory cytokines analyzed by RNase protection assay (RPA) and real time RT-PCR The mice were killed 6 hours and 24 hours after 20 Gy irradiation and RNA was extracted. Four mice from each group were sacrified at the indicated times and the lungs were dissected. Total cellular RNA was isolated using Trizol (Gibco, Grand Island, NY) reagents for RPA and real time RT PCR. Two probe sets-mck2b and mck3b (Pharmingen, San Diego, CA) were used for RPA. The biotin-labeled probes (Ambion, Austin, TX) were hybridized to the target mRNA. The samples were mixed with loading buffer and separated on a 9% sequencing gel by electrophoresis. The intensity of each gene was normalized against with L32 and GAPDH expression. Two μg RNA was subjected to reverse transcription using the superscript II kit (Invitrogen, Carlsbad, CA) with random primer to obtain the first cDNA strand. Primers for IL-1 α/β, IL-6, TNF-α, TGF-β were used for PCR analysis (Applied Biosystems, Foster, CA). To allow for loading differences, a GAPDH primer was used as control. Optimized PCR was performed on an iCycler iQ multicolor real time PCR detection system. Significant PCR fluorescent signals were normalized to a PCR fluorescent signal obtained from the mean value for sham-irradiated control mice. Histologic and immunochemical staining analysis for pulmonary inflammation Treated and control mice were sacrificed by cervical dislocation at 3 and 12 weeks after 10 Gy irradiation. Three mice for each group were used. The whole lungs were perfused via the trachea before they were removed; they were fixed in formalin after removal. For histologic analysis, the lobes were fixed in 10% buffered formalin; paraffin-embedded and sectioned at an average thickness of 5 μm. The mounted sections were subjected to H&E and immunochemical staining. They were incubated overnight with goat anti- mouse TGF-β1antibody (Santa Cruz, CA) After three wash with PBS, the sections were incubated with biotinylated anti-goat IgG followed by peroxidase-avidin staining, washed again in PBS, and treated with 3-amino-9 ethylcarbazole solution as a chromogen. The specimens were counterstaining with hematoxylin. Omission of the primary antibody was used as a negative control, and murine bowel tissue was used as a positive control for the presence of TGF-β1. Results CAPE causes no cytotoxicity or radiosensitization in normal lung cells In our previous report, we demonstrated that 6 μg/ml CAPE caused significant cytotoxicity and increased apoptosis in lung cancer cells [5]. However, the percentage apoptosis showed no obvious increase in WI-38 after CAPE treatment (data not shown). To compare the effects of CAPE in lung cancer and normal lung cells, we treated the cells with 6 μg/ml CAPE for 1 hour prior to radiation and evaluated the cytotoxicity and radiosensitizing effects. As shown in Figure 1, CAPE had no cytotoxicity and radiosensitization effects on WI-38 normal lung cells, in contrast to A549 lung cancer cells. The CAPE- induced decrease of intracellular GSH described in our previous study might have contributed the radiosensitization effect on A549. To explore the possible mechanisms responsible for differential radiosensitization in A549 and WI-38, the intracellular H2O2 and GSH were measured after CAPE treatment. The antioxidant effect of CAPE decreased the intracellular H2O2 within 1 h in WI-38 cells (Fig. 2A), as found in A549 cells in our previous study. However, CAPE did not decrease the intracellular GSH level in WI-38 cells as it does in A549 lung cancer cells (Fig. 2B). This differential effect could explain, at least in part, why CAPE- induced radiosensitization is found in A549 but not WI-38. CAPE attenuates NF-κB expressiondifferentially in normal lung and lung cancer cells CAPE is a specific inhibitor of NF- κB that works by inhibiting the interaction of the transcription factor with DNA. The putative target genes of NF- κB are mainly involved in immune and inflammatory responses. To demonstrate the activation of NF- κB after CAPE treatment, gel shift experiments were conducted in both cell lines (A549 lung cancer cell line and WI-38 normal lung cell) and in mouse lung tissue following various treatments. As shown in Figure 3A and 3B, the nuclear binding of NF-κB is higher in lung cancer cells than in normal lung cells, and 6 μg/ml CAPE treatment significantly decreased binding in lung cancer cells as compared with normal lung cells. Irradiation increased the nuclear binding of NF-κB in both lung cancer cells and normal lung cells. Pre-treatment with 6 μg/ml CAPE for 1 h decreased the augmentation of NF-κB binding activity by irradiation in both cell types. Figure 3C demonstrates the similarity between the nuclear proteins obtained from murine lung tissues and WI-38. There was not change in NF-κB binding in unirradiated lung tissues after CAPE treatment. Irradiation with 20 Gy promoted NF-κB binding, and treatment with CAPE attenuated this effect 12 h after irradiation. The results clearly showed that irradiation activates NF- κB binding and the activation is attenuated by CAPE. Effect of CAPE on the radiation- induced expression of proinflammatory cytokines as revealed by RPA and real- time PCR We used a non-radioactive ribonuclease protection assay to screen the expressions of various cytokines mRNAs. The mRNAs detected were for cytokines involved in acute inflammation and the fibrosis of pneumonitis: TNF-α/β, IL-6, IFN-β/γ, IL-6, IL-10, IL-1α/β, IL12 and MIF. These cytokine mRNAs were barely detectable in control but the levels were elevated after 20 Gy pulmonary irradiation. After combined radiation and CAPE treatment, there was a decrease in the trend towards overexpression of TNF-α, IL-6, IL-1α/β and TGF-β mRNA (data not show). Furthermore, we quantified the expression of these cytokines after radiation and CAPE treatment by real-time RT-PCR. Low expression levels were noted in unirradiated BALB/c mice and there were no obvious changes after 10 mg/kg CAPE treatment. Irradiation (20 Gy) induced a significant increase in TNF-α, IL-6, IL-1α/β and TGF-β mRNA 6 hours and 24 hours after irradiation (Fig. 4A and 4B). The attenuating effect of CAPE measured by RPA was quantified by real-time PCR. At 6 hours, CAPE treatment reduced the increase in the levels of the TNF-α, IL-1α and TGF-β mRNA by half (p < 0.05 verse irradiated untreated mice), and IL-6 and IL-1β mRNA by one quarter. The inhibitory effect of CAPE on IL-6 and IL-1β mRNA levels were more significant 24 h after irradiation (p < 0.05 verse irradiated untreated mice). CAPE attenuates the induction of pulmonary inflammation by irradiation No lesions were observed in non-irradiated lung from control mice. Microscopic examination of the lungs 3 weeks after 10 Gy irradiation revealed an increase in acute inflammatory infiltrate in the interstitium. After CAPE treatment, the degree of interstitial pneumonitis 3 weeks after 10 Gy irradiation was less pronounced (Fig 5A–C). Similar changes persisted 12 weeks after irradiation (Fig 5D–F). Immunochemical analysis showed that unirradiated lung tissue exhibited very low TGF-β1 immunoreactivity in the parenchyma and muscularis propria. By 3 weeks after 10 Gy irradiation, positive TGF-β1 staining had increased. CAPE treatment attenuated the radiation- induced increase. (Fig 6) Discussion In this study, CAPE cause no significant cytotoxicity and radiosensitization in normal lung cell, in contrast to that noted in lung cancer cells. CAPE is a potent and specific inhibitor of activation of nuclear transcription factor NF-κB [12]. NF-κB activation could induce compensated mechanisms and decrease apoptosis. The inhibition of NF-κB by CAPE is a possible mechanism for tumor cell cytotoxicity [16-18]. We found significant higher expression of NF-κB binding in lung cancer cells than normal lung cells. CAPE significantly decreased the NF-κB binding activity in lung cancer cells. In contrast, it caused no significant change in the expression of NF-κB binding in normal lung. The differential downregulation of NF-κB by CAPE might explain why CAPE caused significant cytotoxicity and growth inhibition in lung cancer cells but not normal lung cells. Furthermore, we found that CAPE had a radiosensitizating effect on lung cancer cells but not on normal lung cells. The mechanism underpinning this radiosensitization may be related to intracellular ROS, GSH and NF-κB [19-21] and it needs further investigation. Intracellular GSH might play a role because we noted that GSH levels decrease after CAPE treatment in tumor cells but not in normal lung cells. GSH decreases radiation-induced damage through its function as a free radical scavenger. A high concentration of intracellular thiol is an important way to resist cytotoxic and radiation damage in cancer cells. The depletion of GSH in lung cancer cells is consisting with the differential radiosensitization between tumor cells and normal lung cells. The lung is the major dose-limiting organ for radiotherapy in thoracic region and radiation pneumonitis is a serious complication of lung cancer treatment by radiotherapy. Acute pneumonitis is characterized by edema, infiltration of inflammatory cells and thickening of the alveolar septa. Late radiation- induced lung damage is characterized by pulmonary fibrosis, which is usually proceeded by fibrosing alveolitis [22]. Franko et al [23] pointed out the pathomorphological effects in irradiated lung in relation to the lung function; fibrosis was first observed 8 weeks post-irradiation at 10.3 Gy and the number of fibrotic lesions had increased 10- fold by 14 weeks. The response induced by radiation in vivo is associated with increased expression and activity of inflammatory cytokines. NF-κB is believed to play a pivotal role in the induction of cytokine expression in inflammatory response [24,25]. Haase et al [26] reported that DNA binding by NF-κB is activated for 6 months after irradiation of the rat lung. This might play a role in sustaining chronic inflammation and hyerproliferation of mesenchymal cells after radiation. NF-κB plays a key role in the induction of these cytokines in vivo. The efficacy of CAPE in inhibiting NF-κB activation and proinflammatory production has been demonstrated [13,14,27,28]. Fitzpatrick et al [14] reported that CAPE significantly attenuated bacterial peptidoglycan polysaccharise-induced colitis and reduced the inflammatory cytokine level. Linard et al [27] demonstrated that CAPE treatment reduced the NF-κB activation and attenuated the increase in IL-6 and IL-1β expression in intestine. However, there were few reports about the relationship between CAPE and radiation pneumontits. The present study has demonstrated that CAPE treatment inhibits NF-κB activation and reduces the overexpression of genes involved in the acute inflammatory response, including IL-1α/β, TNF-α and IL-6 after irradiation in the mouse lungs. We found that CAPE treatment inhibited NF-κB activation and reduced the overexpression of genes involved in the acute inflammatory response including IL-6 and IL-1β, consistent with the results of Linard's series. In addition, the finding of the inhibition of TNF-α and IL-1 by the selective blockade of NF-κB pathway is similar to observations reported in alveolar epithelial cells [25,29]. Because these elevated cytokines are closely related to radiation induced pneumonitis [30,31], this could explain why CAPE treatment decreased irradiation-induced interstitial pneumonitis 3 and 12 weeks post-irradiation. Moreover, the results showed CAPE treatment is effective in reducing the expression of TGF-β1 after irradiation. TGF-β is a potent chemoattractant for fibroblasts and triggers the expression of extracellular matrix components in pulmonary fibrosis. The predominant localization of TGF-β in areas of inflammatory cell infiltrates and fibrosis suggests involvement of this cytokine in the pathogenesis of radiation- induced pulmonary fibrosis [32,33]. The mechanisms by which CAPE decreases TGF-β are unclear. It is probably through the NF-κB dependent pathway [34,35] or a process indirectly related to NF-κB [36,37]. In view of theses studies, we propose that CAPE treatment reduces radiation- induced pulmonary inflammation and fibrosis after a longer follow up time. In summary, we have shown CAPE plays an important role in decreasing radiation pneumonitis by inflammatory cytokines, at least in part, without causing significant cytotoxicity. Based on this observation, CAPE is a promising adjuvant agent in the radiation treatment of lung cancer. Competing interests The author(s) declare that they have no competing interests. Authors' contributions MFC performed the study and drafted the manuscript. PCK conceived part of the study and performed the statistical analysis. PYL helped in histology and IHC staining. CTY and SKL conceived of the study and participated in its design and coordination. WCC conceived of the study, participated in its design and coordination and assisted in editing of manuscript. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements The work was supported by NSC 93-2314-B-182A-125 from the National Science Council, Taiwan and CMRP 630007 from Chang Gung Memorial Hospital Figures and Tables Figure 1 The growth curves of A549 lung cancer cells and WI-38 cells in the presence of CAPE and irradiation. Exponentially growing cells were treated with or without 6 μg/ml CAPE for 1 h prior to irradiation. Viable cells were counted on day 2, 4, 6 and 8 after treatment. The growth curves were obtained by plotting the number of viable cells as a function of time in culture for 8 days. (a) Growth curves of A549 cancer cells; (b) Growth curves of WI-38 cells. CAPE significantly reduces in vitro cell growth of A549 with or without irradiation, while no inhibition of growth was detected in WI-38 cells. Three independent experiments were performed for these curves. Y axis represents the relative cell count, normalized to the cell numbers on Day 0. * P < 0.05, day 4, 6 and 8, CAPE treated cells verse CAPE- untreated cells. Figure 2 Intracellular peroxide and glutathione content of WI-38 cells. Determination of peroxides and GSH in intact cells (WI-38 control cells or cell treated with CAPE) was accomplished by fluorescence measurement as described in the Materials and Methods section. (a) Intracellular H2O2 levels. (b) intracellular GSH levels. Data are presented as the mean ± SE of three separate experiments. Y axis represents the relative fluorescence, normalized to the fluorescence value of control cells. * P < 0.05, CAPE treated cells verse CAPE- untreated cells. Figure 3 Inhibition of irradiation- induced activation of DNA binding of NF-κB by CAPE in vitro and in vivo. Nuclear extracts were isolated and mobility gel shift assays were performed as described in the Material and Methods Section. Representative figures are ahown for (a) A549 cancer cells; (b) WI-38 cells 4 hours after 9 Gy irradiation; and (c) normal mice lung 12 hours after 20 Gy irradiation. These demonstrate the binding activity of nuclear NF-κB is higher in A549 cancer cells. CAPE inhibited the binding activity of NF-κB in A549 cancer cells with or without irradiation and in normal lung cells after irradiation. Figure 4 Effects of CAPE treatment on radiation-induced increases in inflammatory cytokine mRNA levels in vivo. The TGF-β, TNF-α, IL-1α/βand IL-6 mRNA levels (a) 6 hours (b) 24 hours after20 Gy irradiation. The results are expressed as RNA ratio. The RNA ratio was determined from the mRNA level at indicated time after treatment to control for a specific gene. Data are the mean ± SE. * P < 0.05, CAPE-treated irradiated mice versus CAPE-untreated irradiated mice. Figure 5 Histologic analyses with H&E staining on irradiated murine lung tissue. Three mice from each group were checked. Representative slides are shown for (a) unirradiated control mice at 3 weeks after sham-irradiation; (b) untreated mice at 3 weeks after 10 Gy irradiation; (c) CAPE-treated mice at 3 week after irradiation 10 Gy irradiation; (d) unirradiated control mice at 12 weeks after sham-irradiation; (e) untreated mice at 12 weeks after 10 Gy irradiation; (f) CAPE-treated mice at 12 week after irradiation 10 Gy irradiation. These demonstrate that interstitial pneumonia with increased acute inflammatory infiltrate in the interstitium was detected at 3 weeks and 12 weeks after irradiation. CAPE treatment attenuated the extent of inflammation. Magnification × 200 Figure 6 Immunohistochemical staining with TGF-β antibody on murine lung tissues. Three mice from each group were checked. Representative slides are shown for (a) unirradiated control mice at 3 weeks after sham-irradiation; (b) untreated mice at 3 weeks after 10 Gy irradiation; (c) CAPE-treated mice at 3 week after irradiation 10 Gy irradiation. These demonstrated that CAPE treatment attenuated the increased TGF-β immunoreactivity after irradiation. Magnification × 250. ==== Refs Chiao C Carothers AM Grunberger D Solomon G Preston GA Barrett JC Apoptosis and altered redox state induced by caffeic acid phenethyl ester (CAPE) in transformed rat fibroblast cells Cancer Res 1995 55 3576 3583 7543016 Sud'ina GF Mirzoeva OK Pushkareva GA Korshunova GA Sumbatyan NV Varfolomeev SD Caffeic acid phenethyl ester as a lipoxygenase inhibitor with antioxidant properties FEBS Lett 1993 329 21 4 7689063 10.1016/0014-5793(93)80184-V Bhimani RS Troll W Grunberger D Frenkel K Inhibition of oxidative stress in Hela cells by chemopreventive agents Cancer Res 1993 53 4528 33 7691399 Chen YJ Shiao MS Wang SY The antioxidant caffeic acid phenethyl ester induces apoptosis associated with selective scavenging of hydrogen peroxide in human leukeic HL-60 cells Anti-Cancer Drugs 2001 12 143 149 11261888 10.1097/00001813-200102000-00008 Chen MF Wu CT Chen YJ Keng PC Chen WC Cell killing and radiosensitization by caffeic acid phenethyl ester in lung cancer cells J Radiat Res 2004 45 253 260 15304968 10.1269/jrr.45.253 Huang MT Ma W Yen P Xie JG Han J Frenkel K Grunberger D Conney AH Inhibitory effects of caffeic acid phenethyl ester (CAPE) on 12-O-tetradecanoylphorbol-13-acetate-induced tumor promotion in mouse skin and the synthesis of DNA, RNA and protein in HeLa cells Carcinogenesis 1996 17 761 5 8625488 Hall EJ The Physics and chemistry of radiation absorption Radiology for the radiologist 2000 5 Philadelphia: J.B. Lippincott 5 16 Seed T Kumar S Whitnall M Srinivasan V Singh V Elliott T Landauer M Miller A Chang CM Inal C Deen J Gehlhaus M Jackson W 3rdHilyard E Pendergrass J Toles R Villa V Miner V Stewart M Benjack J Danilenko D Farrell C New strategies for the prevention of radiation injury: possible implications for countering radiation hazards J Radiat Res 2002 43 S239 S244 12793766 10.1269/jrr.43.S239 Rubin P Johnston CJ Williams JP McDonald S Finkelstein JN A perpetual cascade of cytokines postirradiation leads to pulmonary fibrosis Int J Radiat Oncol Biol Physics 1995 33 99 109 10.1016/0360-3016(95)00095-G Hong JH Chiang CS Tsao CY Lin PY McBride WH Wu CJ Rapid induction of cytokine gene expression in the lung after single and fractionated doses of radiation Int Radiat Biol 1999 75 1421 1427 10.1080/095530099139287 Michaluart P Masferrer JL Carothers AM Subbaramaiah K Zweifel BS Koboldt C Mestre JR Grunberger D Sacks PG Tanabe T Dannenberg AJ Inhibitory effects of caffeic acid phenethyl ester on the activity and expression of cyclooxygenase-2 in human oral epithelial cells and in a rat model of inflammation Cancer Res 1999 59 2347 52 10344742 Orban Z Mitsiades N Burke TR JrTsokos M Chrousos GP Caffeic acid phenethyl ester induces leukocyte apoptosis, modulates nuclear transcription factor NF-κB and suppresses acute inflammation Neuroimmunomodul 2000 7 99 105 10.1159/000026427 Natarajan K Singh S Burke TR JrGrunberger D Aggarwal BB Caffeic acid phenethyl ester is a potent and specific inhibitor of activation of nuclear transcription factor NF-κB Proc Natl Acad Sci USA 1996 93 9090 9095 8799159 10.1073/pnas.93.17.9090 Fitzpatrick LR Wang J Le T Caffeic acid phenethyl ester, an inhibitor of nuclear transcription factor NF-κB, attenuates bacterial peptidoglycan polysaccharide- induced colitis in rats J Pharmacol Exp Ther 2001 299 915 920 11714876 Chen MF Keng PC Shau H Wu CT Hu YC Liao SK Chen WC Inhibition of lung tumor growth and augmentation of radiosensitivity by decreasing peroxiredoxin I expression Int J Radiat Oncol Biol Physics Mirkovic N Voehringer DW Story MD McConkey DJ McDonnell TJ Meyn RE Resistance to radiation- induced apoptosis Bcl-2- expressing cells is reversed by depleting cellular thiols Oncogene 1997 15 1461 70 9333022 10.1038/sj.onc.1201310 Nardini M Leonardi F Scaccini C Virgili F Modulation of ceramide- induced NF-κB binding activity and apoptotic response by caffeic acid in U937 cells: Comparsion with other antioxidants Free Rad Biol Med 2000 30 722 33 10.1016/S0891-5849(00)00515-3 Kucharczak J Simmons MJ Fan Y G'eline C To be, or not to be: NF-κB is the answer – role of Rel/ NF-κB in the regulation of apoptosis Oncogene 2003 22 8961 8982 14663476 10.1038/sj.onc.1207230 Russo SM Tepper JE Baldwin AS JrLiu R Adams J Elliott P Cusack JC Jr Enhancement of radiosensitivity by proteason inhibition: implication for a role of NF-κB Int J Radiat Oncol Biol Physics 2001 50 183 193 10.1016/S0360-3016(01)01446-8 de Ridder M Verovski VN Van den Berge DL Sermeus AB Monsaert C Wauters N Storme GA Lipid A radiosensitizes hypoxic EMT-6 tumor cells: role of the NF-κB signaling pathway Int J Radiat Oncol Biol Physics 2003 57 779 786 10.1016/S0360-3016(03)00662-X Wang JF Jerrells TR Spitzer JJ Decrease production of reactive oxygen intermediates is an early event during in vitro apoptosis of rat thymocytes Free Rad Biol Med 1996 20 533 42 8904294 10.1016/0891-5849(95)02085-3 Molls M van Beuningen D Scherer E, Streffer C, Trott KR Radiation injury of the lung: Experimental studies, observations after radiotherapy and total body irradiation prior to bone marrow transplantation Radiopathology of organs and tissues 1991 Berlin: Springer- Verlag 369 404 Franko AJ Sharplin J Development of fibrosis after lung irradiation in relation to inflammation and lung function in mouse strain prone to fibrosis Radiat Res 1994 140 347 355 7972687 Neurath MF Becker C Barbulescu K NF-κB in immune and inflammatory responses in the gut Gut 1998 43 856 860 9824616 Cuzzocrea S Pisano B Dugo L Ianaro A Patel NS Caputi AP Thiemermann C Tempol reduced the activation of nuclear factor-κB in acute inflammation Free Radical Research 2004 38 813 819 15493454 10.1080/10715760410001710829 Haase MG Klawitter A Geyer P Alheit H Baumann M Kriegel TM Kasper M Baretton GB Sustained elevation of NF-κB DNA binding activity in radiation- induced lung damage in rats Int J Radiat Biol 2003 79 863 877 14698955 10.1080/09553000310001632903 Linard C Marquette C Mathieu J Pennequin A Clarencon D Denis M Acute induction of inflammatory cytokine expression after γ- irradiation in the rat: effect of an NF-κB inhibitor Int J Radiat Oncol Biol Physics 2004 58 427 434 10.1016/j.ijrobp.2003.09.039 Wang Y Meng A Lang H Brown SA Konopa JL Kindy MS Schmiedt RA Thompson JS Zhou D Activation of nuclear factor κB in vivo selectively protects the murine small intestine against ionizing radiation induced damage Cancer Research 2004 64 6240 6246 15342410 10.1158/0008-5472.CAN-04-0591 Haddad JJ Land SC Nuclear Factor-κB blockade attenuates but does not abrogate lipopolysaccharide-dependent tumor necrosis factor-alpha biosynthesis in alveolar epithelial cells Biochem Biophys Res Commun 2001 285 267 272 11444836 10.1006/bbrc.2001.5172 McBride WH Cytokine cascades in late normal tissue radiation responses Int J Radiat Oncol Biol Physics 1995 33 233 234 10.1016/0360-3016(95)02019-8 Kovacs EJ Fibrogenic cytokines: the role of immune mediators in the development of scar tissue Immunology Today 1991 12 17 23 2015044 10.1016/0167-5699(91)90107-5 Rube CE Uthe D Schmid KW Richter KD Wessel J Schuck A Willich N Rube C Dose-dependent induction of transforming growth factor β (TGF-β) in the lung tissue of fibrosis- prone mice after thoracic irradiation Int J Radiat Oncol Biol Physics 2000 47 1033 1042 10.1016/S0360-3016(00)00482-X Sime PJ Xing Z Graham FL Csaky KG Gauldie J Adenovirus- mediated gene transfer of active transforming factor -β induces prolonged severe fibrosis in rat lung J clin Invest 1997 100 768 766 9259574 Rameshwar P Narayanan R Qian J Denny TN Colon C Gascon P NF-κB as a central mediator in the induction of TGF-β in monocytes from patients with idiopathic myelofibrosis: an inflammatory response beyond the real of hemeostasis J of Immunology 2000 165 2271 2277 10925316 Lawrance IC Wu F Leite AZ Willis J West GA Fiocchi C Chakravarti S A murine model of chronic inflammation- induced intestinal fibrosis down- regulated by abtisense NF-Kappa B Gastroenterlogy 2003 125 1750 61 10.1053/j.gastro.2003.08.027 Warshamana GS Corti M Brody AR TNF-α, PDGF, and TGF-beta (1) expression by primary mouse bronchiolar alveolar epithelial and mesenchymal cells: TNF-α induces TGF-beta (1) Exp Mol Patho 2001 71 113 33 Liu JY Brody AR Increased TGF-β1 in the lungs of asbestos-exposed rats and mice: reduced expression in TNF- alpha receptor knockout mice J Environ Pathol Toxicol Oncol 2001 20 97 108 11394717	16336675	PMC1325253	CC BY	2021-01-04 16:03:08	no	BMC Cancer. 2005 Dec 9; 5:158	utf-8	BMC Cancer	2,005	10.1186/1471-2407-5-158	oa_comm
==== Front BMC NeurolBMC Neurology1471-2377BioMed Central London 1471-2377-5-231632116610.1186/1471-2377-5-23Research ArticleThe effect of body mass index on global brain volume in middle-aged adults: a cross sectional study Ward Michael A [email protected] Cynthia M [email protected] Mehul A [email protected] Mark A [email protected] Sterling C [email protected] Geriatric Research Education and Clinical Center, Wm. S. Middleton VA Hospital, Madison, WI, USA2 Geriatrics and Adult Development, University of Wisconsin Medical School, Madison, WI, USA2005 2 12 2005 5 23 23 4 4 2005 2 12 2005 Copyright © 2005 Ward et al; licensee BioMed Central Ltd.2005Ward 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 Obesity causes or exacerbates a host of medical conditions, including cardiovascular, pulmonary, and endocrine diseases. Recently obesity in elderly women was associated with greater risk of dementia, white matter ischemic changes, and greater brain atrophy. The purpose of this study was to determine whether body type affects global brain volume, a marker of atrophy, in middle-aged men and women. Methods T1-weighted 3D volumetric magnetic resonance imaging was used to assess global brain volume for 114 individuals 40 to 66 years of age (average = 54.2 years; standard deviation = 6.6 years; 43 men and 71 women). Total cerebrospinal fluid and brain volumes were obtained with an automated tissue segmentation algorithm. A regression model was used to determine the effect of age, body mass index (BMI), and other cardiovascular risk factors on brain volume and cognition. Results Age and BMI were each associated with decreased brain volume. BMI did not predict cognition in this sample; however elevated diastolic blood pressure was associated with poorer episodic learning performance. Conclusion These findings suggest that middle-aged obese adults may already be experiencing differentially greater brain atrophy, and may potentially be at greater risk for future cognitive decline. ==== Body Background The prevalence of overweight and obese people in the United States and other developed nations has progressively increased over the last twenty years and is now at epidemic proportions[1]. It is estimated that greater than 60% of all Americans are overweight, and approximately one-half of that group are classified as obese[2]. Previous studies have found that obesity reduces life expectancy [3] by causing or exacerbating various medical conditions including coronary heart disease (CHD), type 2 diabetes mellitus, hypertension, obstructive sleep apnea, and stroke[4]. Neurocognitive health may also be related to obesity. A recent study determined that obesity was strongly associated with poorer cognitive function in individuals over 65 years of age [5]. In a population-based sample of women aged 70–89 years, greater body mass index (BMI) in middle and later life was associated with cerebral white matter ischemic change [6], a higher incidence of dementia, particularly Alzheimer's disease (AD) [7], and temporal lobe atrophy [8] in later life. Brain atrophy involves the loss of tissue volume and is commonly seen with increasing age [9-11] and neurodegenerative disease[12]. Vascular factors intrinsic to overweight individuals, such as hypertension[13,14], hypercholesterolemia [13,15], endothelial dysfunction[16,17], and diabetes [18-20] have all been linked to greater risk for dementia or brain atrophy in the elderly. Furthermore, older adults with better cardiovascular fitness demonstrate significant improvements in cognitive function and a significant slowing of age-related atrophy of gray and white matter[21]. Together these findings suggest that older overweight individuals have a higher risk of accelerated brain atrophy and concomitant cognitive decline. While the deleterious effects of obesity on the brain in the geriatric population are now apparent, it is not known whether this relationship occurs in younger persons or is unique to older populations. This is an important question because interventions to reduce the adverse effects of obesity may have a larger public health impact when implemented at younger ages. The purpose of the present study was to determine whether the effect of BMI on brain atrophy previously observed in elderly females [8] might also be observable in cognitively healthy adults between the ages of 40 and 66, and to determine the relationships between this effect and associated cardiovascular factors (hypertension and hypercholesterolemia). Methods One hundred seventeen participants (44 male, 73 female) with a mean age of 54.2 years (SD = 6.5) were studied with magnetic resonance imaging (MRI) and cognitive testing as part of a cross-sectional study analyzing factors related to global brain volume and cognition. Sixty-five participants were recruited from an existing registry known as the Wisconsin Registry for Alzheimers' Prevention (WRAP)[22] consisting of cognitively normal middle-aged adults who had at least one parent with AD. These participants were recruited to enrich the sample with individuals having risk factors for AD. The remaining fifty-two participants were recruited somewhat simultaneously from the University of Wisconsin-Madison (UWM) community. This convenience sample was selected to have no known first-degree family history of AD (with parents surviving until at least age 70 without dementia). All participants in this study were required to be between the ages of 40 and 66 and have no current major Axis I psychiatric disease or history of major medical conditions (i.e., traumatic brain injury, neurovascular infarctions, brain neoplasms or ischemic changes, history of cancer, diabetes, or condition requiring an invasive brain procedure). Additionally, participants were required to have normal cognitive function, and MRI scanner compatibility. Lastly, participants on any medication with potential to affect cerebral perfusion or cognition (such as beta blockers, calcium channel antagonists, Angiotensin-converting Enzyme (ACE) inhibitors, statins, or Selective Serotonin Reuptake Inhibitors (SSRIs)) were excluded from the analysis. All participants completed a detailed health history questionnaire, and were administered a battery of neuropsychological tests and laboratory blood tests. Data were collected on Apolipoprotein E (APOE) genotype, non-fasting total blood cholesterol level, blood pressure (BP), height and weight (for BMI calculation). BP was measured with the subject seated and at rest using an automated BP machine. Body height and weight were collected to the nearest 0.5-inch and one pound respectively. The battery of neuropsychological tests [23] included the following: portions of the Wechsler Adult Intelligence Scale-Third Edition (WAIS-III), the Rey Auditory Verbal Learning Test (RAVLT), Trail Making Test A and B, and the Center for Epidemiological Studies Depression (CES-D) Inventory. All participants gave written informed consent under a protocol approved by the local institutional review board. This study was performed in a manner that was in accordance with the Declaration of Helsinki. Brain imaging MRI was performed using a General Electric 3.0 Tesla SIGNA (Waukesha, WI) MRI system. A 3D IR-prepped fast gradient echo pulse sequence provided high-resolution T1-weighted structural images. The whole brain was imaged in the axial plane with the following parameters: inversion time = 600 ms, fast gradient echo read-out with TR/TE/flip = 9 ms/1.8 ms/20°; acquisition matrix = 256 × 192 × 124 (interpolated to 256 × 256 × 124); field of view = 240 mm; slice thickness = 1.2 mm (124 slices); ± 16 kHz receiver bandwidth. A Fast Recovery Fast Spin Echo 2D T2-weighted axial sequence was also acquired with the same start and stop locations as the T1 weighted images. The parameters were: field of view = 240 mm, matrix 256 × 256 × 64, TR = 9000 ms, TE = 93 ms, flip angle = 90. Seventy slices were acquired; slice thickness = 1.7 mm with 0.3 mm skip. An experienced neuroradiologist examined all images for evidence of any neurovascular disease or structural abnormality that would exclude the subject from the analysis (see exclusions above). Global brain atrophy determination Global brain volumes, calculated from T1-weighted images, were analyzed using the cross sectional method of Structural Image Evaluation, using Normalization, of Atrophy (SIENAX) within the FSL 3.1β software suite[24]. The SIENAX parameters were set for a three-class segmentation of tissue type and at a 0.3 threshold for segmenting the brain from extra-axial soft tissue. SIENAX analysis yielded whole-brain volumetric data, in units of cubic millimeters, for three different tissue types: cerebral spinal fluid (CSF), gray matter, and white matter. Total intracranial volume (TICV) was defined as the sum of the three tissue types, and whole brain parenchyma volume (BV) was defined as the sum of gray matter and white matter. BV was divided by TICV to obtain a normalized brain volume (NBV) with respect to head size. BMI classification BMI values were calculated in standard fashion by dividing the weight in kilograms by the square of height in meters. BMI groups were defined using the World Health Organization's (WHO) classification system: Underweight, less than 18.5; Normal, 18.5–24.9; Overweight, 25.0–29.9; and Obese, greater than 30.0 (in units kg/m2) [25]. There were not enough underweight participants in this study group to analyze the effects of underweight on global brain atrophy. Therefore, individuals classified as underweight were not included in the statistical analysis. Statistical analysis First, Student's t-tests were performed to determine whether the participants differed by family history of AD status/referral source. These analyses revealed that UWM participants were more highly educated than the WRAP participants, mean = 16.9 (SD = 2.4) and mean = 15.9 (SD = 2.5) respectively (t [112] = -2.10; p= 0.038). In addition, the frequency of the APOE ε4 allele was more prevalent in WRAP participants (54.0 %) compared to the UWM participants (13.7 %) (chi-square [1] = 23.5; p < 0.001). No other variables were found to be significantly different between the two groups. Nevertheless, 1st degree family history of AD was treated as an independent variable in the ensuing stepwise regression model. Next, linear regression analyses and stepwise regression analyses were used to predict the effect of BMI, age, non-fasting total blood cholesterol level, systolic and diastolic BP, family history of AD, gender, education, and genotype on the dependent variables of NBV and cognition in separate analyses. Family history of AD, gender, and APOE genotype were entered into the model as dichotomous variables: history/no history of AD, male/female, and presence/absence of the ε4 allele, respectively. A method advocated by Baron et al. [26] was used to determine whether any variable within the stepwise linear regression model mediated the effect of BMI on NBV. Three participants were not included in the brain volume analysis. Two participants were classified as underweight and one participant was discovered to have a previously undiagnosed brain tumor. Therefore the final sample size equaled 114 participants. Nine participants were not included in the cognition analysis; five participants scored a fifteen or higher on the CES-D inventory suggesting symptoms of depression, one participant did not complete the CES-D inventory, and the remaining three participants were the same participants excluded in the brain atrophy analysis (n= 108 participants). Therefore the final sample size for the cognition analysis equaled 108 participants. Results One hundred fourteen middle-aged (mean age 54) adults were administered a comprehensive battery of neuropsychological tests and structural MRI scans in this cross-sectional study examining the brain and cognitive effects associated with various AD risk factors. Baseline demographics, cognitive measures, and predictor variables are shown in Tables 1 and 2. Table 1 Demographic and cognitive measures for men and women Demographic and Cognitive Variables Mean (SD) Age (years) 54.2 (6.6) Education (years) 16.4 (2.5) Classification: n = Normal / Overweight / Obesea 51 / 42 / 21 Trail Making Test B (seconds)b 61.3 (20.2) WAIS-III-Digit Span raw scoreb 17.9 (3.8) RAVLT-Total raw scoreb 49.6 (7.9) CES-Db 4.6 (5.2) Data are presented as mean (standard deviation). Notes: aNormal, overweight, and obese body types are defined using the World Health Organization's classification system. bThe cognitive and mood variables are presented for the 108 participants in the cognitive analysis. Demographic variables are for the full sample of 114. Table 2 Independent associations between each predictor variable and age-adjusted NBV Regressor Mean (SD) Range β-value p-value BMI (kg/m2) 26.0 (4.4) (19.0–39.7) -0.22 0.010 Total Cholesterol (mg/dL) 208.0 (37.8) (114–339) -0.12 0.178 Systolic BP (mm Hg) 132.4 (17.2) (102–205) -0.15 0.092 Diastolic BP (mm Hg) 79.4 (10.4) (54–110) -0.09 0.328 Family History AD (y/n) 63/51 0.07 0.448 APOE Genotype (ε4/no ε4) 41/73 -0.04 0.660 Gender (Male/Female) 43/71 0.13 0.137 n = 114. β-values, acquired using individual linear regression analyses with each of the above predictors entered after age, reflect age-adjusted predictions of NBV. An initial linear regression analysis examined the effect of age and BMI on NBV to determine whether the effect observed in elderly females [8] could be observed in younger men and women. This analysis indicated that both age (β = -0.390; t = -4.614; p < 0.0001) and BMI (β = -0.220; t = -2.605; p = 0.010) were significant predictors of NBV (R2 = 0.206; F [111, 2] = 14.432; p < 0.00001). Figure 1 shows that elevated BMI is associated with decreased NBV (after adjusting for age). To determine whether additional variables may also predict NBV, a stepwise linear regression analysis was used to examine the effects of age, BMI, family history of AD, APOE genotype, total cholesterol, systolic and diastolic BP, and gender on NBV. This analysis found that age (β = -0.389; t = -4.337; p < 0.001) and BMI (β = -0.224; t = -2.495; p < 0.014) together were the best predictors of NBV (R2 = 0.205; F [99, 2] = 12.756; p < 0.0001). No other variables were significant. Figure 1 NBV versus BMI. Age adjusted values of NBV plotted versus BMI scores. NBV is shown to decrease proportionally with increasing BMI (r = -0.232, p < 0.03). Pearson correlations determined that BMI was significantly associated with cholesterol (r = 0.21, p < 0.04) and systolic BP (r = 0.36, p < 0.001) but was not associated with any other predictor variable. Neither cholesterol nor systolic BP were significantly associated with NBV (results are summarized in Table 2), and therefore did not appear to mediate the effect of BMI on NBV according to criteria specified by Baron et al. [26]. Next we applied the same stepwise model above, including education to predict the non-crystallized cognitive abilities of learning, processing speed and working memory in separate regression models. Age (β = -0.315; t = -3.417; p < 0.001) and diastolic BP (β = -0.233; t = -2.526; p = 0.013) were significant predictors of learning (R2 = 0.157; F [101, 2] = 9.252; p < 0.001). Age was the only significant predictor of processing speed (R2 = 0.081; F [100, 1] = 8.454; p < 0.01). There was no significant predictor of working memory in this sample. Average neuropsychological scores are given in Table 1. Discussion This study found that elevated BMI is associated with reduced brain volumes, suggesting greater global brain atrophy in middle-aged adults even after adjusting for age. Episodic learning, working memory, and processing speed abilities were not associated with BMI in this sample indicating that the possible effect on cerebral atrophy had not influenced cognition in participants with a high BMI. Episodic learning was related to diastolic BP, which indicates that certain cardiovascular risk factors may be associated with cognition. However, this finding was isolated to only one of our three categories of cognition. Longitudinal studies are needed to assess future cognitive consequences associated with cardiovascular risk factors. Our finding that BMI affects brain volume at a relatively young age extends prior research on the harmful effects of obesity on the brain. We found a 2.4% difference in brain parenchyma volume for participants classified by the WHO as obese (n = 21 people; average BMI= 33.3 kg/m2) compared to participants at a normal, healthy weight (n = 51 people; average BMI = 22.6 kg/m2). Although the magnitude of this difference is not large in this cross-sectional sample, when we consider these results in the context of the younger age of our sample (54.2 years), it follows that obesity in middle age may render an individual more vulnerable to brain atrophy over subsequent years, and thus more vulnerable to cognitive decline or dementia. The results of this study are consistent with others showing that life choices (i.e., proper diet, physical activity) may be neuroprotective and, therefore, potentially reduce the rate of brain atrophy and concomitant cognitive decline [27,28]. A longitudinal study of obese people at middle age is needed to test these hypotheses. It is unclear through what mechanisms obesity affects brain volume. Central obesity is associated with risk factors composing the metabolic syndrome, including high triglyceride levels, low HDL cholesterol levels, hypertension, insulin resistance, and prothrombotic and proinflammatory states. Prior studies indicate that some of these risk factors such as mid-life elevations in blood pressure, total cholesterol, and inflammatory markers, are risk factors for late-life dementia[13,29-31]. In the present study, neither BP nor non-fasting total cholesterol was associated with current NBV (though systolic BP did show a trend). These vascular risk factors may affect chronic cerebral perfusion and β-amyloid generation, thus influencing neuronal degeneration. Leptin, a peptide related to obesity, also affects β-amyloid regulation [32]. Further studies are needed to determine the mechanism by which BMI affects brain atrophy in middle-aged adults, and whether this may have future deleterious consequences on brain structure or function. Brain tissue is strongly dependent upon oxygen for survival and is thus vulnerable to hypoxic and ischemic conditions. Hypoxia is an inherent consequence of many of the conditions caused or exacerbated by obesity. Obstructive sleep apnea[4], CHD [33], asthma [34], and low cardiovascular fitness[21] may lead to hypoxia which, in chronic situations, may lead to cognitive decline and neuronal death[35]. Hypoxia is often a consequence of hypoperfusion, which is strongly associated with endothelial dysfunction common to individuals with hypertension, atherosclerosis, and CHD. Endothelial dysfunction, as a result of oxidative stress, may induce neuronal damage and initiate neurodegenerative change[17]. Prevention or postponement of the onset of dementia has the potential to drastically impact the prevalence of dementia. The prevalence of AD in the United States alone is projected to quadruple in the next 50 years, and if disease onset could be delayed by only one year, it would result in 800,000 fewer cases[36]. It has been shown that a person's weight is a reflection of their habitual physical activity[4]. Given the rapid increase in obesity in the United States, targeting weight management may significantly impact the prevalence of dementia. High dietary fat intake has been shown to increase the risk of dementia [37]. Physical fitness significantly improves cognitive function in older adults[21] and exercise increases the production of brain-derived neurotrophic factor (BDNF) and insulin-like growth factor I (IGF-I). BDNF and IGF-I are neurotrophic hormones that are important for neurogenesis, which may protect against neurodegeneration seen with aging[38]. In older adults the rate of change in brain volume may be diminished in persons with better cardiovascular fitness[39]. There were some limitations of this study. First, the range of BMI was truncated because of the small-bore radius (55 cm) inherent to the MRI scanner. Therefore, participants in this study could not exceed 260 pounds, which did not permit an analysis of the effects of extreme obesity. Next, insulin-dependent diabetic participants were excluded from this study because diabetes has been shown to be a risk factor for dementia [18-20] and therefore could be an important confounder in the analysis. The incidence of diabetes in our study population was less than 1%, which was not enough to determine its affect on global brain volume. Next, the design of this study was cross-sectional which only allowed for inferences between participants over a single time point. A longitudinal study is needed to clarify if the reduced brain volume associated with elevated BMI is representative of a progressive brain atrophy and concomitant cognitive decline. Additionally, non-fasting, instead of the more accurate fasting blood cholesterol values were used as potential predictors of NBV. Furthermore, the BMI measurement itself is only an estimate of obesity; it does not accurately distinguish fat mass from lean mass. The highest risk for adverse consequences related to overweight individuals are best indicated by excess amounts of adipose tissue, especially that within the intra-abdominal region. Waist circumference and skinfold thickness measurements in conjunction with BMI would have provided a more precise measure of the amount and location of this excess body fat. Finally, this study population was self-selected from a group of highly educated, motivated volunteers; these findings may not be representative of the general population. However, these findings are important in understanding how to potentially target preventive therapies for at-risk individuals. Conclusion This study found that elevated BMI is associated with lesser brain volume in middle-aged adults (mean age = 54 years) even after adjusting for age. This study extends prior research that indicated a similar effect in a population-based sample of females aged 70 to 84 years with elevated BMIs from middle age through later life (46 to 84 years). Further, longitudinal studies are needed to determine whether the effect we have observed at middle age results in greater rate of cerebral atrophy over time, and whether this effect may increase the risk of future cognitive decline and incidence of dementia. Competing interests The author(s) declare that they have no competing interests. Authors' contributions MAW carried out the statistical analysis, drafted the manuscript, and assisted in the acquisition of data. CMC contributed to study conception, participated in its design, and assisted in drafting the manuscript. MAT assisted in data acquisition, statistical analyses, interpreting the results, and drafting the manuscript. MAS assisted in data acquisition, interpreting the results, and writing the manuscript. SCJ drafted the manuscript, conceived of the study concept and design, assisted in the statistical analysis, and interpretation of the results. All authors read and approved the final manuscript. Pre-publication history The pre-publication history for this paper can be accessed here: Acknowledgements This study was supported by AG021155. The assistance of Britta Torgerson, Taylor Schmitz, Michele Ries, Ph. D., Howard Rowley, M.D., Rebecca Koscik, Ph.D. Justin Dunker, Kristi Kalmoe, Allie Wichmann, Michael Anderle, and Ron Fisher is greatly appreciated. We especially thank the participants of this study. ==== Refs Kuczmarski RJ Flegal KM Campbell SM Johnson CL Increasing prevalence of overweight among US adults. The National Health and Nutrition Examination Surveys, 1960 to 1991 Jama 1994 272 205 211 8022039 10.1001/jama.272.3.205 Flegal KM Carroll MD Ogden CL Johnson CL Prevalence and trends in obesity among US adults, 1999-2000 Jama 2002 288 1723 1727 12365955 10.1001/jama.288.14.1723 Olshansky SJ Passaro DJ Hershow RC Layden J Carnes BA Brody J Hayflick L Butler RN Allison DB Ludwig DS A Potential Decline in Life Expectancy in the United States in the 21st Century The New England Journal of Medicine 2005 352 1138 1145 15784668 10.1056/NEJMsr043743 Kopelman PG Obesity as a medical problem Nature 2000 404 635 643 10766250 Jeong SK Nam HS Son MH Son EJ Cho KH Interactive Effect of Obesity Indexes on Cognition Dement Geriatr Cogn Disord 2004 19 91 96 15591798 10.1159/000082659 Gustafson DR Steen B Skoog I Body mass index and white matter lesions in elderly women. An 18-year longitudinal study Int Psychogeriatr 2004 16 327 336 15559756 10.1017/S1041610204000353 Gustafson D Rothenberg E Blennow K Steen B Skoog I An 18-year follow-up of overweight and risk of Alzheimer disease Arch Intern Med 2003 163 1524 1528 12860573 10.1001/archinte.163.13.1524 Gustafson D Lissner L Bengtsson C Bjorkelund C Skoog I A 24-year follow-up of body mass index and cerebral atrophy Neurology 2004 63 1876 1881 15557505 Fotenos AF Snyder AZ Girton LE Morris JC Buckner RL Normative estimates of cross-sectional and longitudinal brain volume decline in aging and AD Neurology 2005 64 1032 1039 15781822 Jernigan TL Archibald SL Berhow MT Sowell ER Foster DS Hesselink JR Cerebral structure on MRI, Part I: Localization of age-related changes Biol Psychiatry 1991 29 55 67 2001446 10.1016/0006-3223(91)90210-D Blatter DD Bigler ED Gale SD Johnson SC Anderson CV Burnett BM Parker N Kurth S Horn SD Quantitative volumetric analysis of brain MR: normative database spanning 5 decades of life AJNR Am J Neuroradiol 1995 16 241 251 7726068 Kantarci K Jack CRJ Neuroimaging in Alzheimer disease: an evidence-based review Neuroimaging Clin N Am 2003 13 197 209 13677801 10.1016/S1052-5149(03)00025-X Kivipelto M Helkala EL Laakso MP Hanninen T Hallikainen M Alhainen K Iivonen S Mannermaa A Tuomilehto J Nissinen A Soininen H Apolipoprotein E epsilon4 allele, elevated midlife total cholesterol level, and high midlife systolic blood pressure are independent risk factors for late-life Alzheimer disease Ann Intern Med 2002 137 149 155 12160362 Heijer T Skoog I Oudkerk M de Leeuw FE de Groot JC Hofman A Breteler MM Association between blood pressure levels over time and brain atrophy in the elderly Neurobiol Aging 2003 24 307 313 12498964 10.1016/S0197-4580(02)00088-X Meyer JS Rauch GM Crawford K Rauch RA Konno S Akiyama H Terayama Y Haque A Risk factors accelerating cerebral degenerative changes, cognitive decline and dementia Int J Geriatr Psychiatry 1999 14 1050 1061 10607973 10.1002/(SICI)1099-1166(199912)14:12<1050::AID-GPS56>3.0.CO;2-Z Togo T Katsuse O Iseki E Nitric oxide pathways in Alzheimer's disease and other neurodegenerative dementias Neurol Res 2004 26 563 566 15265275 10.1179/016164104225016236 Aliev G Obrenovich ME Smith MA Perry G Hypoperfusion, Mitochondria Failure, Oxidative Stress, and Alzheimer Disease J Biomed Biotechnol 2003 2003 162 163 12975531 10.1155/S1110724303305029 Leibson CL Rocca WA Hanson VA Cha R Kokmen E O'Brien PC Palumbo PJ Risk of dementia among persons with diabetes mellitus: a population-based cohort study Am J Epidemiol 1997 145 301 308 9054233 Ott A Stolk RP van Harskamp F Pols HA Hofman A Breteler MM Diabetes mellitus and the risk of dementia: The Rotterdam Study Neurology 1999 53 1937 1942 10599761 Schnaider Beeri M Goldbourt U Silverman JM Noy S Schmeidler J Ravona-Springer R Sverdlick A Davidson M Diabetes mellitus in midlife and the risk of dementia three decades later Neurology 2004 63 1902 1907 15557509 Kramer AF Bherer L Colcombe SJ Dong W Greenough WT Environmental influences on cognitive and brain plasticity during aging J Gerontol A Biol Sci Med Sci 2004 59 M940 57 15472160 Sager MA Hermann BP La Rue A Middle-aged Children of Persons with Alzheimer's Disease: APOE Genotypes and Cognitive Function in the Wisconsin Registry for Alzheimer's Prevention Journal of Geriatric Psychiatry and Neurology Spreen O Strauss E A Compendium of Neuropsychological Tests 1998 New York , Oxford Smith SM Zhang Y Jenkinson M Chen J Matthews PM Federico A De Stefano N Accurate, robust, and automated longitudinal and cross-sectional brain change analysis Neuroimage 2002 17 479 489 12482100 10.1006/nimg.2002.1040 Obesity: preventing and managing the global epidemic. Report of a WHO consultation World Health Organ Tech Rep Ser 2000 894 i xii, 1-253 11234459 Baron RM Kenny DA The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations J Pers Soc Psychol 1986 51 1173 1182 3806354 10.1037/0022-3514.51.6.1173 Scarmeas N Stern Y Cognitive reserve: implications for diagnosis and prevention of Alzheimer's disease Curr Neurol Neurosci Rep 2004 4 374 380 15324603 Whalley LJ Deary IJ Appleton CL Starr JM Cognitive reserve and the neurobiology of cognitive aging Ageing Res Rev 2004 3 369 382 15541707 10.1016/j.arr.2004.05.001 Kivipelto M Helkala EL Laakso MP Hanninen T Hallikainen M Alhainen K Soininen H Tuomilehto J Nissinen A Midlife vascular risk factors and Alzheimer's disease in later life: longitudinal, population based study Bmj 2001 322 1447 1451 11408299 10.1136/bmj.322.7300.1447 Knopman D Boland LL Mosley T Howard G Liao D Szklo M McGovern P Folsom AR Cardiovascular risk factors and cognitive decline in middle-aged adults Neurology 2001 56 42 48 11148234 Schmidt R Schmidt H Curb JD Masaki K White LR Launer LJ Early inflammation and dementia: a 25-year follow-up of the Honolulu-Asia Aging Study Ann Neurol 2002 52 168 174 12210786 10.1002/ana.10265 Fewlass DC Noboa K Pi-Sunyer FX Johnston JM Yan SD Tezapsidis N Obesity-related leptin regulates Alzheimer's Abeta Faseb J 2004 18 1870 1878 15576490 10.1096/fj.04-2572com Bass JL Corwin M Gozal D Moore C Nishida H Parker S Schonwald A Wilker RE Stehle S Kinane TB The effect of chronic or intermittent hypoxia on cognition in childhood: a review of the evidence Pediatrics 2004 114 805 816 15342857 10.1542/peds.2004-0227 Ronmark E Andersson C Nystrom L Forsberg B Jarvholm B Lundback B Obesity increases the risk of incident asthma among adults Eur Respir J 2005 25 282 288 15684292 10.1183/09031936.05.00054304 Lezak MD Neuropsychological assessment 2004 4th Oxford ; New York , Oxford University Press xiv, 1016 Brookmeyer R Gray S Kawas C Projections of Alzheimer's disease in the United States and the public health impact of delaying disease onset Am J Public Health 1998 88 1337 1342 9736873 Morris MC Evans DA Bienias JL Tangney CC Bennett DA Aggarwal N Schneider J Wilson RS Dietary fats and the risk of incident Alzheimer disease Arch Neurol 2003 60 194 200 12580703 10.1001/archneur.60.2.194 Cotman CW Berchtold NC Exercise: a behavioral intervention to enhance brain health and plasticity Trends Neurosci 2002 25 295 301 12086747 10.1016/S0166-2236(02)02143-4 Colcombe SJ Erickson KI Raz N Webb AG Cohen NJ McAuley E Kramer AF Aerobic fitness reduces brain tissue loss in aging humans J Gerontol A Biol Sci Med Sci 2003 58 176 180 12586857	16321166	PMC1325254	CC BY	2021-01-04 16:28:53	no	BMC Neurol. 2005 Dec 2; 5:23	utf-8	BMC Neurol	2,005	10.1186/1471-2377-5-23	oa_comm
==== Front BMC GenetBMC Genetics1471-2156BioMed Central London 1471-2156-6-551630956110.1186/1471-2156-6-55Research ArticleThe variable number of tandem repeats element in DAT1 regulates in vitro dopamine transporter density VanNess Sidney H [email protected] Michael J [email protected] Clinton D [email protected] Laboratory of Biological Psychopathology, Department of Psychiatry and Behavioral Sciences, Emory University School of Medicine, Atlanta, Georgia, USA2 Laboratory of Neuropsychopharmacology, Department of Psychiatry and Behavioral Sciences, Emory University School of Medicine, Atlanta, Georgia, USA2005 27 11 2005 6 55 55 11 7 2005 27 11 2005 Copyright © 2005 VanNess et al; licensee BioMed Central Ltd.2005VanNess et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background A 40-bp variable number of tandem repeats (VNTR) polymorphism exists in the 15th exon of DAT1, the gene encoding the human dopamine transporter (DAT). Though the VNTR resides in a region encoding the 3' untranslated region (UTR) and does not alter the protein's amino acid sequence, the prevalent 10-repeat variant has shown both linkage and association to Attention Deficit Hyperactivity Disorder (ADHD). In this study, we examined the effects of the DAT1 VNTR on measures of in vitro DAT expression and pharmacology. A series of four DAT1 constructs, each containing the DAT1 coding region, but varying with respect to the downstream presence or content of the 3'UTR, were engineered and stably transfected into an HEK-293 variant using Flp-In integration, an enzyme-mediated, site-specific recombination technology. Results [3H] Win 35,428 saturation binding assays and DAT immunoblots revealed statistically significant differences in DAT expression attributable to DAT1 genotype. Cells harboring the 10-repeat DAT1 variant were characterized by a Bmax approximately 50% greater than cells with the 9-repeat VNTR; those containing only the DAT1 coding region or the coding region flanked by a truncated 3' UTR resulted in greater DAT density than either of the naturalistic 9- and 10-repeat variants. Competition binding assays showed no statistically significant DAT1 genotype effects on the DAT affinity for methylphenidate, a finding consistent with the positional location of the VNTR. Conclusion This study identified the DAT1 VNTR as a functional polymorphism and provides an interpretive framework for its association with behavioral phenotypes. ==== Body Background The human dopamine transporter (hDAT), one member of a family of Na+/Cl- dependent transmembrane transport proteins, serves as a critical regulator of dopaminergic neurotransmission throughout much of the brain. The ~64 kb DAT1 gene (SLC6A3) that encodes the DAT includes 15 exons separated by 14 introns [1]. Concurrent with the cloning and chromosomal mapping of DAT1 to the short arm of chromosome 5 [2], a VNTR polymorphism was identified in the 15th exon, a region encoding the transcript's 3' UTR [3,4]. The 40-bp VNTR element is repeated between 3–13 times and in most human populations occurs with greatest frequency in the 9- and 10-repeat forms [5-7]. This polymorphic variation may be evolutionarily young, as a VNTR homologue has been observed in humans, chimpanzees, and cynomologus macaques, but not in lower mammals including the rat and mouse [8-10]. Though the VNTR resides in the 3'UTR and therefore does not affect the protein's amino acid sequence, regulatory factors such as mRNA stability and nuclear transport, and protein synthesis are potentially regulated by such variations [11-14]. Given the prominent role of dopamine neurotransmission in normal and abnormal behaviors, the DAT1 VNTR became the object of numerous genetic linkage and association studies [15-20], in vitro reporter gene experiments [21-26], in vivo SPECT molecular imaging studies [27-30], and pharmacogenetic examinations of the well-documented inter-individual variation in the response to treatment with DAT inhibitors [31-34]. Reports of the association of the DAT1 VNTR with ADHD have garnered particular attention, with the balance of evidence relating the 10-repeat VNTR to symptoms of the disorder [35]. Despite its high frequency in the general population [7] and an absence of studies addressing possible effects of the VNTR on measures of DAT physiology and pharmacology, the 10-repeat DAT1 allele has become generally recognized as a "high risk" allele for ADHD. The present study utilized radioligand binding and immunoblotting techniques to compare in vitro DAT density and ligand affinity across a series of cell lines stably transfected with constructs containing the DAT1 coding region flanked downstream by one of four DAT1 3'UTR variants. Flp-mediated recombination, one of a family of site-specific recombination technologies, was used to integrate the DAT1 constructs into a common, transcriptionally-active region of the genome, eliminating several caveats associated with traditional reporter gene methodologies [36]. Flp recombinase, the enzymatic basis of the system, is a Saccharomyces cerevisiae derived enzyme that utilizes a substrate Flp recombination target (FRT) sequence upstream of a gene of interest (GOI) to site-specifically insert the GOI into a target site in a host cell line, thereby providing a well-controlled approach to establishing the in vitro functional effect of the DAT1 VNTR polymorphism on the DAT [37]. Results DAT binding assays Results of three independent [3H] Win 35,428 saturation binding experiments revealed statistically significant (p < 0.05) differences in Bmax attributable to DAT1 VNTR genotype (Figure 2). Membranes from cells stably transfected with the DAT1 coding region (hDAT) displayed the highest mean Bmax. Extension of hDAT with the most proximal ~800 bp of the DAT1 3'UTR (hDAT Zero) was associated with decreased in vitro DAT density compared to the hDAT line. Of the two prevalent DAT1 VNTR variants (hDAT 9 and hDAT 10), cells stably transfected with construct harboring the 10-repeat allele displayed a mean Bmax 54% greater than those harboring the 9-repeat variant. Statistically significant differences in Kd were observed between DAT1 variants in saturation binding assays, though the magnitude of the effect was modest and considered physiologically insignificant (Figure 3); goodness of fit (R2) exceeded 0.98 for all non-linear regressions. Three independent competition binding experiments also revealed no pharmacologically significant effects of DAT1 genotype on methylphenidate's ability to displace [3H] Win 35,428 from the DAT (Figure 4). R2 values were ≥ 0.95 for all competition binding curves. Figure 1 Summary of DAT1 constructs used in Flp-In recombination protocol. Top left: hDAT, DAT1 coding region necessary and sufficient to produce a functional transporter protein. Top right: hDAT Zero, a construct containing the DAT1 coding region flanked by an ~800 bp fragment of the 3'UTR upstream of the VNTR region. Bottom left: hDAT 9, a construct with the DAT1 coding region upstream of a full length 3'UTR harboring the 9-repeat VNTR. Bottom right: hDAT 10, a construct with the DAT1 coding region upstream of a full length 3'UTR containing the 10-repeat VNTR. Figure 2 Comparative influence of 3'UTR variation on DAT saturation binding across three independent replicates. A) A characteristic binding curve revealing differences in Bmax attributable to DAT1 genotype. Mock-transfected negative control cells displayed no specific DAT binding. B) A one way ANOVA with Holm-Sidak post-hoc analysis demonstrated genotype-dependent differences in DAT density for all multiple comparisons (* = p < 0.05). Figure 3 Mean Kd measurements derived from three independent [3H] Win 35,428 saturation binding assays. A one way ANOVA with Holm-Sidak post-hoc analysis detected VNTR-dependent differences in the DAT ligand affinity (* = p < 0.05) for all but the hDAT-vs-hDAT9 compairson. Figure 4 Representative Competition Binding Curves. No pharmacologically significant differences exist between DAT1 variants in the encoded protein's affinity for methylphenidate. DAT western blots The observed rank order of immunoblotted DAT signal intensity across three independent replicates paralleled that of Bmax values derived from saturation binding experiments for the four different DAT1 constructs (Figure 5). Mock transfected cells displayed no DAT immunoreactivity. Uniformity of 14-3-3β immunoreactivity between the different transfected cell lines confirmed near equivalent between-lane protein loading and served as a basis for normalization. Significant differences (p < 0.05) were detected for all comparisons except hDAT 10-vs-hDAT Zero. Figure 5 Results of western analysis support the rank order of DAT density observed in saturation binding experiments. A) Lane 1: molecular weight marker, Lane 2: negative control, Lane 3: hDAT transfected cell line, Lane 4: hDAT Zero transfected cell line, Lane 5: hDAT 9 transfected cell line, Lane 6: hDAT 10 transfected cell line. Blots were probed both with a rat anti-DAT monoclonal antibody specific for the transporter's N-terminal region as well as a 14-3-3β polyclonal antibody to control for possible protein loading differences. DAT immunoreactivity was detected with an Alexa Fluor 680 conjugated anti-rat IgG (red) and 14-3-3β immunoreactivity with an IRDye-800 conjugated anti-rabbit IgG (green). B) DAT immunoreactivity was normalized to the 14-3-3β signal, densiometrically quantified with respect 14-3-3β content, and analyzed via a one way ANOVA with a Holm-Sidak post-hoc analysis. Statistically significant differences (* = p < 0.05) were detected for all comparisons except the hDAT10-vs-hDAT Zero comparison. Discussion The VNTR polymorphism of DAT1, though residing outside the gene's coding region, is of interest due to its possible influence on the regulation of dopaminergic neurotransmission, its implication in conferring genetic vulnerability for ADHD, and its putative role in modulating response to treatment with a first line medication, methylphenidate. The VNTR has not, however, been characterized as regards its possible influence on DAT physiology and pharmacology. Accumulating evidence that the 3'UTR influences the nuclear export, polyadenylation, subcellular targeting and rates of transcription and degradation of mRNA [11] supports the possibility that a VNTR polymorphism in this region could exert a regulatory influence on gene function. Observed DAT1 VNTR effects on reporter gene expression have provided initial support for an effect of the VNTR on in vitro gene expression [21-26], though the results vary widely (see [25] for discussion). Furthermore, heterologous reporter gene methodologies are subject to context-specific effects that may not reflect a sequence's impact on a gene of interest [38]. The present study provides multiple lines of converging in vitro evidence for the DAT1 VNTR's modulatory effect on DAT expression, thus supporting it as a functional polymorphism that may contribute to the recognized inter-individual differences in DAT density and dopaminergic function. The hDAT cell line, which lacks a 3'UTR altogether, displayed the highest overall DAT density and immunoreactivity, with a Bmax for [3H] Win 35,428 binding roughly 2.5-fold that of the hDAT 9 line. The hDAT Zero cell line had a Bmax for [3H] Win 35,428 binding ~25% less than that of the hDAT construct, a result that implicates the first ~800 bp of the DAT1 3'UTR in diminishing DAT1 transcription, mRNA stability, or translation. The additional extension of the 3'UTR to include either the 9 or 10 repeat variant further attenuated Bmax and DAT immunoreactivity compared to that for the hDAT and hDAT Zero cell lines. These results are consistent with the results of reporter gene assays [13,25] in which constructs containing the VNTR polymorphism were associated with decreased DAT1 expression. Incremental attenuation of DAT density by both the ~800 bp fragment and the VNTR region suggests the presence of at least two regulatory regions in the DAT1 15th exon, with the differential effects of the 9- and 10-repeat variants possibly due to interactions between a VNTR-specific attenuator and one contained within the ~800 bp fragment. Comparison of the 9- and 10-repeat variants, the two most frequent allelic variations of DAT1 [5,7] revealed that DAT binding site density for the 10-repeat VNTR was elevated approximately 50% over that of the 9-repeat allele, with western analysis confirming this finding. This result indicates that the 3'UTR does not simply mediate a length-dependent reduction in DAT availability, as has been previously suggested [22]. Rather, variation in the number of tandem repeats appears to have distinct modulatory influences on in vitro DAT density. At least five groups have reported in vitro reporter gene results examining the relative effects of the DAT1 9- and 10-repeat VNTR's influence on luciferase or GFP expression in transiently transfected systems. Fuke and coworkers compared the impact of DAT1 3'UTRs containing the 7-, 9- and 10-repeat alleles on luciferase reporter activity [21]. Their results suggested that the 10-repeat allele yielded the greatest reporter gene expression. Additionally, it was found that constructs harboring the VNTR polymorphism (of any variety) resulted in lower reporter gene expression than those with no repeat region; our data largely concurs with these observations. However, Inoue-Murayama and colleagues [22], who assessed the relative luciferase activities associated with the human 9-, 10- and 11-repeat alleles in addition to several non-human primate DAT1 VNTRs, reported an inverse relationship between reporter gene activity and repeat number, an observation consistent with possible length-dependent reductions in transfection efficiency, and a finding that differs from our data. Miller and Madras [26] arrived at similar findings (i.e. 9 > 10), though their analysis suggested that an additional SNP located in the 3'UTR may further regulate the VNTR's effects. Lastly, Mill [25], who recently published a well-controlled set of reporter gene analyses using both neuronal and non-neuronal cell lines, found no significant difference in reporter gene activity attributable to VNTR copy number. These discrepant studies underscore the need to examine the DAT1 VNTR effects under carefully controlled experimental conditions using the DAT protein itself as the reporter signal. The veracity of our findings is reinforced by the use of a targeted stable integration protocol that eliminates confounds to construct comparison attributable to variable transfection efficiency or clonal variance, both common pitfalls of transient transfection or conventional non-targeted stable integration approaches. Unlike typical stable integration methods, the enzymatically-mediated Flp-In system employed herein resulted in DAT-expressing cell lines with their respective constructs directionally integrated at a common, transcriptionally-active locus [39-41]. pcDNA5/FRT's hygromycin resistance gene, which was used in the isolation and expansion of cell lines for this study, lacks both a promoter and start codon, thereby preventing selection in instances of random genomic insertion. When enzymatically integrated into the host cell line, however, the gene becomes properly aligned with both a promoter and initiation sequence, permitting expression of the hygromycin resistance gene and DAT1 variant from a common site. This targeted stable integration approach eliminates the need to normalize data to account for differential transfection efficiency, a significant confound of most traditional reporter gene methodologies. The current findings are particularly interesting considering the divergent results of four in vivo human [123I] β-CIT SPECT studies examining the impact of the DAT1 VNTR polymorphism on measures of striatal DAT binding potential [27-30]. The first study, conducted in a mixed population of abstinent alcoholics and healthy controls, found that 10-repeat homozygotes displayed a 22% higher DAT binding density than 9/10 heterozygotes [29]. A similar study conducted in a population of healthy subjects revealed sharply different results, with 10-repeat homozygotes characterized by striatal DAT binding 13.4% lower than subjects carrying a 9-repeat allele [28]; van Dyck [30] recently arrived at similar findings. Martinez [27], also utilizing [123I] β-CIT, studied a mixed population of 59 healthy controls and schizophrenics and found no difference in binding potential between 10-repeat homozygotes and carriers of the 9-repeat allele. Though these studies differ widely in their conclusions, a number of factors may have contributed to the varied results. Foremost, subject populations varied from healthy individuals to those with psychiatric diagnoses characterized by probable dopaminergic dysregulation. Second, the use of [123I] β-CIT, a high affinity DAT ligand, but one lacking selectivity for the DAT over the serotonin transporter, highlights the need to conduct experiments using more selective ligands, preferably in a PET setting [42,43]. Third, two of these studies focused exclusively on the DAT1 VNTR polymorphism and did not consider other polymorphisms capable of modulating [123I] β-CIT binding (e.g. SERT promoter VNTR polymorphism known to modulate 5-HT transporter density [44]); given [123I] β-CIT's binding profile and the partially overlapping brain regional distribution of DAT and SERT, a portion of the between-study variance may be attributable to variations in genes encoding other monoamine transporters. In light of these limitations, PET imaging with highly selective DAT ligands [45] remains the best in vivo approach to assessing whether the observed in vitro effects of the DAT1 VNTR on DAT expression generalizes to the in situ DAT. Of interest are the small, though statistically significant, effects of 3'UTR variations on DAT affinity for [3H] Win 35,428. Given the location of the VNTR polymorphism, genotype-dependent variations in the DAT affinity would not be predicted. Preliminary data from our laboratory suggests that VNTR-dependent rates of DAT protein glycosylation may account for the observed variance in ligand affinity, a hypothesis subject to ongoing testing in our laboratory. While the present study provides evidence supporting the DAT1 VNTR as a functional polymorphism, the strength of this assertion is tempered by the inherent limitations of the experimental approach. First, despite the magnitude and reproducibility of the observed effects in the present in vitro study, these effects may not generalize to the in vivo condition. The association of the 10-repeat allele with an increase in the amount of in vitro DAT protein could be attenuated by in vivo counter-regulatory mechanisms including upregulation of lysosomal DAT targeting and modulation in rates of endosome-mediated DAT internalization [46-48]. The current experiments, conducted in membrane homogenates, do not discriminate between DAT proteins functionally inserted into the cell membrane and those contained within cellular vesicles. Secondly, the scope of the present study was limited to the effects of the VNTR region independent of other SNP or haplotype effects that may contribute to variability in dopaminergic function; our experiments, by design, do not reflect the recognized complexity of interactions affecting DAT1 expression [17,24]. Third, our experiments utilized the powerful CMV promoter; while the use of strong viral promoters is common in experiments such as this, possible promoter/VNTR interactions have been suggested [26]. Therefore, different effects may be observed if the CMV promoter was substituted with either the native DAT1 promoter or a weaker constitutively active sequence. Lastly, our results are limited to the effects of the DAT1 VNTR in Flp-In 293 cells, a cell line that may lack certain tissue-specific regulatory factors found in DAT rich tissues. While it would be of interest to study these effects in dopaminergic neurons in combination with Flp-In integration, such an approach would render it difficult to pharmacologically distinguish between native and transfected DAT and would significantly compromise the otherwise high degree of experimental control characteristic of the present experiments. Though the DAT1 3'UTR is lengthier than the transcript's entire coding region, is a likely target for modulating DAT1 translation, and has been associated with symptoms of ADHD, there exists the possibility of linkage disequilibrium between the DAT1 3'UTR and a separate, yet unidentified, causal genetic element. Still, these findings have interpretive significance to the often-observed association of the 10-repeat DAT1 VNTR with ADHD. ADHD is viewed as being related to a deficit in synaptic dopamine subserving task-specific signaling. An increase in DAT expression in 10-allele carriers could result in such a dopamine deficit state, depleting dopamine's ability to task-specifically increase signal to noise ratios in target neurons [49,50] and result in deficits in attention and impulse control. Methods Cloning of DAT1 variants Four DAT1 constructs were created, each containing a common promoter, coding region, and poly (A) tail, but varying with respect to the presence and length of the 3'UTR. All constructs were restriction mapped and sequence verified throughout the length of the insert prior to targeted stable transfection in Flp-In 293 cells. Though the sequence of these clones is nearly indistinguishable from that predicted by public genomic databases, they are not eligible for Genbank submission due to the mixed origin (cDNA and genomic DNA) of the constructs. Since the sequence variance in and around DAT1 is complex, we have included as supplemental material to this manuscript the complete annotated sequence of each construct (Additional file 1). The constructs, presented in order of increasing length, are depicted in Figure 1 and were generated as follows: hDAT The pRC/CMV plasmid (Invitrogen, Carlsbad, California, USA) containing the DAT1 coding region was obtained (courtesy of Dr. Marc Caron, Duke University) and the insert released via digestion with PmeI (New England Biolabs, Beverly, MA, USA). The resultant fragments were electrophoretically resolved on 1.5% agarose and the band containing the DAT1 coding region was purified using the QIAquick Gel Extraction Kit (Qiagen, Venlo, The Netherlands). Similarly, pcDNA5/FRT was digested with PmeI, the vector backbone purified (QIAquick PCR Purification Kit) and treated with shrimp alkaline phosphatase. The DAT1 insert underwent blunt end ligation into the pcDNA5/FRT backbone by overnight reaction at 14°C with T4 DNA ligase (New England Biolabs). hDAT Zero An ~800 bp segment of the DAT1 3'UTR beginning upstream of the stop codon and terminating proximal to the VNTR element was cloned from banked human DNA into pCR 2.1 TOPO using the TOPO TA Cloning Kit (Invitrogen). Oligos DAT1E15F (5'-CAACCACAGTCTCGCGGCTTT-3') and DAT1E15ZeroR (5'CTCAGGCCGTTCCCTACACC-3') were used in conjunction with Platinum Pfx DNA Polymerase (Invitrogen) to drive 35 cycles of PCR using the manufacturer's recommended protocol. Following initial amplification, 1 U of Taq polymerase was added to the reaction mixture and incubated at 72°C for 10 minutes; the resultant PCR product with 3' A-overhangs was used immediately in a TOPO-TA cloning reaction. After transforming competent bacteria and identifying clones containing the insert of interest, the fragment containing the 3'UTR was released via restriction digestion with BsmBI and XbaI (New England Biolabs), band purified, and cloned into a BsmBI/XbaI site in the pRC/CMV plasmid immediately downstream of the DAT1 coding region. The resultant clone contained an insert consisting of the DAT1 coding region flanked by a ~800 bp fragment of the 3'UTR up to, but not including, the VNTR region. This construct was subcloned into the pcDNA5/FRT plasmid via the same blunt end cloning method used in the generation of the hDAT plasmid. hDAT9 and hDAT10 The DAT1 15th exon from a previously genotyped 9/10 heterozygote was PCR amplified and cloned into pCR2.1-TOPO using a protocol similar to that used in the generation of the hDAT Zero construct. Primers DAT1E15F (sequence previously noted) and DAT1E15R (5'-AGGGACCCACACGATGCTGA-3') were used to produce a ~2100 bp PCR product that was subsequently made TOPO-TA compatible through the previously described A-overhang procedure. TOPO-TA reactions were immediately performed and used to transform competent bacteria. Ampicillin-resistant bacterial colonies were isolated and grown overnight using standard culture conditions. Plasmid DNA was isolated and genotyped to differentiate those containing the 9-repeat versus the 10-repeat VNTR via a protocol that has been previously described [4]. Following identification of clones containing the 9 and 10 repeat VNTRs, respectively, the insert containing the entire 3'UTR was released via restriction digestion with BsmBI and XbaI. The band was purified via gel extraction and cloned into a BsmBI/XbaI site in the original pRC/CMV plasmid. The resultant clones contained an insert consisting of the DAT1 coding region flanked by a ~1900 bp fragment of the 3'UTR harboring either the 9 or 10 repeat variant. The insert was subcloned into the pcDNA5/FRT plasmid via the previously described protocol. Creation and maintenance of stably transfected model cell system The Flp-In 293 host cell line (Invitrogen), an HEK-293 variant containing a Flp-In recombination site at a transcriptionally active locus, was grown in complete medium (D-MEM with 2 mM L-glutamine, 10% FBS, 1% penn/strep) supplemented with 100 μg/mL Zeocin in a 37°C incubator with 5% CO2. 48-hours prior to transfection, cells were split into 6-well plates and grown to ~80% confluence on the day of transfection and incubated in complete medium lacking Zeocin. Cells were co-transfected with one of the four DAT1 variants and pOG44 (Invitrogen), a plasmid encoding the Flp-recombinase enzyme necessary for targeted stable integration [51]. The Lipofectamine 2000 reagent was used to transfect the cells according to the manufacturer's recommended protocol (Invitrogen). Briefly, 3.6 μg pOG44 and 0.4 μg of one of the four DAT1 constructs were diluted to a 250 μl volume in Opti-MEM reduced serum media (Invitrogen); similarly, 10 μL Lipofectamine 2000 was diluted to a volume of 250 μL and incubated at room temperature for 5 minutes. The diluted DNA and Lipofectamine solutions were combined and incubated at room temperature for 20 minutes and then added to culture medium. 24-hours after transfection, cells were washed with PBS and fresh complete medium was again added. 48-hours after transfection, cells were split into fresh medium, plated on 150 mm × 25 mm cell culture dishes, and incubated at 37°C until cells attached. Medium was then removed and replaced with complete medium supplemented with the selecting antibiotic hygromycin (100 μg/mL). Selective media was replaced every 4 days until hygromycin-resistant foci were identified. Single resistant colonies were encircled with a cloning cylinder, dislodged with 0.25% trypsin, and expanded. Cells exhibiting the phenotype for proper Flp-In recombination and demonstrating the presence of a cocaine-sensitive DAT as assessed by a rapid single concentration [3H] dopamine uptake assay were selected; clones were expanded to confluence in 2-tray (1264 cm2) Nunc Cell Factories (Nalge Nunc International, Rochester, NY, USA). Once confluent, the four cell lines were harvested with 37°C PBS containing 0.53 mmol/L ethylenediaminetetraacetic acid, separated into aliquots, and centrifuged at 2000 × g for 10 min. Supernatants were decanted and the pellets were rinsed with 37°C PBS. Pellets were again centrifuged at 2000 × g for 10 min, the supernatants decanted, and the pellets stored at -80°C until the time of assay. Mock transfected cells were confirmed to have no specific DAT binding or immunoreactivity prior to initiation of assays. General radioligand binding methods Binding assays were carried out in 0.03 M phosphate buffer (pH 7.4) containing 0.32 M sucrose. Membrane suspensions were prepared by resuspending the pellet in 7 mL of ice cold assay buffer followed by homogenization with a Polytron PT3000 for 20 seconds at 20,000 rpm. All assays were performed in 12 × 75 mm polystyrene tubes in a 1,000 μL final volume consisting of 800 μL of assay buffer, 100 μL of [3H] Win 35,428 (Perkin Elmer, Boston, MA), and 100 μL of cell membrane suspension. Competition assays were performed in a final volume of 1,000 μl consisting of 700 μl assay buffer, 100 μL of [3H] Win 35,428, 100 μL of methylphenidate HCl (10-10 to 10-6 M) (Sigma-Aldrich, St. Louis, MO), and 100 μl of cell suspension. Nonspecific binding was defined via the addition of 25 μM cocaine HCl (Sigma-Aldrich). Incubations were terminated via rapid vacuum filtration through GF/B filters presoaked in assay buffer containing 0.3% polyethyleneimine and rinsed with three washes (5 mL) of ice cold assay buffer. Filters were punched, placed in scintillation vials, and equilibrated overnight in 6 mL of liquid scintillation cocktail (Ultima Gold, Packard, Meriden, CT, USA). Vials were shaken and radioactivity determined in a liquid scintillation counter at 50% efficiency for 240 sec/vial. Saturation binding assays The DAT radioligand [3H] Win 35,428 was isotopically diluted with freshly weighed cold ligand (courtesy of Dr. Michael Kuhar, Emory University) in silanized borosilicate glass tubes. Drug was initially dissolved in assay buffer containing 5 mmol/L HCl at a concentration of 1 μg/μL followed by serial dilution with assay buffer. Saturation binding assays utilized final [3H] Win 35,428 concentrations over a range of 0.5 to 75 nM. Total and nonspecific binding was determined in triplicate at each concentration of ligand. Assays were initiated by the addition of membrane suspension and incubated at 4°C for 4 hours. Competition binding assays [3H] Win 35,428 was isotopically diluted with freshly weighed cold ligand in silanized borosilicate glass tubes as previously described. The concentration used in competition assays (5.85 nM) approximated the mean Kd for [3H] Win 35,428 binding to the DAT from the binding site saturation studies. Methylphenidate was dissolved in assay buffer containing 5 mmol/L HCl at an initial concentration of 1 μg/μl. The concentrated solution subsequently underwent serial dilution to generate final concentrations of competing ligand ranging from 0.01 to 10,000 nM. Assays were initiated by the addition of membrane suspension, and incubated at 25°C for 1 hour. Protein assays At the time of each saturation binding assay, twelve 100 μL aliquots of each membrane homogenate were collected for purposes of measuring total mean protein content with a BCA protein assay (Pierce Biotechnology, Rockford, IL, USA). Protein assay results were used in the normalization of saturation binding data to account for minor differences in total added tissue. Western blots Cell pellets were homogenized in PBS in the presence of DNAse and protease inhibitors (Roche Applied Science, Indianapolis, IN, USA). Samples were then enzymatically deglycosylated (Enzymatic Protein Deglycosylation Kit, Sigma-Aldrich) prior to the loading and electrophoretic resolution of 20 μg total protein in sample buffer (40 mM Tris, pH 6.8, 0.1% SDS, 10% glycerol, 0.025% Bromphenol blue) on 10% acrylamide. DAT, a highly glycosylated protein, typically runs as a smear of high molecular weight bands; deglycosylation reduces the core protein to a mass of ~50 kD and greatly aids densitometry analysis. An example of mature, fully-glycosylated DAT and the resultant high molecular weight smear is included as supplemental material (Additional file 2) to demonstrate the significant benefit of sample deglycosylation. BioRad's Precision Plus Protein Standards (Hercules, CA, USA) were loaded into the first lane of each gel. Gels were transferred to PVDF membranes (Millipore, Billerica, MA, USA) overnight at 3 mA. Membranes were blocked for 1 hour at RT (SuperBlock Blocking Buffer, Pierce Biotechnology) on an orbital shaker and probed with a rat anti-DAT monoclonal antibody specific for the protein's N-terminal region (1:250 dilution, courtesy Dr. Allan Levey, Emory University) and rabbit anti 14-3-3β polyclonal antibody (1:500 dilution, Santa Cruz Biotechnology, Santa Cruz, CA, USA) at room temperature for 1 hour. Following rinsing, the blot was incubated with a 1:10,000 dilution of both Alexa Fluor 680 conjugated anti-rat IgG (Molecular Probes, Eugene, OR, USA) and IRDye-800 conjugated Anti-Rabbit IgG (Rockland Immunochemicals, Gilbertsville, PA, USA) for 1 h at room temperature. Blots were rinsed, dried and visualized using a LI-COR Biosciences Odyssey Infrared imager (Lincoln, NE, USA). The ubiquitously expressed 14-3-3β was used as a protein loading control. Western analyses were conducted in triplicate. Data analysis Radioligand binding data Saturation and competition binding curves were analyzed by the iterative, non-linear, curve fitting program Prism 4.0 (GraphPAD Software, Inc., San Diege, CA). Values for Kd, Bmax and Ki are expressed as mean values ± S.E.M (Table 1). A one-way ANOVA with a Holm-Sidak post-hoc analysis was used to identify statistically significant differences between VNTR variants. The threshold for statistical significance was set at p < 0.05. Table 1 In vitro pharmacological characteristics of DAT1 variants. hDAT hDAT Zero hDAT 9 hDAT 10 Bmax (± S.E.M) 7747 (± 360) 5913 (± 308) 3021 (± 286) 4669 (± 480) Kd (± S.E.M) 5.44 (± 0.55) 7.77 (± 0.41) 4.98 (± 0.20) 6.69 (± 0.57) Ki (± S.E.M) 24.77 (± 1.14) 29.13 (± 1.81) 25.00 (± 0.10) 33.39 (± 4.15) Western blot analysis Western blots underwent densitometric analysis using the program ImageJ 1.33 u (National Institutes of Health, Bethesda, Maryland, USA). The DAT signal was normalized to the 14-3-3β signal within each lane (expressed in arbitrary units), then subsequently examined for main effects of DAT1 genotype via a one way ANOVA with a Holm-Sidak post hoc comparison. The threshold for statistical significance was set at p < 0.05. Authors' contributions SHV performed all molecular biology, radioligand binding experiments, immunoassays, statistical analysis and drafted the manuscript. MJO participated in the design of the pharmacological aspects of the study, provided supervision of radioligand binding experiments, and aided in the revision of the manuscript. CDK participated in the overall design and coordination of the study and aided in the drafting of the manuscript. All authors read and approved the final manuscript. Supplementary Material Additional File 1 ZIP archive containing full length annotated sequence files (.gb format) of the four DAT1 constructs. dat1_clones_in_pcDNA5-FRT.zip (17 K) Click here for file Additional File 2 Western blot run without sample deglycosylation. Samples are variably glycosylated, creating a high molecular weight smear. Sample deglycosylation yields tight, discrete bands that are more easily analyzed. Click here for file Acknowledgements The authors would like to acknowledge Dr. Marc Caron for generously supplying the DAT1 starting material, Dr. Allan Levey and Mr. Craig Heilman for assistance with DAT western blots, Dr. Michael Kuhar for advice, and Elyse Katz and Dina Ghoneim for technical assistance. This work was supported by the National Institute of Mental Health (R21 MH067939) and by the National Institute on Drug Abuse (T32 DA015040). ==== Refs Kawarai T Kawakami H Yamamura Y Nakamura S Structure and organization of the gene encoding human dopamine transporter Gene 1997 195 11 18 9300814 10.1016/S0378-1119(97)00131-5 Giros B el Mestikawy S Godinot N Zheng K Han H Yang-Feng T Caron MG Cloning, pharmacological characterization, and chromosome assignment of the human dopamine transporter Mol Pharmacol 1992 42 383 390 1406597 Vandenbergh DJ Persico AM Uhl GR A human dopamine transporter cDNA predicts reduced glycosylation, displays a novel repetitive element and provides racially-dimorphic TaqI RFLPs Brain Research Molecular Brain Research 1992 15 161 166 1359373 10.1016/0169-328X(92)90165-8 Vandenbergh DJ Persico AM Hawkins AL Griffin CA Li X Jabs EW Uhl GR Human dopamine transporter gene (DAT1) maps to chromosome 5p15.3 and displays a VNTR Genomics 1992 14 1104 1106 1478653 10.1016/S0888-7543(05)80138-7 Mitchell RJ Howlett S Earl L White NG McComb J Schanfield MS Briceno I Papiha SS Osipova L Livshits G Leonard WR Crawford MH Distribution of the 3' VNTR polymorphism in the human dopamine transporter gene in world populations Hum Biol 2000 72 295 304 10803661 Doucette-Stamm LA Blakely DJ Tian J Mockus S Mao JI Population genetic study of the human dopamine transporter gene (DAT1) Genetic Epidemiology 1995 12 303 308 7557351 10.1002/gepi.1370120307 Kang AM Palmatier MA Kidd KK Global variation of a 40-bp VNTR in the 3'-untranslated region of the dopamine transporter gene (SLC6A3) Biol Psychiatry 1999 46 151 160 10418689 10.1016/S0006-3223(99)00101-8 Shimada S Kitayama S Lin CL Patel A Nanthakumar E Gregor P Kuhar M Uhl G Cloning and expression of a cocaine-sensitive dopamine transporter complementary DNA Science 1991 254 576 578 1948034 Wu X Gu HH Molecular cloning of the mouse dopamine transporter and pharmacological comparison with the human homologue Gene 1999 233 163 170 10375632 10.1016/S0378-1119(99)00143-2 Miller GM De La Garza RD Novak MA Madras BK Single nucleotide polymorphisms distinguish multiple dopamine transporter alleles in primates: implications for association with attention deficit hyperactivity disorder and other neuropsychiatric disorders Molecular Psychiatry 2001 6 50 58 11244485 10.1038/sj.mp.4000809 Conne B Stutz A Vassalli JD The 3' untranslated region of messenger RNA: A molecular 'hotspot' for pathology? Nat Med 2000 6 637 641 10835679 10.1038/76211 Nakamura Y Koyama K Matsushima M VNTR (variable number of tandem repeat) sequences as transcriptional, translational, or functional regulators J Hum Genet 1998 43 149 152 9747025 10.1007/s100380050059 Fuke S Sasagawa N Ishiura S Identification and characterization of the Hesr1/Hey1 as a candidate trans-acting factor on gene expression through the 3' non-coding polymorphic region of the human dopamine transporter (DAT1) gene J Biochem (Tokyo) 2005 137 205 216 15749835 Mignone F Gissi C Liuni S Pesole G Untranslated regions of mRNAs Genome Biol 2002 3 REVIEWS0004 11897027 10.1186/gb-2002-3-3-reviews0004 Cook EHJ Stein MA Krasowski MD Cox NJ Olkon DM Kieffer JE Leventhal BL Association of attention-deficit disorder and the dopamine transporter gene Am J Hum Genet 1995 56 993 998 7717410 Daly G Hawi Z Fitzgerald M Gill M Mapping susceptibility loci in attention deficit hyperactivity disorder: preferential transmission of parental alleles at DAT1, DBH and DRD5 to affected children Molecular Psychiatry 1999 4 192 196 10208453 10.1038/sj.mp.4000510 Barr CL Xu C Kroft J Feng Y Wigg K Zai G Tannock R Schachar R Malone M Roberts W Nothen MM Grunhage F Vandenbergh DJ Uhl G Sunohara G King N Kennedy JL Haplotype study of three polymorphisms at the dopamine transporter locus confirm linkage to attention-deficit/hyperactivity disorder Biological Psychiatry 2001 49 333 339 11239904 10.1016/S0006-3223(00)01053-2 Gill M Daly G Heron S Hawi Z Fitzgerald M Confirmation of association between attention deficit hyperactivity disorder and a dopamine transporter polymorphism Mol Psychiatry 1997 2 311 313 9246671 10.1038/sj.mp.4000290 Waldman ID Rowe DC Abramowitz A Kozel ST Mohr JH Sherman SL Cleveland HH Sanders ML Gard JM Stever C Association and linkage of the dopamine transporter gene and attention-deficit hyperactivity disorder in children: heterogeneity owing to diagnostic subtype and severity Am J Hum Genet 1998 63 1767 1776 9837830 10.1086/302132 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 Mol Psychiatry 2003 8 393 396 12740596 10.1038/sj.mp.4001238 Fuke S Suo S Takahashi N Koike H Sasagawa N Ishiura S The VNTR polymorphism of the human dopamine transporter (DAT1) gene affects gene expression The Pharmacogenomics Journal 2001 1 152 156 11911442 Inoue-Murayama M Adachi S Mishima N Mitani H Takenaka O Terao K Hayasaka I Ito S Murayama Y Variation of variable number of tandem repeat sequences in the 3'-untranslated region of primate dopamine transporter genes that affects reporter gene expression Neurosci Lett 2002 334 206 210 12453630 10.1016/S0304-3940(02)01125-4 Michelhaugh SK The dopamine transporter gene (SLC6A3) variable number of tandem repeats domain enhances transcription in dopamine neurons Journal of Neurochemistry 2001 79 1033 1038 11739616 10.1046/j.1471-4159.2001.00647.x Greenwood TA Kelsoe JR Promoter and intronic variants affect the transcriptional regulation of the human dopamine transporter gene Genomics 2003 82 511 520 14559208 10.1016/S0888-7543(03)00142-3 Mill J Asherson P Craig I D'Souza UM Transient expression analysis of allelic variants of a VNTR in the dopamine transporter gene (DAT1) BMC Genet 2005 6 3 15683546 10.1186/1471-2156-6-3 Miller GM Madras BK Polymorphisms in the 3'-untranslated region of human and monkey dopamine transporter genes affect reporter gene expression Mol Psychiatry 2002 7 44 55 11803445 10.1038/sj.mp.4001921 Martinez D Gelernter J Abi-Dargham A van Dyck CH Kegeles L Innis RB Laruelle M The variable number of tandem repeats polymorphism of the dopamine transporter gene is not associated with significant change in dopamine transporter phenotype in humans Neuropsychopharmacology 2001 24 553 560 11282255 10.1016/S0893-133X(00)00216-5 Jacobsen LK Staley JK Zoghbi SS Seibyl JP Kosten TR Innis RB Gelernter J Prediction of dopamine transporter binding availability by genotype: a preliminary report American Journal of Psychiatry 2000 157 1700 1703 11007732 10.1176/appi.ajp.157.10.1700 Heinz A Goldman D Jones DW Palmour R Hommer D Gorey JG Lee KS Linnoila M Weinberger DR Genotype influences in vivo dopamine transporter availability in human striatum Neuropsychopharmacology 2000 22 133 139 10649826 10.1016/S0893-133X(99)00099-8 van Dyck CH Malison RT Jacobsen LK Seibyl JP Staley JK Laruelle M Baldwin RM Innis RB Gelernter J Increased Dopamine Transporter Availability Associated with the 9-Repeat Allele of the SLC6A3 Gene Journal of Nuclear Medicine 2005 46 745 751 15872345 Kirley A Lowe N Hawi Z Mullins C Daly G Waldman I McCarron M O'Donnell D Fitzgerald M Gill M Association of the 480 bp DAT1 allele with methylphenidate response in a sample of Irish children with ADHD Am J Med Genet 2003 121B 50 54 10.1002/ajmg.b.20071 Winsberg BG Comings DE Association of the dopamine transporter gene (DAT1) with poor methylphenidate response [see comments] J Am Acad Child Adolesc Psychiatry 1999 38 1474 1477 10596245 10.1097/00004583-199912000-00006 Stein MA Waldman ID Sarampote CS Seymour KE Robb AS Conlon C Kim SJ Cook EH Dopamine transporter genotype and methylphenidate dose response in children with ADHD Neuropsychopharmacology 2005 30 1374 1382 15827573 Cheon KA Ryu YH Kim JW Cho DY The homozygosity for 10-repeat allele at dopamine transporter gene and dopamine transporter density in Korean children with attention deficit hyperactivity disorder: relating to treatment response to methylphenidate Eur Neuropsychopharmacol 2005 15 95 101 15572278 10.1016/j.euroneuro.2004.06.004 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 Sauer B Site-specific recombination: developments and applications Curr Opin Biotechnol 1994 5 521 527 7765467 10.1016/0958-1669(94)90068-X O'Gorman S Fox DT Wahl GM Recombinase-mediated gene activation and site-specific integration in mammalian cells Science 1991 251 1351 1355 1900642 Zhang K Kurachi S Kurachi K Limitation in use of heterologous reporter genes for gene promoter analysis. Silencer activity associated with the cloramphenicol acetyltransferase reporter gene J Biol Chem 2003 278 4826 4830 12473656 10.1074/jbc.M211361200 Jones JR Shelton KD Magnuson MA Strategies for the use of site-specific recombinases in genome engineering Methods Mol Med 2004 103 245 258 15542911 Kolb AF Genome engineering using site-specific recombinases Cloning Stem Cells 2002 4 65 80 12006158 10.1089/153623002753632066 Wirth D Hauser H Flp-mediated integration of expression cassettes into FRT-tagged chromosomal loci in mammalian cells Methods Mol Biol 2004 267 467 476 15269443 Kula NS Baldessarini RJ Tarazi FI Fisser R Wang S Trometer J Neumeyer JL [3H]beta-CIT: a radioligand for dopamine transporters in rat brain tissue Eur J Pharmacol 1999 385 291 294 10607889 10.1016/S0014-2999(99)00695-0 Davis MR Votaw JR Bremner JD Byas-Smith MG Faber TL Voll RJ Hoffman JM Grafton ST Kilts CD Goodman MM Initial human PET imaging studies with the dopamine transporter ligand 18F-FECNT J Nucl Med 2003 44 855 861 12791810 Du L Faludi G Palkovits M Demeter E Bakish D Lapierre YD Sotonyi P Hrdina PD Frequency of long allele in serotonin transporter gene is increased in depressed suicide victims Biol Psychiatry 1999 46 196 201 10418694 10.1016/S0006-3223(98)00376-X Goodman MM Kilts CD Keil R Shi B Martarello L Xing D Votaw J Ely TD Lambert P Owens MJ Camp VM Malveaux E Hoffman JM 18F-labeled FECNT: a selective radioligand for PET imaging of brain dopamine transporters Nucl Med Biol 2000 27 1 12 10755640 10.1016/S0969-8051(99)00080-3 Melikian HE Buckley KM Membrane trafficking regulates the activity of the human dopamine transporter J Neurosci 1999 19 7699 7710 10479674 Pristupa ZB McConkey F Liu F Man HY Lee FJ Wang YT Niznik HB Protein kinase-mediated bidirectional trafficking and functional regulation of the human dopamine transporter Synapse 1998 30 79 87 9704884 10.1002/(SICI)1098-2396(199809)30:1<79::AID-SYN10>3.0.CO;2-K Loder MK Melikian HE The dopamine transporter constitutively internalizes and recycles in a protein kinase C-regulated manner in stably transfected PC12 cell lines J Biol Chem 2003 278 22168 22174 12682063 10.1074/jbc.M301845200 Nicola SM Woodward Hopf F Hjelmstad GO Contrast enhancement: a physiological effect of striatal dopamine? Cell Tissue Res 2004 318 93 106 15503151 10.1007/s00441-004-0929-z O'Donnell P Dopamine gating of forebrain neural ensembles Eur J Neurosci 2003 17 429 435 12581161 10.1046/j.1460-9568.2003.02463.x Dymecki SM Flp recombinase promotes site-specific DNA recombination in embryonic stem cells and transgenic mice Proc Natl Acad Sci U S A 1996 93 6191 6196 8650242 10.1073/pnas.93.12.6191	16309561	PMC1325255	CC BY	2021-01-04 16:30:39	no	BMC Genet. 2005 Nov 27; 6:55	utf-8	BMC Genet	2,005	10.1186/1471-2156-6-55	oa_comm
==== Front J Negat Results BiomedJournal of Negative Results in Biomedicine1477-5751BioMed Central London 1477-5751-4-101635430210.1186/1477-5751-4-10ResearchExclusion of PINK1 as candidate gene for the late-onset form of Parkinson's disease in two European populations Schlitter Anna Melissa [email protected] Martin [email protected] Jan P [email protected] Dirk [email protected] Thomas [email protected] Joerg T [email protected] Gabriele [email protected] Department of Human Genetics, Ruhr-University Bochum, Germany2 Department of Neurology, Heinrich-Heine-University Düsseldorf, Germany3 Department of Neurology, Stavanger University Hospital, Norway4 Department of Neurology, St. Josef-Hospital, Ruhr-University Bochum, Germany2005 14 12 2005 4 10 10 10 7 2005 14 12 2005 Copyright © 2005 Schlitter et al; licensee BioMed Central Ltd.2005Schlitter 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 Parkinson's disease (PD) is the second most common neurodegenerative disorder. Recently, mutations in the PINK1 (PARK6) gene were shown to rarely cause autosomal-recessively transmitted, early-onset parkinsonism. In order to evaluate whether PINK1 contributes to the risk of common late-onset PD we analysed PINK1 sequence variations. A German (85 patients) and a Norwegian cohort (90 patients) suffering from late-onset PD were screened for mutations and single nucleotide polymorphisms (SNPs) in the PINK1 gene. Both cohorts consist of well-characterized patients presenting a positive family history of PD in ~17%. Investigations were performed by single strand conformation polymorphism (SSCP), denaturating high performance liquid chromatography (DHPLC) and sequencing analyses. SNP frequencies were compared by the χ2 test Results Several common SNPs were identified in our cohorts, including a recently identified coding variant (Q115L) in exon 1. Genotyping of the Q115L variation did not reveal significant frequency differences between patients and controls. Pathogenic mutations in the PINK1 gene were not identified, neither in the German nor in the Norwegian cohort. Conclusion Sequence variation in the PINK1 gene appears to play a marginal quantitative role in the pathogenesis of the late-onset form of PD, in German and Norwegian cohorts, if at all. ==== Body Background PD is the second most common neurodegenerative disorder after Alzheimer disease affecting more than 1% of the population by the age of 65 years. Mutations in the alpha-synuclein (PARK1), Parkin (PARK2) and DJ-1 (PARK7) gene cause fairly rare familial forms of PD characterized by an early age of onset. Mutations in the recently identified LRRK2 (PARK8) gene, especially the common mutation G2019S, occur more frequently in patients suffering from early as well as late-onset PD [1,2]. Recently, mutations in the PINK1 (PARK6) gene were shown to cause autosomal recessively transmitted early-onset parkinsonism [3,4]. The PINK1 (PTEN-induced kinase 1) gene encodes a putative protein kinase. The protein is targeted to mitochondria and shows a serine-threonine kinase domain with homology to kinases of the Ca2+/calmodulin family[3]. It appears to exert protective effects against cellular stress within mitochondria[3]. These findings link mitochondria directly to the pathogenesis of PD [3,5]. An additional link between mitochondrial dysfunction and PD is obvious via the identification of disease causing mutations in the Omi/HtrA2 gene [6]. The hypothesis of mitochondrial impairment was further emphasized by postmortem studies of PD brains [7] and observation of PD syndromes after intoxication with mitochondrial complex I inhibitors, such as MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) and rotenone [8]. Mutations in the PINK1 gene are the second most common cause of autosomal-recessively inherited early-onset PD after mutations in the Parkin gene. On the other hand, strong evidence was reported for a possible role of Parkin gene variations in the late-onset form of PD (age of onset >45 years): Parkin mutations appear to contribute to the common late-onset form and mutations, especially in exon 7 in heterozygous state, may play a role as susceptibility alleles for sporadic PD[9,10]. The question arises as to whether the PINK1 gene is also a candidate gene for late-onset forms of Parkinson's disease, similar to the suggested role of the Parkin gene. Here we report of a population-based analysis of sequence variations within the PINK1 gene. The German cohort includes 85 patients suffering from late-onset form of PD and represents a modern urban population with a genetic heterogeneity. In contrast, the Norwegian cohort represents a more homogeneous population [11] and includes 90 patients suffering from late-onset form of PD. Results All 8 coding exons of the PINK1 gene were screened for sequence variation using SSCP and sequencing analyses. Several SNPs were identified in our cohorts (L63L, Q115L, Ivs4-5A>G het, Ivs6+43C>T het, N521T, c.1783A>T). Allelic frequencies of several SNPs differed significantly between the two European cohorts confirming homogeneity of the Norwegian cohort (Table 1). Patients and controls were genotyped for the recently identified variation Q115L [12] using DHPLC analysis (Table 2). The observed frequencies did not differ significantly between patients and controls, neither in the German (p = 0.27) nor in the Norwegian cohort (p = 0.8). This screening did not reveal any disease-relevant mutation in our cohorts. Table 1 Allelic frequencies of identified SNPs in Norwegian (NW) and German (G) patient cohorts Exon Sequence variation Allelic frequencies p-value distribution is significant NW G 1 Q115L 0.072 0.035 0.13 1 L63L 0.083 0.218 0.0004 5 Ivs4-5A>G 0.017 0.091 0.002 * 6 Ivs6+43C>T 0.041 0.102 0.029 * 8 N521T 0.128 0.188 0.13 8 c.1783A>T 0.201 0.2 0.98 Table 2 Genotyping of the Q115L variation Norwegian cohort German cohort Patients (n = 90) Controls (n = 136) p Patients (n = 85) Controls (n = 210) p Q115 (Wildtype) 80 118 0.8 79 190 0.27 Q115L 7 18 6 16 L115 3 0 0 4 Discussion As recently shown, mutations in the PINK1 gene rarely cause autosomal-recessively transmitted PD [3]. Besides an early age of onset, the observed clinical symptoms in PD caused by PINK1 are similar to symptoms in idiopathic PD. In this population-based study, we investigated whether sequence variations in the PINK1 gene play a role in the late-onset form of PD. A recently described variation (Q115L) of the PINK1 gene [12] was identified in the German and the Norwegian cohorts. We calculated allele frequencies in cases and controls and show here that the Q115L variant was not associated with late-onset PD in our study. These findings correspond to previously published data of no leading association of other coding SNPs within the PINK1 gene and PD [13]. In addition, several common SNPs were identified. We did not find any pathogenic mutation of the PINK1 in our cohorts composed of patients suffering from late-onset form of PD. The risk to miss potential mutations was minimized by a well established and optimized SSCP analyses. The most likely reason to explain the absence of mutations in our cohorts is the lack of influence of the PINK1 gene in the pathogenesis of late-onset PD. Individual tagging SNPs and tag-defined haplotypes in 500 PD patients likewise did not reveal associations with PD [14]. Future investigations should include screening of potential promoter as well as enhancer/silencer regions of the gene to finally exclude any lack of influence of PINK1 variation on PD manifestation. Yet, functional investigations of the PINK1 protein are necessary to identify potential interaction partners as candidates for additional mutation screening. Conclusion Investigations of other genes involved in the mitochondrial pathway of the PINK1 gene are necessary to evaluate the exact role of mitochondrial impairment for common forms of PD. Materials and methods Patients and controls The Norwegian cohort (n = 90) consists of patients suffering from late-onset PD (median age of onset 64.4 years, range from 49 to 78 years, standard deviation 7.9) originated from the Stavanger area of Western Norway. This population is known to be genetically quite homogeneous [11] and has previously been described in several clinical PD studies [15,16]. 16.7% of the patients presented a positive family history for PD concerning first degree relatives (siblings or parents). All patients meet the criteria for PD [15,17] and were thoroughly clinically examined. An ethnically matched control group of healthy blood donors was recruited in Bergen, Norway. The German cohort (n = 85) consists of patients of the Ruhr area suffering from late-onset PD (median age of onset 58.7 years, range from 45 to 79 years, standard deviation 8.7) diagnosed according to the UK Brain Bank criteria [18]. 16.5% of the patients presented a positive family history for PD concerning first degree relatives (siblings or parents). Ethnically matched control samples from senior healthy blood donors (median age 57.2 years, range from 42 to 68 years, standard deviation 5.7) were recruited at the neighbouring University Hospital of Essen (Germany). Population stratification was excluded for the controls by multiple microsatellites analyses. After receiving informed consent from the patients, peripheral blood samples were taken and genomic DNA was extracted following standard protocols. German (Bochum and Düsseldorf) and Norwegian (Bergen) ethics committees approved this study. SSCP, DHPLC, sequencing The 8 coding exons of the PINK1 gene were amplified by polymerase chain reaction (PCR) in all patients using designed primer pairs adapted to the SSCP technique (Table 3). SSCP analysis according to standard procedure [19] was used to identify mutations and SNPs. In order to optimize mutation screening by SSCP analyses, PCR products were digested with different restriction enzymes depending on the lengths of their fragments [19] and screened in two different conditions. Selected samples with band shifts evidenced in SSCP analyses were confirmed by direct sequencing. The sequence reactions were run on an automated DNA sequencer (Applied Biosystems 377 XL, Foster City, USA) and analyzed with the ABI Prism™ 377 XL collection and convenient sequencing analysis software. SNP frequencies of the Q115L variation in patients and controls were determinated by using DHPLC analyses (WAVE® system, Cheshire, UK, using software Wavemaker 4.1) according to established procedures. Table 3 Primers for PINK1 gene analysis Exon Primer sequence Product size (bp) Ex 1 F 5'-AAGTTTGTTGTGACCGGCG-3' R 5'-CTTAGCTCCGTCCTCCGCT-3' 507 Ex 2 F 5'-CCTTCCTAGGCTCCCTGGC-3' R 5'-AAGATGGGCATTTTGAGAACATCT-3' 387 Ex 3 F 5'-GCTTACAAGGAACTTACCATTCTGC-3' R 5'-GTGCTGAGGACATAAGTGATGGAT-3' 240 Ex 4 F 5'-GATGTATCAGCTCCAGGCCCT-3' R 5'-TATTCTTTCCAGGTGTTGTATCTGATG-3' 286 Ex 5 F 5'-AAACGTATTGGGAGTCGTCGA-3' R 5'-CTCTAGTGCCCCTGGAGAGCT-3' 266 Ex 6 F 5'-CGAGTCTCCTGCATTCAGTGG-3' R 5'-GACATAGCAGGGCCTCTCAGAG-3' 265 Ex 7 F 5'-TCAGGTGATGTGCAGGACATG-3' R 5'-CAGAGGTTTCTACCCACACCG-3' 358 Ex 8 F 5'-GGACCAGAGAAGGGAAGACCC-3' R 5'-TCACGACACAGAGGATGCCA-3' 410 Statistical analyses SNP frequencies were compared by the χ2 test. We considered P-values < 0.05 as significant. Authors' contributions AMS carried out the molecular genetic studies, performed the statistical analysis and drafted the manuscript. MK participated in devising the study based on thoroughly clinical analysis of the patients. JPL supervised data collection and diagnosis of the Norwegian cohort. DW and TM provided the samples and performed clinical diagnostics of the German patient group. JTE conceived of the study, and participated in its design and coordination and helped to draft the manuscript. GD supervised AMS, especially the molecular studies. All authors read and approved the final manuscript. Acknowledgements We sincerely thank all the participants in this study. We also thank Dr. O.-B. Tysnes (Department of Neurology, Haukeland University Hospital, Bergen, Norway) for providing control samples. AMS gratefully acknowledges an Alma and Heinrich Vogelsang Foundation fellowship. ==== Refs Gilks WP Abou-Sleiman PM Gandhi S Jain S Singleton A Lees AJ Shaw K Bhatia KP Bonifati V Quinn NP Lynch J Healy DG Holton JL Revesz T Wood NW A common LRRK2 mutation in idiopathic Parkinson's disease Lancet 2005 365 415 416 15680457 Di Fonzo A Rohe CF Ferreira J Chien HF Vacca L Stocchi F Guedes L Fabrizio E Manfredi M Vanacore N Goldwurm S Breedveld G Sampaio C Meco G Barbosa E Oostra BA Bonifati V A frequent LRRK2 gene mutation associated with autosomal dominant Parkinson's disease Lancet 2005 365 412 415 15680456 Valente EM Abou-Sleiman PM Caputo V Muqit MM Harvey K Gispert S Ali Z Del Turco D Bentivoglio AR Healy DG Albanese A Nussbaum R Gonzalez-Maldonado R Deller T Salvi S Cortelli P Gilks WP Latchman DS Harvey RJ Dallapiccola B Auburger G Wood NW Hereditary early-onset Parkinson's disease caused by mutations in PINK1 Science 2004 304 1158 1160 15087508 10.1126/science.1096284 Rogaeva E Johnson J Lang AE Gulick C Gwinn-Hardy K Kawarai T Sato C Morgan A Werner J Nussbaum R Petit A Okun MS McInerney A Mandel R Groen JL Fernandez HH Postuma R Foote KD Salehi-Rad S Liang Y Reimsnider S Tandon A Hardy J St George-Hyslop P Singleton AB Analysis of the PINK1 gene in a large cohort of cases with Parkinson disease Arch Neurol 2004 61 1898 1904 15596610 10.1001/archneur.61.12.1898 Shen J Cookson MR Mitochondria and dopamine: new insights into recessive parkinsonism Neuron 2004 43 301 304 15294138 10.1016/j.neuron.2004.07.012 Strauss KM Martins LM Plun-Favreau H Marx FP Kautzmann S Berg D Gasser T Wszolek Z Muller T Bornemann A Wolburg H Downward J Riess O Schulz JB Kruger R Loss of function mutations in the gene encoding Omi/HtrA2 in Parkinson's disease Hum Mol Genet 2005 Beal MF Mitochondria, oxidative damage, and inflammation in Parkinson's disease Ann N Y Acad Sci 2003 991 120 131 12846981 Dauer W Przedborski S Parkinson's disease: mechanisms and models Neuron 2003 39 889 909 12971891 10.1016/S0896-6273(03)00568-3 Foroud T Uniacke SK Liu L Pankratz N Rudolph A Halter C Shults C Marder K Conneally PM Nichols WC Parkinson Study G Heterozygosity for a mutation in the parkin gene leads to later onset Parkinson disease Neurology 2003 60 796 801 12629236 Oliveira SA Scott WK Martin ER Nance MA Watts RL Hubble JP Koller WC Pahwa R Stern MB Hiner BC Ondo WG Allen FHJ Scott BL Goetz CG Small GW Mastaglia F Stajich JM Zhang F Booze MW Winn MP Middleton LT Haines JL Pericak-Vance MA Vance JM Parkin mutations and susceptibility alleles in late-onset Parkinson's disease.[see comment] Annals of Neurology 2003 53 624 629 12730996 10.1002/ana.10524 Borg A DAHKMLHEMP BRCA1 1675delA and 1135insA account for one third of Norwegian familial breast-ovarian cancer and are associated with later disease onset than less frequent mutations Dis Markers 1999 Oct;15(1-3):79-84. Klein C Djarmati A Hedrich K Schafer N Scaglione C Marchese R Kock N Schule B Hiller A Lohnau T Winkler S Wiegers K Hering R Bauer P Riess O Abbruzzese G Martinelli P Pramstaller PP PINK1, Parkin, and DJ-1 mutations in Italian patients with early-onset parkinsonism Eur J Hum Genet 2005 15970950 Groen JL Kawarai T Toulina A Rivoiro C Salehi-Rad S Sato C Morgan A Liang Y Postuma RB St George-Hyslop P Lang AE Rogaeva E Genetic association study of PINK1 coding polymorphisms in Parkinson's disease Neurosci Lett 2004 372 226 229 15542245 10.1016/j.neulet.2004.09.043 Healy DG Abou-Sleiman PM Ahmadi KR Muqit MM Bhatia KP Quinn NP Lees AJ Latchmann DS Goldstein DB Wood NW The gene responsible for PARK6 Parkinson's disease, PINK1, does not influence common forms of parkinsonism Ann Neurol 2004 56 329 335 15349859 10.1002/ana.20206 Kurz M Alves G Aarsland D Larsen JP Familial Parkinson's disease: a community-based study European Journal of Neurology 2003 10 159 163 12603291 10.1046/j.1468-1331.2003.00532.x Tandberg E Larsen JP Nessler EG Riise T Aarli JA The epidemiology of Parkinson's disease in the county of Rogaland, Norway Movement Disorders 1995 10 541 549 8552103 10.1002/mds.870100503 Larsen JP Dupont E Tandberg E Clinical diagnosis of Parkinson's disease. Proposal of diagnostic subgroups classified at different levels of confidence Acta Neurologica Scandinavica 1994 89 242 251 8042440 Hughes AJ Daniel SE Kilford L Lees AJ Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico-pathological study of 100 cases J Neurol Neurosurg Psychiatry 1992 55 181 184 1564476 Jaeckel S Epplen JT Kauth M Miterski B Tschentscher F Epplen C Polymerase chain reaction-single strand conformation polymorphism or how to detect reliably and efficiently each sequence variation in many samples and many genes Electrophoresis 1998 19 3055 3061 9932793 10.1002/elps.1150191802	16354302	PMC1325256	CC BY	2021-01-04 16:24:35	no	J Negat Results Biomed. 2005 Dec 14; 4:10	utf-8	J Negat Results Biomed	2,005	10.1186/1477-5751-4-10	oa_comm
==== Front Immun AgeingImmunity & Ageing1742-4933BioMed Central London 1742-4933-2-171631647010.1186/1742-4933-2-17ReviewMelatonin, immune function and aging Srinivasan V [email protected] GJM [email protected] DP [email protected] AI [email protected] SR Pandi [email protected] SC [email protected] Department of Physiology, School of Medical Sciences, University Sains Malaysia 16150, Kubang Kerian, Kelantan, Malaysia2 Center for Experimental Pathology, Cantonal Institute of Pathology, Via In Selva 24, PO Box 660, Locarno, Switzerland3 Departamento de Fisiología, Facultad de Medicina, Universidad de Buenos Aires, 1121 Buenos Aires, Argentina4 Departamento de Bioquímica y Biología Molecular, Facultad de Medicina, Universidad Complutense, 28040, Madrid, Spain5 Comprehensive Center for Sleep Medicine, Department of Pulmonary, Critical Care and Sleep Medicine, Mount Sinai School of Medicine, 1176 - 5th Avenue, 6th Floor, New York, NY 10029, USA6 Department of Anatomy and Cell Biology, Strathcona Anatomy & Dentistry Building, McGill University, Montreal, PQ, H3A 2B2, Canada2005 29 11 2005 2 17 17 19 7 2005 29 11 2005 Copyright © 2005 Srinivasan et al; licensee BioMed Central Ltd.2005Srinivasan 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. Aging is associated with a decline in immune function (immunosenescence), a situation known to correlate with increased incidence of cancer, infectious and degenerative diseases. Innate, cellular and humoral immunity all exhibit increased deterioration with age. A decrease in functional competence of individual natural killer (NK) cells is found with advancing age. Macrophages and granulocytes show functional decline in aging as evidenced by their diminished phagocytic activity and impairment of superoxide generation. There is also marked shift in cytokine profile as age advances, e.g., CD3+ and CD4+ cells decline in number whereas CD8+ cells increase in elderly individuals. A decline in organ specific antibodies occurs causing reduced humoral responsiveness. Circulating melatonin decreases with age and in recent years much interest has been focused on its immunomodulatory effect. Melatonin stimulates the production of progenitor cells for granulocytes-macrophages. It also stimulates the production of NK cells and CD4+ cells and inhibits CD8+ cells. The production and release of various cytokines from NK cells and T-helper lymphocytes also are enhanced by melatonin. Melatonin presumably regulates immune function by acting on the immune-opioid network, by affecting G protein-cAMP signal pathway and by regulating intracellular glutathione levels. Melatonin has the potential therapeutic value to enhance immune function in aged individuals and in patients in an immunocompromised state. ==== Body Introduction Aging is a complex physiological process that involves a number of biochemical reactions, with molecular changes that are manifested in single cells as well as in the whole organism. Aging reflects the sum total of all changes that occur in living organisms with the passage of time that lead to functional impairment and increased pathology. Aging is characterized by a diminished ability to respond to stress [1]. Among the many theories proposed for aging, the Oxidative Theory of Aging put forth by Harman in 1956 [2] has received wide support. Aging is associated with a decline in immune function known as immunosenescence. This situation implies increased susceptibility to infectious diseases and cancer due to a decreased capacity of the immune system to respond to antigenic stimulation [3]. This results in altered cytokine microenvironment and impairment of both innate and adaptive immunity [4]. It is interesting to note that many hormones that are associated with maintenance of immune function also decline with advancing age and the interrelationship between the endocrine system and the immune system is considered of crucial importance in normal human physiology and in mediating age-associated degenerative diseases [5-8]. The decline in the production of a number of hormones associated with aging such as growth hormone (GH), estrogen and dehydroepiandrosterone, as well as of the pineal substance melatonin, have been proposed to play a significant role in contributing to immunosenesecence [5]. Among these, melatonin has been demonstrated to bear a general immunoenhancing effect in many animal species as well as in humans [9]. Melatonin is a natural antioxidant with significant anti-aging properties [10]. Indeed, any search for a therapeutic agent that can improve the quality of life in the elderly implies the identification of substances that have both antioxidant and immunoenhancing qualities. In this vein, the role for melatonin has been put forth [11-13] and in this paper the evidence indicating that melatonin is effective to combat age associated decline in immune function will be reviewed with the aim of advocating melatonin as a possible therapeutic agent for enhancing the quality of life in the elderly. Aging and immune function Immunosenescence is associated with increased incidence of cancer and of degenerative and infectious diseases. The progressive functional T cell and B cell deficits may be the main responsible factors for age-associated disorders [4,14,15]. The involution of thymus with age results in alterations of gene expression [16]; indeed, immunosenescence is reflected at cellular, molecular and genetic levels [17]. Individuals of the same chronological age may exhibit variations in the degree of senescence associated functional impairment [18]. The role of immunity as a predictor of individual longevity in human beings has been suggested by several studies like OCTO and NONA longitudinal studies and they all reveal the existence of "immunological risk phenotype", that can predict the life span in the elderly [19]. Aging and innate immunity Aging affects the innate immune system [20]. In the innate immune system natural killer (NK) cells play an important role for inhibiting cancer and metastases. Longer life in centenarians has been associated with increased NK cell number, augmented interferon (IFN)-gamma production and phagocytosis [21-23]. The age-associated increases in NK cells (21) have been interpreted as a compensatory response to overcome the generally decreased immune function and has been considered helpful in arresting the growth of neoplastic cells. For example, in human NK cells from healthy subjects over 90 years of age, the ability to synthesize chemotactic cytokines upon stimulation by IL-12 or IL-2, or to express the corresponding chemokine receptors are maintained (24). However most investigators are of the opinion that functional competence of individual human NK cells declines with age [21,25]. Indeed, NK cells of aged people exhibited a diminished production of IFN gamma and chemokines in response to interleukin (IL)-2 and IL-6 [25]. Recently, Albright and her coworkers [26] found severe impairment in the production of mRNA transcripts representing several cytokines in NK/LAK cells of aged mouse. The cytotoxic capacity of NK cell is well preserved in peripheral blood of the centenarians [27]. Functional impairment of macrophages and granulocytes are reported in the elderly. Diminished intracellular phagocytic activity, degranulation and decrease in chemotactic and phagocytic activity have all been found in polymorphonuclear leukocytes of elderly individuals [24,28]. In a study in centenarians, Miyaji and his coworkers [22] found that granulocytes exhibited decreased superoxide production, irrespective of subject's health conditions. A decrease in superoxide production in elderly subjects has also been reported in other studies [29-31], the decreased production of superoxide in the granulocytes being attributed to the reduction in signal transduction in granulocytes [29]. The attenuation of Fc mediated superoxide generation and phagocytosis in the elderly has been suggested as the major factor for the age-related decline in neutrophil function [28,32]. With regard to macrophages, increased production of proinflammatory mediators like IL-1, IL-6 and IL-8 occurs in both healthy aged subjects and people showing pathological aging [33,34]. Macrophages are important for phagocytosis and destruction of microorganisms and also for cytokine production that regulates the functional ability of other cells of innate immunity. Diminished IL-1 levels and diminished generation of reactive oxygen species (ROS) from monocytes of elderly subjects has been reported (reviewed by [35]). IL-6 (which has been termed as a "cytokine for gerontologists", [36]) increases in aged subjects [37,38]. The increase in IL-6 occurs in healthy individuals older than 85 years of age [39]. The increase in IL-6 seen in aged subjects may contribute to age-ssociated diseases [40] and mortality [41]. Plasma concentrations of soluble intercellular adhesion molecule-1 (ICAM-1) increased with age [39,42,43]. Collectively, the results suggest that it is this shift in cytokine profile that is largely responsible for triggering immunosenescence and increased morbidity and mortality in the elderly [39]. Aging and humoral immunity Aging results in changes in humoral immunity such as an increase in the levels of serum immunoglobulins like IgA and IgG, and decrease in the number of B and T lymphocytes [44,45]. A decline in organ specific autoantibodies together with an increase in non-organ specific autoantibodies have been found in the elderly [46]. Reduction in CD 27+ memory B cells has been reported and this correlated with low T cell number [47]. A decrease in CD5+ B cells independent of T cell decline was also reported in aging [48]. Therefore, the reduced humoral responsiveness and altered antibody-mediated defense mechanisms seen in aged individuals are explained mainly by an intrinsic primary cell deficit [49]. The ability of T cells to promote B-cell activation and antibody production may be compromised in elderly individuals, as suggested by studies using cytometric phenotypic analysis [50]. A significant decrease in IL-2 production with aging plays a role in reducing antibody production [23,49]. Aging and cellular immunity Aging not only causes changes in innate immunity and humoral immunity, but also causes changes in cellular immunity. A significant decrease in CD3+, CD4+, CD8+ cells and naïve T lymphocytes (CD45RA+CD4+) occurs with increase in age. An extensive review on T cell function in aging was published by Pawelec et al. [14]. With aging, alterations in signal transduction may also occur. The age-associated decline in T cell function is preceded by involution of the thymus [35]. The striking feature of T cell alterations in aging is the marked shift from naive to memory cells with an imbalance of virgin and memory cells being noted especially in CD8+ T lymphocytes [45]. Naïve T cells, which are concerned with the mounting primary immune response, are dependent upon CD28, a co-stimulatory signal for their proliferation [45]. Both the decrease in the number of naïve T cells and in their responsiveness with aging cause the decline of specific immunization response in aged individuals [51]. Large increases in CD8+ T cells with receptors for single epitopes of cytomegalovirus are common in the elderly [52]. Longitudinal studies (OCTO) suggest that the cluster of immune parameters like low CD4+ cells, an increase in CD8+ cells and a low IL-2 production are all predictive of mortality [53-55]. The decline in naïve T cells is one of the factors that cause a decreased IL-2 production[56]. Melatonin Melatonin (N-acetyl-5-methoxytryptamine) is formed mainly in the pineal gland of most mammals including man [57]. In the pineal gland, serotonin is converted in to melatonin through a two-step enzymatic process involving N acetylation followed by O-methylation. In humans, plasma melatonin level begins to increase steadily after 1900 h to 2300 h to attain the peak values at around 0200 – 0400 h [7]. The study of plasma melatonin among subjects of different age groups reveals a consistent decrease as aging progresses. With some exceptions [58,59] the decline of melatonin with age has been repeatedly reported [60-66]. The melatonin day/night rhythm has been found altered with phase advance in the elderly as compared to young women [67]. Great variations in the amplitude of nocturnal melatonin secretions are found among individuals suggesting that some individuals produce significantly less melatonin during lifetime than others; this may have an impact in terms of aging [7,68]. The loss of amplitude of melatonin rhythm in the advanced age is both an indication as well as a cause of age-related disturbances in the circadian pacemaker leading to chronobiological disorders [69]. This is accompanied by a general deterioration of cognitive, psychological and social functioning as well as by sleep disturbances [70-72]. The age-related impairment of the immune system first appears around 60 years of age coinciding with the decrease of plasma melatonin concentration. Indeed, melatonin has a defined immunomodulatory role both in animals and humans [13,73]. The diurnal and seasonal changes in the immune system have been shown to correlate with melatonin synthesis and secretion [74]. Melatonin is synthesized by human lymphocytes and this finding adds further support to the hypothesis that melatonin plays a role in the regulation of human the immune system [75]. Melatonin receptors Melatonin exerts its many physiological actions by acting on membrane and nuclear receptors although many of its actions are receptor-independent (e.g., scavenging of free radicals, interaction with cytosol proteins like calmodulin). The two melatonin receptors cloned (MT1 and MT2) are membrane receptors that have seven membrane domains and belong to the superfamily of G-protein coupled receptors [76]. Melatonin receptor activation induces a variety of responses that are mediated both by pertussis-sensitive and insensitive G proteins [77]. In the cytosol melatonin interacts with calmodulin [78]. Nuclear binding receptors have been identified in human lymphocytes and monocytes [79]. Melatonin and immune function In recent years much attention has been devoted to the possible interaction between melatonin and the immune system [13,73,80]. Melatonin has significant immunomodulatory roles in immunocompromised states. In 1986, Maestroni et al. first showed that inhibition of melatonin synthesis causes inhibition of cellular and humoral responses in mice [81]. Mice kept under constant light, or receiving injections of betaadrenergic blockers (propranolol) to inhibit melatonin synthesis, exhibited an inability to mount a primary antibody response to sheep red blood cells (SRBC), a decreased cellularity in thymus and spleen and a depressed autologous mixed lymphocyte reaction; all these were reversed by melatonin administration at the late afternoon [81]. Late afternoon injection of melatonin increases both the primary and secondary antibody responses to SRBC [82]. Indeed, the immunoenhancing effect of melatonin was evident only when melatonin was administered in the afternoon or in the presence of T-dependent antigenic stimulation. Since melatonin was ineffective in vitro, Maestroni and co-workers concluded that it exerts its immunostimulating effect through other neuroendocrine mechanisms in antigen-activated cells [83]. Hamsters exposed to short photoperiods had increased spleen weight and number of splenic lymphocytes and macrophages [84]. A key finding-albeit in young adult humans – with respect to the interplay of melatonin and the immune system, was the observation that the nocturnal rise of blood melatonin in humans correlated with the increase of thymic production of peptides like thymosin-1 alpha and thymulin [85]. Melatonin and innate immunity A number of studies support the immunoregulatory action of melatonin on the body's innate immunity [80]. Melatonin stimulates the production of progenitor cells for granulocytes and macrophages (GM-CFU) and has a general stimulatory action on hemopoiesis [86,87]. Melatonin receptors are detectable in monocyte/macrophage lineage [79] and melatonin binding to these receptors stimulates the production of GMCFU cells [88,89]. A recent pivotal study, although carried out in young adult mice, has revealed a profound, time-dependent influence of melatonin on certain cells fundamentally important to the immune system. Exogenous melatonin augments NK cells and monocytes in both the bone marrow and the spleen with a latency of 7 to 14 days [90]. As both these cells are components of the non-specific immune system, the findings suggest that melatonin could be an effective way for arresting neoplastic growth and for destroying virus infected cells. The action of melatonin on monocyte production can be partly due to its direct action on melatonin receptors or may be due to an increase of monocyte sensitivity to stimulants like IL-3, IL-4, IL-6 or GM-colony stimulating factor (GM-CSF) [88-90]. As stromal cells contain receptors for kappa opioid cytokine peptides, melatonin-induced release of opioid peptides from these stromal cells in bone marrow could be involved in the regulation of hemopoietic cell proliferation [91]. In addition to monocytes, the bone marrow precursor cells for the granulocyte lineage increase in absolute numbers after melatonin administration. The study of Currier et al. [90] revealed that melatonin increases the actual production of the GM-cell lineage and not the inter-organ trafficking of myeloid precursors. An increased activation of monocytes/macrophages by melatonin has been reported in yet another study in rodents [92]. As both macrophage cells and neutrophils form important components of the innate immune system, the stimulatory action of melatonin reflects a significant immunoenhancing property. Melatonin treatment restores the decreased total leukocyte count in peripheral blood and bone marrow of pinealectomized squirrels [93]. Macrophages have been shown to form large amounts of nitric oxide (NO) upon activation by ROS that mediate their microbiocidal properties. This excessive production of NO can be harmful to the body as it can result in the development of degenerative diseases [94]. In a recent study melatonin was found to decrease NO concentration in macrophages by suppressing inducible NO synthase expression [95]. When melatonin's effects on phagocytic activity of macrophages were tested at different concentrations, the greatest phagocytic stimulation was obtained when melatonin was added resembling the unstressed situation [96]. NK cells play an important role in immunosurveillance against neoplasia and virus infected cells [97,98]. IFN-gamma enhances NK cell activity [99]. An observation of potentially high prophylactic significance, was the demonstration that exogenous melatonin given acutely at 1800 h to young healthy males increased their responsiveness to IFN while the chronic administration of melatonin augmented the spontaneous NK cell activity and also the circulating number of NK cells [100]. The increased NK cell number brought about by melatonin administration was attributed partly to the increased production of cytokines by melatonin-stimulated T helper cells. IL-2, IL-6, IL-12 and IFN-gamma have all been suggested as the possible cytokines that mediate melatonin-induced increase of NK cell number [90]. T helper cells contain melatonin receptors that presumably mediate melatonin action in releasing cytokines [101-103]. Melatonin and cytokine production Melatonin has been proposed to regulate the immune system by affecting cytokine. production in immunocompetent cells [104]. Melatonin enhances the production of IL2, IFN-gamma and IL-6 by cultured human mononuclear cells [101]. Melatonin, by. activating monocytes [105], increases the production of IL-1, IL-6, TNF-alpha and ROS. Melatonin also increases IL-12 production by monocytes [105]. Repeated stimulation of T helper (Th) cells in the presence of IL-12 causes Th cells to differentiate into Th1 cells, which produce IL-2 and IFN-gamma and are particularly effective in enhancing immune responses that involve macrophages and other phagocytes. Melatonin augments IFN-gamma production by Th1 cells [104]. The enhancement of NK cell activity by melatonin is attributed to the increased production of IL-2 and IL-12 [104,106,107]. Human lymphocytes themselves play an important role in stimulating IL-2. production in an autocrine or paracrine fashion [75]. After melatonin treatment, up-regulation of gene expression for TGF-β, M-CSF, TNF-α, and stem cell factor (CSF) in peritoneal exudate cells, and the level of gene expression of IL-1β, M-CSF, TNF-α, IFN-γ, and SCF in splenocytes were reported [108]. Melatonin's immunoenhancing effect depends upon its ability to enhance the production of cytokines as well as its anti-apoptotic and antioxidant action. As a functional impairment of macrophages and granulocytes (as shown by the diminished intracellular phagocytic activity, degranulation and decrease in chemotactic activity) has been reported in the elderly [28,44] and a parallel decrease in melatonin production occurs [60-66] it may not be unreasonable to speculate that immunosenescence can be partly attributed to a decreased production of melatonin. To restore the defective phagocytic function the use of adjuvants with immunizations and nutritional supplements has been proposed [109]. Micronutrients like zinc, selenium and vitamin E play a vital role in phagocytic function [110]. Since melatonin can stimulate the immune response and correct immunodeficiencies by causing up-regulation of cytokine production it can be used therapeutically for correcting the immunodeficiency state associated with aging. Melatonin and cellular and humoral immunity Besides its stimulatory action on the production of several cytokines that regulate immune function, melatonin's immunoenhancing properties have been attributed to a direct action on the immunocompetent cells (e.g. granulocyte-macrophage cells, NK cells and lymphocytes). Earlier studies demonstrated that the thymus is a primary target of melatonin's action. The thymus is an organ of youth in mammals, yet any influences on the thymus in youth will have profound effects on the immune system of elderly mammals. A milestone, earlier demonstration revealed that pinealectomized, young mice underwent accelerated involution of the thymus [111]. The presence of melatonin binding sites in membrane preparations of non-mammalian (duck) thymus has also been reported [112]. Mice kept under constant light, or administered with beta-adrenergic blockers exhibited decreased cellularity of thymus and spleen that was reversed by late afternoon administration of exogenous melatonin [81,82,113]. The severe loss of thymocytes with age is the main cause of structural thymic atrophy and thymic weight loss. Melatonin administration increased the total number of thymocytes in old mice [114]. In that study, thymic cell number in 2 months-old mice was 12.6 × 107, while it dropped to 7.3 × 107 cells in 24 months-old animals; in melatonin treated old mice the total number of thymocytes was 9.1 × 107 cells [114]. This protective effect of melatonin on thymocytes was attributed to its antiapoptotic action. Melatonin inhibited glucocorticoid- or hydroxyl radical-induced thymic apoptosis [115,116]. The reversal of age-associated thymic involution by melatonin added further support to the concept that melatonin can be a potential therapeutic agent for correcting immunodeficiency state associated with aging and possibly other immunocompromised states like severe stress [117]. Finally, Yu et al. [118] have demonstrated that orally administered melatonin can substantially promote the survival (anti-apoptosis) of precursor B lymphocytes (responsible for humoral immunity) in the B lymphocyte generating site, i.e., the bone marrow. This indicates that melatonin treatment can boost the survival of mature B cells which are the functional elements in humoral immunity. Melatonin and T lymphocyte function Melatonin enhances both cell-mediated and humoral immunity. The administration of melatonin to normal or immunocompromised mice elevated in vitro and in vivo antibody responses [73]. The immunoenhancing effect of melatonin involves opiod peptides; melatonin stimulates Th cells to secrete opiod peptides that have upregulatory effects on a variety of immune cells [73]. According to Nelson and Drazen [119], melatonin is a part of a complex physiological system that coordinates reproductive, immunological and other physiological processes to cope up with energetic stressors during winter. Studies in birds also indicate that melatonin stimulates both cellular and humoral responses and that the response involves opiate intermediates [120,121]. The immunostimulatory role of melatonin is exerted mainly on Th cells and on T lymphocyte precursors. There is a possibility that melatonin could act as an autocoid in bone marrow as shown by the demonstration of melatonin synthesis in bone marrow cells of mice and humans [122]. The existence of specific melatonin binding sites inlymphoid cells provides evidence for a direct effect of melatonin in the regulation of the immune system. By using the melatonin agonist 2 [125I]-melatonin high affinity binding sites and a signal tranduction pathway for melatonin have been characterized in human lymphocytes [123,124]. Melatonin also counteracted the inhibitory effect of prostaglandin E2 on IL-2 production in human lymphocytes via its MT1 membrane receptor [125]. Melatonin augments CD4+ lymphocytes and decreases CD8+ lymphocytes in rat submaxillary lymph nodes [126]. Collectively, these studies indicate that melatonin possesses important immunoenhancing properties and suggest that melatonin may favor a Th-1 response. During the natural history of human immunodeficiency virus type I (HIV-1) infection, an impairment of IL-12 production precedes a switch from a Th-1 to a Th-2 stage of cellular immunity. A recent study indicated a correlation of serum levels of melatonin and IL-12 in a cohort of 77 HIV-1 infected individuals, the decreased levels of serum melatonin found in HIV-1-infected individuals being possibly instrumental in the impairment of Th-1 immune response [127]. Besides the release of proinflammatory Th-1 cytokines, such as IFN-gamma and IL2 administration of melatonin to antigen-primed mice increased the production of IL10, indicating that melatonin can also activate anti-inflammatory Th-2-like immune responses in certain circumstances [128]. Therefore, it is not yet clear whether melatonin acts only on Th-1 cells or also affects Th-2 cells. This is an important subject as the Th-1/Th-2 balance is significant for the immune response [73]. Relevant to this, melatonin treatment suppressed the subsequent in vitro stimulation by the mitogenic agents LPS (that stimulates B cells) and Con A (that stimulates T cells) in submaxillary lymph nodes [126]. In addition, an inhibitory influence of melatonin on parameters of the immune function has also been demonstrated, i.e., in human NK cell activity, DNA synthesis, IFN-gamma and TNF-alpha synthesis, as well as the proliferation of T lymphocytes and lymphoblastoid cell lines were depressed by melatonin [73]. Melatonin can correct immunodeficiencies secondary to acute stress, viral diseases and drug treatment. In immunodepressed conditions, the immunoenhancing action of melatonin seems to be restricted to T lymphocyes [129]. In conditions of immunodeficiency, as in other pathologies and the normal, melatonin appears to favour a Th1 lymphocyte response [108]. Finally, a recent study (130) has estabished a significant role for melatonin, i.e., as an adjuvant with vaccination in sheep afflicted with ovine footrot, indicating that this agent clearly has significant benefits in health maintenance and disease treatment. Mechanism of action of melatonin in immune responses Studies by Drazen and Nelson [102] indicated that melatonin receptor subtype MT2 but not MT1, is involved in melatonin-induced enhancement of cell-mediated and humoral function in mice. cAMP signal transduction plays an important role in regulating lymphocyte function and this pathway appeared to be abnormal in aged mice[131]. Melatonin antagonized partially forskolin-induced increase of cAMP levels of lymphocytes; indeed, G1 protein coupled adenylate cyclase-cAMP signal pathway may be one of the important mechanisms for the anti-inflammatory immunoregulation by melatonin [132]. Melatonin enhanced significantly met-enkephalin in 2 and 11 months old mice, and the effect was blocked by nifedipine, a Ca2+ antagonist [132]. This suggests that melatonin promotes the production of met-enkephalin through L-type Ca2+ channel. Melatonin-induced immunoregulation may depend upon immuno-opiod interaction [133]. It has been suggested that Th-1 responses are readily transformed into Th-2 dominance through depletion of intracellular GSH [134]. GSH in its reduced form is the single most important protective and regulatory antioxidant in cells. The work of Peterson and his coworkers [135] showed that depletion of glutathione from antigen presenting cells in vivo resulted in lowered Th-1 activity and higher Th-2 activity. Murata et al. [136,137] showed that oxidized macrophages exhibited higher levels of oxidized glutathione as they polarized to type Th-2 cells. Thus the immune activity can have Th-1 or Th-2 characteristics depending upon the relative antioxidant status of the cells. Since melatonin stimulates the production of glutathione [138] its immunoenhancing role may be partly due to its influence on the maintenance of intracellular glutathione level. Indeed, melatonin acts as a hypnotic-chronobiotic [139,140] with cytoprotective properties [141,142] as well as an immunoenhancing agent. Indeed, melatonin not only acts as a hypnotic-chronobiotic with cytoprotective properties but also as an immunoenhancing agent. Melatonin provides a functional link between the neuroendocrine and immuno-hematopoietic systems [143]. Recent studies reveal that not only melatonin but also its oxidation product N1 acetyl-N2-formyl-5-methoxykynuramine (AFMK) is very effective in acting onneutrophils [144,145]. Both melatonin and AFMK have been shown to inhibit IL-8 release from neutrophils and AFMK has been found to be more active than melatonin in this aspect. The production of TNF-alpha by neutrophils is also inhibited by melatonin and AFMK. Since TNF-alpha and IL-8 contribute to the severity of inflammatory conditions [146], the finding of melatonin inhibiting the release of IL-8 and TNF-alpha assumes significance for it may help to reduce acute and chronic inflammation. Neutrophils are more responsive than monocytes to AFMK suggesting that melatonin biosynthesis and metabolism participate in the chemical communication among leukocytes. Melatonin may be effective in optimizing intrinsic immune responses rather than acting simply as an antioxidant [147]. Dietary supplementation of melatonin has been shown to change mRNA levels of many genes and to arrest the attenuated immune responses associated with senescence [147]. Melatonin and season-dependent immune function A number of recent studies point out that seasonal changes exert influence on immune function and melatonin may play an important role in this aspect. Seasonal changes of immune function in animals are mediated by the duration of melatonin secretion, which acts as a photoperiodic signal [119]. Such seasonal changes in immune function have been observed in humans also. Increased production of proinflammatory cytokines IFN-gamma and alpha occurred during winter [148]. Highest production of IL-6 was reported in healthy volunteers during autumn/winter season [149]. In humans the seasonal changes in immune functions can be mediated by the changes in duration of melatonin secretion. Seasonal changes in cytokines like IL-6, IFN-alpha, IFN or the balance of Th-1 and Th-2 response can account for seasonal changes in mood and behavior, such as Seasonal Affective Disorder. Summary The age-associated decline in immune function, known as immunosenescence, is characterized by a decrease in the functional activity of NK cells, granulocytes and macrophages. There is significant reduction in IL-1 and diminished generation of ROS from monocytes. In addition, there is an increase of IL-6 production. Besides causing changes in innate immunity, aging is associated with changes in cellular and humoral immunity. Decreases of CD3 and CD4 and increases of CD8 cells occur in elderly individuals. The decrease in IL-2 production that occurs during aging causes a reduced antibody formation. Melatonin seems to play a significant immunomodulatory role. Melatonin enhances both innate and cellular immunity. It stimulates the production of progenitor cells of granulocytes and macrophages and of NK cells. Production of IL-2, IL-6 and IL-12 is stimulated by melatonin. Increased T-helper production, particularly of CD4+ cells, occurs after melatonin supplementation. Melatonin decreases CD8+ cells. Melatonin may act through the immune-opiod network. The regulation of immune function by melatonin appears to involve cAMP signal transduction, L-type Ca2+ channels and glutathione. The seasonal changes in immune function observed in animals and humans are likely to be mediated by the changes in the duration of melatonin secretion. Acknowledgements The author (VS) thanks Puan Rosnida Said of the Department of Physiology, School of Medical Sciences, University Sains Malaysia, for secretarial assistance in the preparation of this manuscript. ==== Refs Weinert BT Timiras PS Invited review: Theories of aging J Appl Physiol 2003 95 1706 1716 12970376 Hardman D Aging: a theory based on free radical and radiation chemistry J Gerontol 1956 11 298 300 13332224 Pawelec G Akbar A Caruso C Effros R Grubeck-Loebenstein B Wikby A Is immunosenescence infectious? Trends Immunol 2004 25 406 410 15275638 10.1016/j.it.2004.05.006 Hakim FT Flomerfelt FA Boyiadzis M Gress RE Aging, immunity and cancer Curr Opin Immunol 2004 16 151 156 15023406 10.1016/j.coi.2004.01.009 Arlt W Hewison M Hormones and immune function: implications of aging Aging Cell 2004 3 209 216 15268754 10.1111/j.1474-9728.2004.00109.x Srinivasan V Melatonin oxidative stress and neurodegenerative diseases Indian J Exp Biol 2002 40 668 679 12587715 Karasek M Melatonin, human aging, and age-related diseases Exp Gerontol 2004 39 1723 1729 15582288 10.1016/j.exger.2004.04.012 Pawelec G Ouyang Q Colonna-Romano G Candore G Lio D Caruso C Is human immunosenescence clinically relevant? Looking for 'immunological risk phenotypes' Trends Immunol 2002 23 330 332 12103341 10.1016/S1471-4906(02)02255-X Nelson RJ Seasonal immune function and sickness responses Trends Immunol 2004 25 187 192 15039045 10.1016/j.it.2004.02.001 Reiter RJ Tan D Osuna C Gitto E Actions of melatonin in the reduction of oxidative stress. A review J Biomed Sci 2000 7 444 458 11060493 Reiter RJ Calvo JR Karbownik M Qi W Tan DX Melatonin and its relation to the immune system and inflammation Ann N Y Acad Sci 2000 917 376 386 11268363 Hotchkiss AK Nelson RJ Melatonin and immune function: hype or hypothesis? Crit Rev Immunol 2002 22 351 371 12803316 Esquifino AI Pandi-Perumal SR Cardinali DP Circadian organization of the immune response: A role for melatonin Clin Appl Immunol Rev 2004 4 423 433 10.1016/j.cair.2004.08.002 Pawelec G Effros RB Caruso C Remarque E Barnett Y Solana R T cells and aging (update february 1999) Front Biosci 1999 4 D216 D269 10051456 Pawelec G Immunosenescence: impact in the young as well as the old? Mech Ageing Dev 1999 108 1 7 10366035 10.1016/S0047-6374(99)00010-X Fry TJ Mackall CL Current concepts of thymic aging Springer Semin Immunopathol 2002 24 7 22 11974583 10.1007/s00281-001-0092-5 Tarazona R Solana R Ouyang Q Pawelec G Basic biology and clinical. impact of immunosenescence Exp Gerontol 2002 37 183 189 11772502 10.1016/S0531-5565(01)00182-6 Sohal RS Allen RG Oxidative stress as a causal factor in differentiation and aging: a unifying hypothesis Exp Gerontol 1990 25 499 522 2097168 10.1016/0531-5565(90)90017-V Pawelec G Solana R Immunoageing – the cause or effect of morbidity Trends Immunol 2001 22 348 349 11505929 10.1016/S1471-4906(01)01956-1 van den Biggelaar AH Huizinga TW de Craen AJ Gussekloo J Heijmans BT Frolich M Westendorp RG Impaired innate immunity predicts frailty in old age. The Leiden 85-plus study Exp Gerontol 2004 39 1407 1414 15489064 10.1016/j.exger.2004.06.009 Borrego F Alonso MC Galiani MD Carracedo J Ramirez R Ostos B Pena J Solana R NK phenotypic markers and IL2 response in NK cells from elderly people Exp Gerontol 1999 34 253 265 10363791 10.1016/S0531-5565(98)00076-X Miyaji C Watanabe H Toma H Akisaka M Tomiyama K Sato Y Abo T Functional alteration of granulocytes, NK cells, and natural killer T cells in centenarians Hum Immunol 2000 61 908 916 11053634 10.1016/S0198-8859(00)00153-1 Bonafe M Valensin S Gianni W Marigliano V Franceschi C The unexpected contribution of immunosenescence to the leveling off of cancer incidence and mortality in the oldest old Crit Rev Oncol Hematol 2001 39 227 233 11500264 Mariani E Meneghetti A Neri S Ravaglia G Forti P Cattini L Facchini A Chemokine production by natural killer cells from nonagenarians Eur J Immunol 2002 32 1524 1529 12115634 10.1002/1521-4141(200206)32:6<1524::AID-IMMU1524>3.0.CO;2-E Ginaldi L De Martinis M D'Ostilio A Marini L Loreto MF Quaglino D The immune system in the elderly: III. Innate immunity Immunol Res 1999 20 117 126 10580637 Albright JW Bream JH Bere EW Young HA Winkler-Pickett R Ortaldo JR Aging of innate immunity: functional comparisons of NK/LAK cells obtained from bulk cultures of young and aged mouse spleen cells in high concentrations of interleukin-2 Exp Gerontol 2004 39 73 82 14724067 10.1016/j.exger.2003.09.017 Krishnaraj R Senescence and cytokines modulate the NK cell expression Mech Ageing Dev 1997 96 89 101 9223113 10.1016/S0047-6374(97)00045-6 Fulop T JrForis G Worum I Paragh G Leovey A Age related variations of some polymorphonuclear leukocyte functions Mech Ageing Dev 1985 29 18 10.1016/0047-6374(85)90041-7 Lipschitz DA Udupa KB Boxer LA The role of calcium in the age-related decline of neutrophil function Blood 1988 71 659 665 2830926 Indelicato SR Udupa KB Balazovich KJ Boxer LA Lipschitz DA Effect of age on phorbol-ester stimulation of human neutrophils J Gerontol 1990 45 B75 B80 2159497 Lipschitz DA Udupa KB Indelicato SR Das M Effect of age on second messenger generation in neutrophils Blood 1991 78 1347 1354 1652310 Lord JM Butcher S Killampali V Lascelles D Salmon M Neutrophil ageing and immunesenescence Mech Ageing Dev 2001 122 1521 1535 11511394 10.1016/S0047-6374(01)00285-8 Rich EA Mincek MA Armitage KB Duffy EG Owen DC Fayen JD Hom DL Ellner JJ Accessory function and properties of monocytes from healthy. elderly humans for T lymphocyte responses to mitogen and antigen Gerontology 1993 39 93 108 8514204 Franceschi C Bonafe M Valensin S Olivieri F De Luca M Ottaviani E De Benedictis G Inflamm-aging. An evolutionary perspective on immunosenescence Ann N Y Acad Sci 2000 908 244 254 10911963 Castle SC Clinical relevance of age-related immune dysfunction Clin Infect Dis 2000 31 578 585 10987724 10.1086/313947 Ershler WB Interleukin-6: a cytokine for gerontologists J Am Geriatr Soc 1993 41 176 181 8426042 Mysliwska J Bryl E Foerster J Mysliwski A Increase of interleukin 6 and decrease of interleukin 2 production during the ageing process are influenced by the health status Mech Ageing Dev 1998 100 313 328 9578118 10.1016/S0047-6374(97)00154-1 Straub RH Miller LE Scholmerich J Zietz B Cytokines and hormones as possible links between endocrinosenescence and immunosenescence J Neuroimmunol 2000 109 10 15 10969175 10.1016/S0165-5728(00)00296-4 Forsey RJ Thompson JM Ernerudh J Hurst TL Strindhall J Johansson B Nilsson BO Wikby A Plasma cytokine profiles in elderly humans Mech Ageing Dev 2003 124 487 493 12714257 10.1016/S0047-6374(03)00025-3 Ferrucci L Harris TB Guralnik JM Tracy RP Corti MC Cohen HJ Penninx B Pahor M Wallace R Havlik RJ Serum IL-6 level and the development of disability in older persons J Am Geriatr Soc 1999 47 639 646 10366160 Harris TB Ferrucci L Tracy RP Corti MC Wacholder S Ettinger WH JrHeimovitz H Cohen HJ Wallace R Associations of elevated interleukin-6 and C-reactive protein levels with mortality in the elderly Am J Med 1999 106 506 512 10335721 10.1016/S0002-9343(99)00066-2 Morisaki N Saito I Tamura K Tashiro J Masuda M Kanzaki T Watanabe S Masuda Y Saito Y New indices of ischemic heart disease and aging: studies on the serum levels of soluble intercellular adhesion molecule-1 (ICAM-1) and soluble vascular cell adhesion molecule-1 (VCAM-1) in patients with hypercholesterolemia and ischemic heart disease Atherosclerosis 1997 131 43 48 9180243 10.1016/S0021-9150(97)06083-8 Miles EA Thies F Wallace FA Powell JR Hurst TL Newsholme EA Calder PC Influence of age and dietary fish oil on plasma soluble adhesion molecule concentrations Clin Sci (Lond) 2001 100 91 100 11115423 Ginaldi L Loreto MF Corsi MP Modesti M De Martinis M Immunosenescence and infectious diseases Microbes Infect 2001 3 851 857 11580980 10.1016/S1286-4579(01)01443-5 Fagnoni FF Vescovini R Passeri G Bologna G Pedrazzoni M Lavagetto G Casti A Franceschi C Passeri M Sansoni P Shortage of circulating naive CD8(+) T cells provides new insights on immunodeficiency in aging Blood 2000 95 2860 2868 10779432 Weksler ME Szabo P The effect of age on the B-cell repertoire J Clin Immunol 2000 20 240 249 10939711 10.1023/A:1006659401385 Breitbart E Wang X Leka LS Dallal GE Meydani SN Stollar BD Altered memory B-cell homeostasis in human aging J Gerontol A Biol Sci Med Sci 2002 57 B304 B311 12145356 Colonna-Romano G Bulati M Aquino A Scialabba G Candore G Lio D Motta M Malaguarnera M Caruso C B cells in the aged: CD27, CD5, and CD40 expression Mech Ageing Dev 2003 124 389 393 12714244 10.1016/S0047-6374(03)00013-7 Ginaldi L De Martinis M D'Ostilio A Marini L Loreto MF Corsi MP Quaglino D The immune system in the elderly: I. Specific humoral immunity Immunol Res 1999 20 101 108 10580635 Cossarizza A Ortolani C Monti D Franceschi C Cytometric analysis of immunosenescence Cytometry 1997 27 297 313 9098622 10.1002/(SICI)1097-0320(19970401)27:4<297::AID-CYTO1>3.0.CO;2-A Linton PJ Haynes L Tsui L Zhang X Swain S From naive to effector-alterations with aging Immunol Rev 1997 160 9 18 9476661 Ouyang Q Wagner WM Zheng W Wikby A Remarque EJ Pawelec G Dysfunctional CMV-specific CD8+ T cells accumulate in the elderly Exp Gerontol 2004 39 607 613 15050296 10.1016/j.exger.2003.11.016 Ferguson FG Wikby A Maxson P Olsson J Johansson B Immune parameters in a longitudinal study of a very old population of Swedish people: a comparison between survivors and nonsurvivors J Gerontol A Biol Sci Med Sci 1995 50 B378 B382 7583794 Wikby A Maxson P Olsson J Johansson B Ferguson FG Changes in CD8 and CD4 lymphocyte subsets, T cell proliferation responses and non survival in the very old: the Swedish longitudinal OCTO-immune study Mech Ageing Dev 1998 102 187 198 9720651 10.1016/S0047-6374(97)00151-6 Olsson J Wikby A Johansson B Lofgren S Nilsson BO Ferguson FG Age related change in peripheral blood T-lymphocyte subpopulations and cytomegalovirus infection in the very old: the Swedish longitudinal OCTO immune study Mech Ageing Dev 2000 121 187 201 11164473 10.1016/S0047-6374(00)00210-4 Ginaldi L De Martinis M D'Ostilio A Marini L Loreto MF Quaglino D The immune system in the elderly: III. Innate immunity Immunol Res 1999 20 117 126 10580637 Cardinali DP Pévet P Basic aspects of melatonin action Sleep Med Rev 1998 2 175 190 15310500 10.1016/S1087-0792(98)90020-X Zeitzer JM Daniels JE Duffy JF Klerman EB Shanahan TL Dijk DJ Czeisler CA Do plasma melatonin concentrations decline with age? Am J Med 1999 107 432 436 10569297 10.1016/S0002-9343(99)00266-1 Fourtillan JB Brisson AM Fourtillan M Ingrand I Decourt JP Girault J Melatonin secretion occurs at a constant rate in both young and older men and women Am J Physiol Endocrinol Metab 2001 280 E11 E22 11120654 Iguchi H Kato KI Ibayashi H Age-dependent reduction in serum melatonin. concentrations in healthy human subjects J Clin EndocrinolMetab 1982 55 27 29 Dori D Casale G Solerte SB Fioravanti M Migliorati G Cuzzoni G Ferrari E Chrono-neuroendocrinological aspects of physiological aging and senile dementia Chronobiologia 1994 21 121 126 7924632 Girotti L Lago M Ianovsky O Carbajales J Elizari MV Brusco LI Cardinali DP Low urinary 6-sulphatoxymelatonin levels in patients with coronary artery disease J Pineal Res 2000 29 138 142 11034110 10.1034/j.1600-079X.2000.290302.x Siegrist C Benedetti C Orlando A Beltran JM Tuchscherr L Noseda CM Brusco LI Cardinali DP Lack of changes in serum prolactin, FSH, TSH, and estradiol after melatonin treatment in doses that improve sleep and reduce benzodiazepine consumption in sleep-disturbed, middle-aged, and elderly patients J Pineal Res 2001 30 34 42 11168905 10.1034/j.1600-079X.2001.300105.x Mishima K Okawa M Hozumi S Hishikawa Y Supplementary administration of artificial bright light and melatonin as potent treatment for disorganized circadian rest-activity and dysfunctional autonomic and neuroendocrine systems in institutionalized demented elderly persons Chronobiol Int 2000 17 419 432 10841214 10.1081/CBI-100101055 Luboshitzky R Shen-Orr Z Tzischichinsky O Maldonado M Herer P Lavie P Actigraphic sleep-wake patterns and urinary 6-sulfatoxymelatonin excretion in patients with Alzheimer's disease Chronobiol Int 2001 18 513 524 11475420 10.1081/CBI-100103973 Mishima K Okawa M Shimizu T Hishikawa Y Diminished melatonin secretion in the elderly caused by insufficient environmental illumination J Clin Endocrinol Metab 2001 86 129 134 11231989 10.1210/jc.86.1.129 Skene DJ Swaab DF Melatonin rhythmicity: effect of age and Alzheimer's disease Exp Gerontol 2003 38 199 206 12543278 10.1016/S0531-5565(02)00198-5 Bergiannaki JD Soldatos CR Paparrigopoulos TJ Syrengelas M Stefanis CN Low and high melatonin excretors among healthy individuals J Pineal Res 1995 18 159 164 7562374 Srinivasan V Melatonin, biological rhythm disorders and phototherapy Indian J Physiol Pharmacol 1997 41 309 328 10235654 Pandi-Perumal SR Seils LK Kayumov L Ralph MR Lowe A Moller H Swaab DF Senescence, sleep, and circadian rhythms Ageing Res Rev 2002 1 559 604 12067601 10.1016/S1568-1637(02)00014-4 Monti JM Alvarino F Cardinali D Savio I Pintos A Polysomnographic study of the effect of melatonin on sleep in elderly patients with chronic primary insomnia Arch Gerontol Geriatr 1999 28 85 98 15374088 10.1016/S0167-4943(98)00129-0 Brzezinski A Vangel MG Wurtman RJ Norrie G Zhdanova I Ben Shushan A Ford I Effects of exogenous melatonin on sleep: a meta-analysis Sleep Med Rev 2005 9 41 50 15649737 10.1016/j.smrv.2004.06.004 Maestroni GJ The immunotherapeutic potential of melatonin Expert Opin Investig Drugs 2001 10 467 476 11227046 10.1517/13543784.10.3.467 Skwarlo-Sonta K Melatonin in immunity: comparative aspects Neuroendocrinol Lett 2002 23 61 66 12019354 Carrillo-Vico A Calvo JR Abreu P Lardone PJ Garcia-Maurino S Reiter RJ Guerrero JM Evidence of melatonin synthesis by human lymphocytes and its physiological significance: possible role as intracrine, autocrine, and/or paracrine substance FASEB J 2004 18 537 539 14715696 Dubocovich ML Cardinali DP Delagrange P Krause DN Strosberg D Sugden D Yocca FD IUPHAR Melatonin receptors The IUPHAR Compendium of Receptor Characterization and Classification 2000 2 London: IUPHAR Media 271 277 McArthur AJ Hunt AE Gillette MU Melatonin action and signal transduction in the rat suprachiasmatic circadian clock: activation of protein kinase C at dusk and dawn Endocrinology 1997 138 627 634 9002996 10.1210/en.138.2.627 Benitez-King G Huerto-Delgadillo L Anton-Tay F Melatonin modifies calmodulin cell levels in MDCK and N1E-115 cell lines and inhibits phosphodiesterase activity in vitro Brain Res 1991 557 289 292 1747759 10.1016/0006-8993(91)90146-M Garcia-Maurino S Pozo D Calvo JR Guerrero JM Correlation between nuclear melatonin receptor expression and enhanced cytokine production in human lymphocytic and monocytic cell lines J Pineal Res 2000 29 129 137 11034109 10.1034/j.1600-079X.2000.290301.x Guerrero JM Reiter RJ Melatonin-immune system relationships Curr Top Med Chem 2002 2 167 179 11899099 10.2174/1568026023394335 Maestroni GJ Conti A Pierpaoli W Role of the pineal gland in immunity. Circadian synthesis and release of melatonin modulates the antibody response and antagonizes the immunosuppressive effect of corticosterone J Neuroimmunol 1986 13 19 30 2944914 10.1016/0165-5728(86)90047-0 Maestroni GJ Conti A Pierpaoli W Role of the pineal gland in immunity: II.Melatonin enhances the antibody response via an opiatergic mechanism Clin Exp Immunol 1987 68 384 391 3308215 Maestroni GJ Conti A Pierpaoli W The pineal gland and the circadian, opiatergic, immunoregulatory role of melatonin Ann N Y Acad Sci 1987 496 67 77 3474997 Peters BA Sothmann M Wehrenberg WB Blood leukocyte and spleen lymphocyte immune responses in chronically physically active and sedentary hamsters Life Sci 1989 45 2239 2245 2601577 10.1016/0024-3205(89)90065-9 Molinero P Soutto M Benot S Hmadcha A Guerrero JM Melatonin is responsible for the nocturnal increase observed in serum and thymus of thymosin alpha1 and thymulin concentrations: observations in rats and humans J Neuroimmunol 2000 103 180 188 10696913 10.1016/S0165-5728(99)00237-4 Maestroni GJ Conti A Melatonin and the immune-hematopoietic system therapeutic and adverse pharmacological correlates Neuroimmunomodulation 1996 3 325 332 9266542 Maestroni GJ The photoperiod transducer melatonin and the immune hematopoietic system J Photochem Photobiol B 1998 43 186 192 9718719 10.1016/S1011-1344(98)00107-9 Maestroni GJ Conti A Lissoni P Colony-stimulating activity and hematopoietic rescue from cancer chemotherapy compounds are induced by melatonin via endogenous interleukin 4 Cancer Res 1994 54 4740 4743 7914829 Maestroni GJ Covacci V Conti A Hematopoietic rescue via T-cell dependent, endogenous granulocyte-macrophage colony-stimulating factor induced by the pineal neurohormone melatonin in tumor-bearing mice Cancer Res 1994 54 2429 2432 8162592 Currier NL Sun LZ Miller SC Exogenous melatonin: quantitative enhancement in vivo of cells mediating non-specific immunity J Neuroimmunol 2000 104 101 108 10713348 10.1016/S0165-5728(99)00271-4 Maestroni GJ Is hematopoiesis under the influence of neural and neuroendocrine mechanisms? Histol Histopathol 1998 13 271 274 9476656 Kaur C Ling EA Effects of melatonin on macrophages/microglia in postnatal rat brain J Pineal Res 1999 26 158 168 10231729 Rai S Haldar C Pineal control of immune status and hematological changes in blood and bone marrow of male squirrels (Funambulus pennanti) during their reproductively active phase Comp Biochem Physiol C Toxicol Pharmacol 2003 136 319 328 15012903 10.1016/j.cca.2003.10.008 Minagar A Shapshak P Fujimura R Ownby R Heyes M Eisdorfer C The role of macrophage/microglia and astrocytes in the pathogenesis of three neurologic disorders: HIV-associated dementia, Alzheimer disease, and multiple sclerosis J Neurol Sci 2002 202 13 23 12220687 10.1016/S0022-510X(02)00207-1 Zhang S Li W Gao Q Wei T Effect of melatonin on the generation of nitric oxide in murine macrophages Eur J Pharmacol 2004 501 25 30 15464059 10.1016/j.ejphar.2004.08.015 Barriga C Martin MI Ortega E Rodriguez AB Physiological concentrations of melatonin and corticosterone in stress and their relationship with phagocytic activity J Neuroendocrinol 2002 14 691 695 12213130 10.1046/j.1365-2826.2002.00823.x Herberman RB Ortaldo JR Natural killer cells: their roles in defenses against disease Science 1981 214 24 30 7025208 Trinchieri G Perussia B Human natural killer cells: biologic and pathologic aspects Lab Invest 1984 50 489 513 6201674 Cox WI Holbrook NJ Grasso RJ Specter S Friedman H Suppression of the natural killer cell activity of murine spleen cell cultures by dexamethasone (41489) Proc Soc Exp Biol Med 1982 171 146 150 7178117 Angeli A Gatti G Sartori ML Ponte D Carignola R Gupta D, Attanasio A, Reiter RJ Effect of exogenous melatonin on human natural killer (NK) cell activity An approach to the immunomodulatory role of the pineal gland 1988 Tubingen: Brain Research Promotion 145 156 Garcia-Maurino S Gonzalez-Haba MG Calvo JR Rafii-el-Idrissi M Sanchez-Margalet V Goberna R Guerrero JM Melatonin enhances IL-2, IL-6, and IFN-gamma production by human circulating CD4+ cells: a possible nuclear receptor-mediated mechanism involving T helper type 1 lymphocytes and monocytes J Immunol 1997 159 574 581 9218571 Drazen DL Nelson RJ Melatonin receptor subtype MT2 (mel 1b) and not MT1 (mel 1a) is associated with melatonin-induced enhancement of cell mediated and humoral immunity Neuroendocrinology 2001 74 178 184 11528219 10.1159/000054684 Carrillo-Vico A Garcia-Perganeda A Naji L Calvo JR Romero MP Guerrero JM Expression of membrane and nuclear melatonin receptor mRNA and protein in the mouse immune system Cell Mol Life Sci 2003 60 2272 2278 14618273 10.1007/s00018-003-3207-4 Garcia-Maurino S Pozo D Carrillo-Vico A Calvo JR Guerrero JM Melatonin activates Th1 lymphocytes by increasing IL-12 production Life Sci 1999 65 2143 2150 10579467 10.1016/S0024-3205(99)00479-8 Morrey KM McLachlan JA Serkin CD Bakouche O Activation of human monocytes by the pineal hormone melatonin J Immunol 1994 153 2671 2680 8077674 Barjavel MJ Mamdouh Z Raghbate N Bakouche O Differential expression of the melatonin receptor in human monocytes J Immunol 1998 160 1191 1197 9570533 Lissoni P Rovelli F Brivio F Brivio O Fumagalli L Circadian secretions of IL-2, IL-12, IL-6 and IL-10 in relation to the light/dark rhythm of the pineal hormone melatonin in healthy humans Nat Immun 1998 16 1 5 9789119 10.1159/000069464 Liu F Ng TB Fung MC Pineal indoles stimulate the gene expression of immunomodulating cytokines J Neural Transm 2001 108 397 405 11475007 10.1007/s007020170061 Fulop T JrWagner JR Khalil A Weber J Trottier L Payette H Relationship between the response to influenza vaccination and the nutritional status in institutionalized elderly subjects J Gerontol A Biol Sci Med Sci 1999 54 M59 M64 10051856 Khanna KV Markham RB A perspective on cellular immunity in the elderly Clin Infect Dis 1999 28 710 713 10825023 Csaba G Barath P Morphological changes of thymus and the thyroid gland after postnatal extirpation of pineal body Endocrinol Exp 1975 9 59 67 1084274 Liu ZM Pang SF [I125]-labelled iodomelatonin binding sites in the duck bursa of Fabricius: binding characteristics and diurnal variation Neurosci Lett 1992 146 163 166 1337193 10.1016/0304-3940(92)90068-I Maestroni GJ Conti A Pierpaoli W Role of the pineal gland in immunity. III. Melatonin antagonizes the immunosuppressive effect of acute stress via an opiatergic mechanism Immunology 1988 63 465 469 3350581 Tian YM Zhang GY Dai YR Melatonin rejuvenates degenerated thymus and redresses peripheral immune functions in aged mice Immunol Lett 2003 88 101 104 12880677 10.1016/S0165-2478(03)00068-3 Sainz RM Mayo JC Reiter RJ Antolin I Esteban MM Rodriguez C Melatonin regulates glucocorticoid receptor: an answer to its antiapoptotic action in thymus FASEB J 1999 13 1547 1556 10463946 Tian YM Li PP Jiang XF Zhang GY Dai YR Rejuvenation of degenerative thymus by oral melatonin administration and the antagonistic action of melatonin against hydroxyl radical-induced apoptosis of cultured thymocytes in mice J Pineal Res 2001 31 214 221 11589755 10.1034/j.1600-079X.2001.310304.x Maestroni GJ The immunoneuroendocrine role of melatonin J Pineal Res 1993 14 1 10 8483103 Yu Q Miller SC Osmond DG Melatonin inhibits apoptosis during early B cell development in mouse bone marrow J Pineal Res 2000 29 86 93 10981821 10.1034/j.1600-079X.2000.290204.x Nelson RJ Drazen DL Melatonin mediates seasonal changes in immune function Ann N Y Acad Sci 2000 917 404 415 11268368 Moore CB Siopes TD Effects of lighting conditions and melatonin supplementation on the cellular and humoral immune responses in japanese quail coturnix coturnix japonica Gen Comp Endocrinol 2000 119 95 104 10882554 10.1006/gcen.2000.7496 Moore CB Siopes TD Melatonin enhances cellular and humoral immune responses in the Japanese quail (Coturnix coturnix japonica) via an opiatergic mechanism Gen Comp Endocrinol 2003 131 258 263 12714007 10.1016/S0016-6480(03)00011-X Conti A Conconi S Hertens E Skwarlo-Sonta K Markowska M Maestroni JM Evidence for melatonin synthesis in mouse and human bone marrow cells J Pineal Res 2000 28 193 202 10831154 10.1034/j.1600-079X.2000.280401.x Gonzalez-Haba MG Garcia-Maurino S Calvo JR Goberna R Guerrero JM High-affinity binding of melatonin by human circulating T lymphocytes. (CD4+) FASEB J 1995 9 1331 1335 7557023 Garcia-Perganeda A Pozo D Guerrero JM Calvo JR Signal transduction for melatonin in human lymphocytes: involvement of a pertussis toxin-sensitive G protein J Immunol 1997 159 3774 3781 9378964 Carrillo-Vico A Garcia-Maurino S Calvo JR Guerrero JM Melatonin counteracts the inhibitory effect of PGE2 on IL-2 production in human lymphocytes via its mt1 membrane receptor FASEB J 2003 17 755 757 12594180 Castrillón P Esquifino AI Varas A Zapata A Cutrera RA Cardinali DP The effect of melatonin treatment on 24 hr variations in response to mitogens and lymphocyte subpopulations in rat submaxillary lymph nodes J Neuroendocrinol 2000 12 758 765 10929088 10.1046/j.1365-2826.2000.00519.x Nunnari G Nigro L Palermo F Leto D Pomerantz RJ Cacopardo B Reduction of serum melatonin levels in HIV-1-infected individuals' parallel disease progression: correlation with serum interleukin-12 levels Infection 2003 31 379 382 14735378 Raghavendra V Singh V Shaji AV Vohra H Kulkarni SK Agrewala JN Melatonin provides signal 3 to unprimed CD4+ T cells but failed to stimulate LPS primed B cells Clin Exp Immunol 2001 124 414 422 11472402 10.1046/j.1365-2249.2001.01519.x Maestroni GJM Cardinali DP Esquifino AI Pandi-Perumal SR Does melatonin play a disease-promoting role in rheumatoid arthritis? J Neuroimmunol 2004 158 106 111 15589043 10.1016/j.jneuroim.2004.08.015 Regodon S Martin-Palomino P Fernandez-Montesinos R Hererra JL Carrascosa-Salmoral MP Piriz S Vadillo S Guerrero JM Pozo D The use of melatonin as a vaccine agent Vaccine 2005 23 5321 5327 16055232 10.1016/j.vaccine.2005.07.003 Kammer GM The adenylate cyclase-cAMP-protein kinase A pathway and regulation of the immune response Immunol Today 1988 9 222 229 2855581 10.1016/0167-5699(88)91220-0 Wei W Shen YX Dai M Chen Q Effects and mechanisms of melatonin on immune responses in mice of different months Acta Pharmacol Sin 2003 24 719 723 12852842 Maestroni GJ Conti A Anti-stress role of the melatonin-immuno-opioid network: evidence for a physiological mechanism involving T cell-derived, immunoreactive beta-endorphin and MET-enkephalin binding to thymic. opioid receptors Int J Neurosci 1991 61 289 298 1668619 Kidd P Th1/Th2 balance: the hypothesis, its limitations, and implications for health and disease Altern Med Rev 2003 8 223 246 12946237 Peterson JD Herzenberg LA Vasquez K Waltenbaugh C Glutathione levels in antigen-presenting cells modulate Th1 versus Th2 response patterns Proc Natl Acad Sci U S A 1998 95 3071 3076 9501217 10.1073/pnas.95.6.3071 Murata Y Shimamura T Hamuro J The polarization of Th 1/Th2 balance is dependent on the intracellular thiol redox status of macrophages dueto the distinctive cytokine production Int Immunol 2002 14 201 212 11809739 10.1093/intimm/14.2.201 Murata Y Amao M Hamuro J Sequential conversion of the redox status of macrophages dictates the pathological progression of autoimmune diabetes Eur J Immunol 2003 33 1001 1011 12672066 10.1002/eji.200323575 Urata Y Honma S Goto S Todoroki S Iida T Cho S Honma K Kondo T Melatonin induces gamma-glutamylcysteine synthetase mediated byactivator protein-1 in human vascular endothelial cells Free Radic Biol Med 1999 27 838 847 10515588 10.1016/S0891-5849(99)00131-8 Arendt J Skene DJ Melatonin as a chronobiotic Sleep Med Rev 2005 9 25 39 15649736 10.1016/j.smrv.2004.05.002 Cardinali DP Furio AM Reyes MP Brusco LI The use of chronobiotics in the resynchronization of the sleep/wake cycle Cancer Causes & Control 2005 Reiter RJ Tan DX Pappolla MA Melatonin relieves the neural oxidative burden that contributes to dementias Ann N Y Acad Sci 2004 1035 179 196 15681808 10.1196/annals.1332.012 Cardinali DP Parmeggiani PL, Velluti R The use of melatonin as a chronobiotic-cytoprotective agent in sleep disorders The Physiological Nature of Sleep 2005 London: Imperial College Press Claustrat B Brun J Chazot G The basic physiology and pathophysiology of melatonin Sleep Med Rev 2005 9 11 24 15649735 10.1016/j.smrv.2004.08.001 Silva SO Rodrigues MR Carvalho SR Catalani LH Campa A Ximenes VF Oxidation of melatonin and its catabolites, N-acetyl-N-formyl-5 methoxykynuramine and N-acetyl-5-methoxykynuramine, by activated leukocytes J Pineal Res 2004 37 171 175 15357661 Silva SO Rodrigues MR Ximenes VF Bueno-da-Silva AE Amarante-Mendes GP Campa A Neutrophils as a specific target for melatonin and kynuramines: effects on cytokine release J Neuroimmunol 2004 156 146 152 15465605 10.1016/j.jneuroim.2004.07.015 Adams DH Lloyd AR Chemokines: leucocyte recruitment and activation cytokines Lancet 1997 349 490 495 9040590 10.1016/S0140-6736(96)07524-1 Bondy SC Lahiri DK Perreau VM Sharman KZ Campbell A Zhou J Sharman EH Retardation of brain aging by chronic treatment with melatonin Ann N Y Acad Sci 2004 1035 197 215 15681809 10.1196/annals.1332.013 Katila H Cantell K Appelberg B Rimon R Is there a seasonal variation in the interferon-producing capacity of healthy subjects? J Interferon Res 1993 13 233 234 8366290 Maes M Stevens W Scharpe S Bosmans E De Meyer F D'Hondt P Peeters D Thompson P Cosyns P De Clerck L Seasonal variation in peripheral blood leukocyte subsets and in serum interleukin-6, and soluble interleukin 2 and -6 receptor concentrations in normal volunteers Experientia 1994 50 821 829 7523180 10.1007/BF01956463	16316470	PMC1325257	CC BY	2021-01-04 16:23:27	no	Immun Ageing. 2005 Nov 29; 2:17	utf-8	Immun Ageing	2,005	10.1186/1742-4933-2-17	oa_comm
==== Front J Circadian RhythmsJournal of Circadian Rhythms1740-3391BioMed Central London 1740-3391-3-131632116410.1186/1740-3391-3-13ResearchDaily illumination exposure and melatonin: influence of ophthalmic dysfunction and sleep duration Jean-Louis Girardin [email protected] Daniel F [email protected] Jeffrey A [email protected] Ferdinand [email protected] Arthur H [email protected] Douglas R [email protected] Department of Psychiatry and Ophthalmology, SUNY Downstate Medical Center, New York, NY2 Brooklyn Research Foundation on Minority Health, KJMC, New York, NY3 Department of Psychiatry, Maimonides Medical Center, New York, NY4 Department of Psychiatry, University of California, San Diego, CA2005 1 12 2005 3 13 13 13 10 2005 1 12 2005 Copyright © 2005 Jean-Louis et al; licensee BioMed Central Ltd.2005Jean-Louis 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 Ocular pathology lessens light's efficacy to maintain optimal circadian entrainment. We examined whether ophthalmic dysfunction explains unique variance in melatonin excretion of older adults over and above the variance explained by daily illumination, medical, and sociodemographic factors. We also examined whether ophthalmic dysfunction influences relationships between ambient illumination and melatonin. Methods Thirty older adults (mean age = 69 years; Blacks = 42% and Whites = 58%) of both genders participated in the study. Demographic and health data were collected at baseline. Participants underwent eye exams at SUNY Downstate Medical Center, wore an actigraph to monitor illumination and sleep, and collected urine specimens to estimate aMT6s concentrations. Results Hierarchical regression analysis showed that illumination factors explained 29% of the variance in aMT6s mesor. The proportion of variance explained by ophthalmic factors, sleep duration, and race was 10%, 2%, and 2%, respectively. Illumination factors explained 19% of the variance in aMT6s acrophase. The proportion of variance explained by ophthalmic factors, sleep duration, and race was 11%; 17%; and 2%, respectively. Controlling for sleep duration and race reduced the correlations between illumination and melatonin, whereas controlling for ophthalmic factors did not. Conclusion Ophthalmic exams showed that elevated intraocular pressure and large cup-to-disk ratios were independently associated with earlier melatonin timing. Lower illumination exposure also had independent associations with earlier melatonin timing. Conceivably, ophthalmic and illumination factors might have an additive effect on the timing of melatonin excretion, which in turn might predispose individuals to experience early morning awakenings. ==== Body Introduction Light influences numerous biological and behavioral functions [1-3]. In the laboratory, exposure to light of varying intensities, wavelengths, and durations entrains the circadian pacemaker [4], suppresses melatonin rhythms [2,3,5], and modulates pupillary reflexes [6-8]. Recent evidence suggests that these processes involve specialized signal transduction mechanisms of intrinsically photosensitive retinal ganglion cells [9]. These cells are believed to express melanopsin, the primary candidate photopigment in the synchronization of circadian rhythms [7,10-12]. Studies performed in the natural environment have shown that ambient illumination affects melatonin rhythms [13,14], rest-activity cycles [15,16], and mood [17,18]. Naturalistic studies have also demonstrated that several factors impinge on the level and timing of ambient illumination. They include age [16], gender [19], race/ethnicity [15,19], time standard [16], season [20-23], and latitude [20]. Notwithstanding the importance of these factors, the integrity of the visual and photic system remains the overriding component governing light's ability to entrain circadian rhythms. Generally, blind patients without conscious light perception show a loss of circadian entrainment and do not experience light-induced suppression of melatonin [2,24-28]. Emerging evidence suggests, however, that a minority of blind patients maintain the capacity for photic entrainment, as demonstrated through melatonin-suppression tests [2,25]. Thus, light transmission is not necessarily abolished in all patients with no conscious light perception, particularly where no optic diseases are suspected. A recent study, investigating adolescents and young adults ages 12–20 years from the Missouri School for the Blind, found significantly greater circadian dysfunction (e.g., more daytime napping and variable timing of awakening), among patients with optic diseases relative to those without such diseases [29]. It appears that blind patients exhibiting incapacity for photic entrainment represent a unique category. Much less is known regarding effects of age-related photic impairment on circadian rhythm functions. There are suggestions that several ophthalmic diseases could attenuate photic transmission to the circadian pacemaker. Senile miosis is one of those diseases; it is characterized by an age-related reduction in pupil diameter that could reduce retinal illumination [6,30]. Opacification and yellowing of the crystalline lens of the eye, as seen in patients with cataracts, can also substantially reduce photic transmission [31]. Loss of retinal ganglion cells, which afflicts primarily glaucoma patients, might negatively affect retinal phototransduction to the pacemaker [32,33]. It of great interest to ascertain how each of these ophthalmic diseases compromises light input to the circadian system. Judging from the available evidence, it is reasonable to hypothesize that age-associated ocular pathology lessens light's efficacy to maintain optimal circadian entrainment [34-36]. In the present study, we tested the hypothesis that ophthalmic dysfunction explains unique variance in melatonin excretion of older adults over and above that explained by daily illumination, medical, and sociodemographic factors. A parallel hypothesis examined in this study was that ophthalmic dysfunction influences the relationships between ambient illumination and endogenous melatonin rhythms. Methods Participants Data presented in this paper were from a study investigating relations of ambient illumination to depression and melatonin excretion. Associations of daily illumination exposure with depression have been reported elsewhere [18]. The present report focuses on relationships of daily illumination and ophthalmic measures to melatonin excretion. Respondents to study advertisements completed baseline questionnaires. They were included if they had no current eye diagnosis, their self-stated race was Black or White, were 60 years old or older, and provided informed consent under the supervision of the Institutional Review Boards at SUNY and UCSD. They were excluded if they indicated major depression or lithium use, sleep apnea, drugs that influence endogenous melatonin, a history of ocular surgery or laser treatment, or impaired cognitive or functional ability. Respondents were compensated for participating in the study. Volunteers meeting study criteria provided demographic and health-related data, underwent eye exams, provided illumination and sleep data, and collected urine specimens. Thirty participants (mean age = 69.03 ± 6.84 years) provided complete data for the present analyses. The sample comprised Black (43%) and White (57%) Americans of both genders (women = 80% and men = 20%), with a BMI averaging 26.89 ± 6.11 kg/m2; 87% received at least a high school diploma and the median household income was $16,500. Procedures Baseline data were acquired using the Comprehensive Assessment and Referral Evaluation (CARE), the 30-item Geriatric Depression Scale (GDS), and the Pittsburgh Sleep Quality Index (PSQI). The CARE has been widely used to assess physical health of older individuals in minority communities. It has shown good construct validity [37] as well as concurrent and predictive validity [38]. Five sub-scales were included in the present analysis: vision disorder, respiratory disease, diabetes, hypertension, and sleep disorder (Cronbach α = 0.78; 0.86; 0.82; 0.91; and 0.92, respectively). The GDS measures depressed moods. It comprises five main factors described as: sad mood, lack of energy, positive mood, agitation, and social withdrawal. According to a study that examined depressed moods among adults (≥ 60 years old) attending primary-care clinics, the GDS had a sensitivity of 100% and a specificity of 84% in screening for major depression, using a cut-off score of 10 [39]. By contrast, the original psychometric study, which used a cut-off score of 11, found a sensitivity of 81% and a specificity of 61% for major depression (DSMIII-R) [40]. Although the PSQI is not highly specific, it is a valid measure of subjective sleep quality in clinical research. A psychometric study has shown good overall reliability coefficient for the PSQI (Cronbach α = 0.77) [41]. When investigators used a cut-off point of 5.5 in the global score, sensitivity and specificity estimates were respectively 85.7% and 86.6% for primary insomnia, 80.0% and 86.6% for major depression, 83.3% and 86.6% for generalized anxiety disorder, and 83.3% and 86.6% for schizophrenia. Nonetheless, this scale does not necessarily distinguish between conditions disturbing subjective sleep. Ophthalmic assessment A trained technician performed standard examinations to assess ophthalmic disorders. These provided data on visual acuity, visual field defects (mean deviation), intraocular pressure (IOP), vertical and horizontal cup-to-disk ratios (CDR), and nerve-fiber-layer (NFL) thickness; a large CDR is an indicator of glaucoma. An ophthalmologist graded ocular photos. Snellen best-corrected visual acuity was obtained and converted into logMAR units; higher scores denoted worse visual acuity. The SITA standard program of the Humphrey Field Analyzer was used for visual field testing to estimate ocular nerve loss [42]. Results of the Ocular Hypertension Treatment Study suggested that 97% of visual field examinations are reliable [43]. Tonometry was used to assess intraocular pressure [44,45]. The Egna-Neumarkt Glaucoma study revealed that the sensitivity and specificity of tonometry in recognizing glaucoma are 80% and 98%, respectively [44]. Fundus photography was used to examine the retina and the macula [45]. Vertical and horizontal CDR in the optic disk were derived, with higher scores indicating greater abnormality. According to the Early Treatment Diabetic Retinopathy Study, agreement rates range from 78% to 83% between retinal specialists and photographic graders [46]. Peripapillary NFL thickness, a measure of atrophy of the retinal ganglion cells, was assessed with a scanning laser polarimeter (Nerve Fiber Analyzer GDX) [47]. The GDX can detect glaucomatous eyes with a sensitivity of 71% and a specificity of 91% [48]. Illumination and sleep assessment Upon completion of eye exams, participants wore the Actiwatch-L (Mini Mitter Co., Inc.) for a week at home to monitor ambient illumination and sleep. The Actiwatch-L is a monitoring device worn on the wrist, which incorporates a photometer and a linear accelerometer. The photometer registers illumination that ranges from 1 to 150,000 lux. Registered lux values are averaged across each minute and stored in memory. Illumination time-series data were imported into a computer program for least-squares cosine analyses using Action3 software. This technique is preferred because it corrects for biases due to the time of day when the recordings began and for missing data due to actigraph removal. Cosine analyses were performed on the logarithm of measured illumination. Derived circadian measures were: 1) the mesor, the fitted 24-hour average of logged illumination levels and 2) the acrophase, the timing of the peak of the fitted cosine; goodness of fit for the cosines averaged 0.65 ± 0.12. Acrophases could be linearized before performing statistical analyses, since their distribution did not cover the whole range of 360 degrees. The accelerometer of the Actiwatch-L is sensitive to 0.01 g. and has a sampling frequency of 32 Hz; it summates and records the degree and intensity of motion on a minute-by-minute basis. Actigraphic sleep time was estimated using an automatic algorithm provided by the Actiwatch manufacturer [49]. Acceptable correlations have been found between actigraphic and polysomnographic estimates of sleep duration, but the accuracy of the algorithm has not been systematically ascertained for use among older adults. Illumination and sleep log data were used to verify time-in-bed intervals before estimating sleep and wakefulness. Sleep duration was averaged across all seven days, and this was used as a measure of habitual sleep time. Melatonin assessment Urine samples were collected for approximately 24 hours near the end of the Actiwatch-L recording. Participants collected each fractional urine specimen, measured and recorded its time and total volume, and froze duplicate aliquots in two 2 cc vials. Most volunteers provided the required 10 samples spanning at least 24 hours, and most included at least one mid-sleep collection. Samples were retrieved by a staff member and sent to UCSD where they were stored at -70°C until assay of 6-sulfatoxymelatonin (aMT6s), the major urinary melatonin metabolite using 96 well ELISA kits (Buhlmann Labs, EK-M6S) purchased from ALPCO, Ltd. (Windham, NH). This is a competitive immunoassay that uses a highly specific rabbit anti-6-sulfatoxymelatonin antibody and a second antibody capture technique. Assay performance has been extensively validated by the manufacturer and results correlate well with the Arendt (Stockgrand, Ltd) RIA (r = 0.987). At the usual dilution of 1:200 the analytical sensitivity of the ELISA is 0.35 ng/ml and the functional least detectable dose (for CV < 20%) is 1.3 ng/ml. In our laboratory, control urine samples averaging 4–6 ng/ml give intra- and inter-assay CVs of 4% and 7%, respectively. To ensure reliability of the aMT6s data, we visually analyzed excretion curves of all participants to record an overall quality score for each 24-hour profile. This evaluation was performed blind to all other information about the participants and was mainly based on the shape and completeness of the ng/h curve, but agreement between ng/h and ng/ml temporal patterns, smoothness of the baseline, and reliability of the patient log were also considered. As a circadian pattern that is clear and free of irregularities is required to estimate acrophase reliably, onset, and offset, profiles with poor quality scores were excluded. Accordingly, we selected 30 suitable profiles from a total of 59 considered. Data excluded from the final batch were not assayed due mostly to missing samples or inaccurate record keeping. Volunteers providing complete melatonin data were not significantly different in clinical presentation compared to those who did not. Of note, Blacks provided a greater number of unusable melatonin samples. The aMT6s excretion rate for each urine sample was computed and transformed into 5-min epoch data and the resulting time series data were imported into Action3 software (Ambulatory Monitoring Inc., Ardsley, NY), where they were aligned with illumination data and further checked for accuracy. Twenty-four-hour least-squares cosine fits were computed for the full aMT6s collection (average duration, excluding missing data intervals was 24.00 h) yielding aMT6s mesors and acrophases. To estimate the duration of nocturnal aMT6s excretion, the onset and the offset of the excretion were estimated by interpolation of times at which the excretion rate (ng/h) crossed the mesor level. The time of onset of aMT6s excretion was estimated as the upward crossing and offset as the downward crossing of the mesor level; aMT6s duration was defined as the interval between onset and offset times. Goodness of fit for the cosines averaged 0.81 ± 0.11. Statistical analysis All acquired data were merged into SPSS 10.0 for final analyses. These included ophthalmic, sociodemographic, medical, mood, illumination, sleep, and melatonin data. Distributions were checked for normality and were transformed where necessary using appropriate statistical techniques. Frequency and measures of central tendency were used to describe the sample. MANCOVA was used to examine race effects on ophthalmic, illumination, sleep, and melatonin measures. This procedure allowed correction for multicolinearity, if detected, and adjustment for multiple comparisons. To examine which factors were predictive of the dependent variables: aMT6s mesor (fitted mean) and acrophase (timing), we employed two hierarchical regression models. This statistical modeling technique yields the proportion of variance in the dependent variable that can be explained by an additional set of factors, over and above that explained by the initial set. Accordingly, one can opt to use the restricted model component, providing results only for the initial set. One can also use the expanded model, which sequentially analyzes the independent contribution of additional sets. In the present analysis, the initial set comprised the mesor and the acrophase of illumination. Three other sets of factors: demographic, medical, and ophthalmic were entered in a stepwise manner. The first regression model used aMT6s mesor as the dependent variable and the illumination data plus three sets of factors as predictors. In the second model, aMT6s acrophase timing was used as the dependent variable, and the above factors were entered as in the first model. Factors in these analyses were chosen because of their associations with the dependent measures and/or because of their hypothesized connection to melatonin. The selection process was based on preliminary results of the Pearson and Spearman correlations that were run to examine the magnitude of the correlation between each factor and the dependent variables and by examination of their collinearity. Results of these preliminary analyses revealed that race/ethnicity was the most important factor for the sociodemographic set (i.e., age, sex, race, education, and income). Of the medical set (BMI, hypertension, diabetes, mood, sleep duration, and sleep quality), sleep duration was chosen. Of the ophthalmic set (i.e., visual acuity, CDR ratios, IOP, visual fields mean deviation, and NFL thickness), IOP and horizontal CDR were selected; these two factors were chosen because they showed similar coefficients and because of their theoretical importance as indicators of glaucoma in the regression model. To assess whether associations between illumination and melatonin were mediated by ophthalmic factors, partial correlations were used. In that analysis, the ophthalmic factors were controlled. In separate partial correlation analyses, effects of the demographic and medical factors were controlled. Results Most participants (79%) were in good health. None were legally blind, but 30% were visually impaired based on standard criteria (best corrected vision worse than 20/40 and better than 20/200 in the better eye) [50]. Of the sample, 83% reported being satisfied with their sleep, although 61% indicated either difficulty initiating sleep, difficulty maintaining sleep, early morning awakening, or daytime napping. Moreover, 23% reported a respiratory condition, 60% hypertension, 77% arthritis, 43% vision problems, and 14% diabetes. Fifty-two percent reported social drinking, 15% indicated consumption of sleep aids, and 7% were current smokers. On average, volunteers had a GDS score of 7.07 ± 3.69 and a PSQI score of 4.68 ± 2.80. Subjective and actigraphic estimates of total sleep time averaged 6.40 ± 1.04 hours and 7.55 ± 1.74 hours, respectively. Median ambient illumination was 565.68 lux. Median aMT6s excretion was 324.60 ng/h. The medians for the acrophases of illumination and aMT6s were 14.12 hours and 3.18 hours (after midnight), respectively. As seen in Table 1, race had significant effects on ophthalmic measures, indicating greater ophthalmic dysfunction for Blacks. In Table 2, we present results of race effects on illumination, melatonin, and sleep measures. Table 1 Values represent adjusted mean ± standard error of ophthalmic measures. Data obtained for visual acuity were converted into logMAR units. Intraocular pressure and horizontal and vertical cup-to-disk ratios were log-transformed. For visual field mean deviation and nerve-fiber-layer thickness, a z-transformation procedure was used. Values were adjusted for effects of age and gender. MANCOVA: Race Effects on Ophthalmic Measures Variable Black (mean ± SE) White (mean ± SE) F p Visual Acuity (logMAR) -0.27 ± 0.07 -0.18 ± 0.06 0.872 0.359 Intraocular Pressure (mmHg) 1.25 ± 0.02 1.18 ± 0.02 4.991 0.034 Vertical Cup/Disk Ratio (mm2) -0.39 ± 0.06 -0.56 ± 0.05 6.090 0.020 Horizontal Cup/Disk Ratio (mm2) -0.44 ± 0.05 -0.60 ± 0.04 5.060 0.033 Visual Field Mean Deviation (dB) -0.52 ± 0.33 0.13 ± 0.27 2.064 0.163 Nerve-Fiber-Layer Thickness (μm) -0.36 ± 0.32 0.53 ± 0.28 4.011 0.056 Table 2 Adjusted mean values ± standard error for illumination (lux), melatonin (aMT6s), and sleep measures. Values were adjusted for effects of age and gender. MANCOVA: Race Effects on Illumination, Melatonin, and Sleep Variable Black (mean ± SE) White (mean ± SE) F p Light Mesor [log] 1.03 ± 0.10 1.38 ± 0.08 6.033 0.022 Light Phase [hr] 13.57 ± 0.34 14.54 ± 0.27 4.306 0.049 aMT6s Mesor [log] 2.67 ± 0.14 2.31 ± 0.11 3.311 0.082 aMT6s Phase [hr, after midnight] 2.35 ± 0.37 3.45 ± 0.29 4.745 0.040 Phase Angle Between aMT6s and Sleep Timing [hr] 2.55 ± 0.43 2.25 ± 0.29 0.305 0.586 Mid-Sleep [hr] 4.76 ± 0.34 5.51 ± 0.27 2.580 0.122 Sleep Duration [hr] 5.75 ± 0.28 6.57 ± 0.21 4.800 0.039 Analysis indicated that the mesor and the acrophase of aMT6s were both associated with the sociodemographic, medical, and ophthalmic factors. The multiple correlation (r2) of aMT6s mesor to these factors added individually was: [r2 = 0.24; r2 = 0.23; r2 = 0.15, respectively]; for aMT6s acrophase, it was: [r2 = 0.15; r2 = 0.21; r2 = 0.28, respectively]. However, in the interest of developing parsimonious regression models and because our sample was too small for a detailed analysis of the overlapping effects of all of the factors on the dependent variables, we selected representative factors from each set of factors. Accordingly, besides the mesor and acrophase of illumination only race, sleep duration, CDR and IOP were entered into the hierarchical regression models as predictors. With a sample size of 30 and an alpha value set at 0.05, it was determined a priori that the study would have power of 0.85 to construct a reliable model with six predictors, accounting for 41% of the variance in the dependent variable. Results of the first hierarchical regression analysis showed that illumination factors explained 29% of the variance in aMT6s mesor; illumination acrophase was the main contributor, indicating that individuals showing later timing had lower aMT6s mesors. Sequential addition of the other factors (i.e., CDR and IOP, entered as a set, sleep duration, and race) showed that the proportion of variance explained by each was 10%, 2%, and 2%, respectively. Overall, the expanded model accounted for 43% of the variance in aMT6s mesor [F = 3.47, p < 0.05]. The adjusted stepwise correlations of each of the factors to aMT6s mesor were: illumination mesor [rp = -0.08], illumination acrophase [rp = -0.49], race [rp = -0.08], sleep duration [rp = -0.25], IOP [rp = 0.31], and CDR [rp = -0.25]. For each of these correlations, effects of the other five factors were simultaneously adjusted. In the second hierarchical regression analysis, where aMT6s acrophase was the dependent variable, the illumination factors explained 19% of the variance; individuals receiving greater daily illumination level and showing later illumination timing were likely to show later aMT6s timing. The proportion of variance explained by the factors: CDR and IOP (entered as a set), sleep duration, and race was 11%; 17%; and 2%, respectively. Altogether, the expanded model accounted for 49% of the variance in aMT6s acrophase [F = 2.64, p < 0.05]. The adjusted stepwise correlations of each of the factors to aMT6s acrophase were: illumination mesor [rp = 0.41], illumination acrophase [rp = 0.09], race [rp = -0.03], sleep duration [rp = 0.48], IOP [rp = -0.29], and CDR [rp = -0.32]. In Table 3, we present results of the partial correlation analyses, examining associations of illumination factors with melatonin measures. Consistent with regression results, later timing of illumination was significantly associated with lower aMT6s mesor. Controlling for sleep duration and race somewhat reduced this association, whereas controlling for IOP and CDR affected them little. Trends suggested that greater illumination was associated with later aMT6s timing. Table 3 Values represent correlation coefficients (Coef.) for associations of ambient illumination with melatonin measures from three separate analyses. First, Pearson correlations were run with no control for the covariates. Second, partial correlations were run with control for sleep duration and race. Third, partial correlations were run with control for intraocular pressure (IOP) and cup-to-disk ratio (CDR). Relationships Between Illumination and Melatonin aMT6s Mesor aMT6s Phase Variable Coef. p Coef. p No control [r] Light Mesor 0.07 0.73 0.31 0.09 Light Phase -0.50 0.01 0.18 0.35 Control for sleep and race [rp] Light Mesor 0.19 0.35 0.28 0.16 Light Phase -0.43 0.03 0.03 0.88 Control for IOP and CDR [rp] Light Mesor 0.07 0.71 0.22 0.25 Light Phase -0.49 0.01 0.08 0.61 Discussion The data show that ophthalmic dysfunction was associated with the endogenous melatonin rhythms of community-residing older adults. Ophthalmic factors explained a significant proportion of the variance in 24-hr 6-sulphatoxymelatonin excretion (mesor) and timing (acrophase), over and above the variance explained by daily illumination, sleep duration, and race. Although most of the volunteers were in good health, ophthalmic exams showed significant evidence of photic impairment anchored by elevated intraocular pressure and large cup-to-disk ratios, which were independently associated with earlier melatonin timing. We observed that lower illumination levels also had independent associations with earlier melatonin timing. Conceivably, ophthalmic and illumination factors might have an additive effect on the timing of melatonin excretion, which in turn might predispose individuals to experience early morning awakenings. As greater intraocular pressure and cup-to-disk ratio may be indicative of optic nerve loss, a common finding among glaucoma patients [32], their effects on melatonin rhythms might be mediated by a defect in retinohypothalamic stimulation. Unfortunately, this study did not offer direct support for this hypothesis. Ophthalmic dysfunction does not seem to have a mediating effect on the relationships between ambient illumination and melatonin rhythms, as these relationships remained virtually unchanged when we controlled for differences in ophthalmic factors. Hence, abnormalities in both IOP and CDR may have had a direct effect on the timing of melatonin excretion of White and Black participants. However, associations of IOP and CDR with the amount of melatonin excretion were mixed, with greater IOP predicting greater excretion while greater CDR predicted lower excretion rates, which was in the expected direction. This discrepancy merits further examination, but we might consider that previous studies of melatonin rhythms in uncontrolled environments have shown that the acrophase, rather than the mesor of melatonin excretion, strongly correlated to ambient illumination [13], depression scores [51], activity rhythms [52], napping behavior [53], and duration and timing of sleep [52,54]. Habitual illumination pattern was the best predictor of aMT6s rhythms of all the factors in the regression models. Both brighter and later illumination exposure correlated to later aMT6s timing, although illumination level was a better predictor in the regression model. We noted that the timing of illumination exposure, rather than its mesor, correlated significantly to the mesor of aMT6s. It might be that a later illumination acrophase reflects less illumination exposure in the morning before the endogenously timed offset of melatonin secretion. Therefore, a later illumination acrophase might be associated with less morning light suppression of melatonin and, in turn, a delayed acrophase of aMT6s excretion. The timing of daily illumination might be a better index of the amount of aMT6s excretion, irrespective of individuals' sociodemographic and medical characteristics. Evidently, this must be balanced against the observation that the timing of melatonin excretion can be influenced by age-related weakening of the circadian pacemaker as well as by individual preferences in the timing of outdoor daylight activities [55-57]. Other factors influencing melatonin excretion in the natural setting include day length, age, duration and timing of sleep, and usage of certain medications [13,16,19,23,58]. Our analysis considered the relative contribution of all these factors, except for day length (season), but scheduling of the recordings was balanced across seasons throughout the study period. Sleep duration is another factor that played an important role in the analyses. That sleep duration correlated with both the mesor and the acrophase of aMT6s is consistent with previous findings [59-61]. We would have expected that shorter sleep duration would correlate with reduced aMT6s excretion, as predicted by data suggesting a longer experience of nocturnal darkness (as might be associated with a longer sleep duration) results in a longer duration of melatonin excretion [62]. The inverse correlation found in our study may have been influenced by the finding that Blacks slept less than did Whites while showing greater mesors of aMT6s excretion. It is noteworthy that in our preliminary analyses aMT6s measures had stronger correlations to sleep duration than to a history of hypertension, diabetes, or respiratory disease, BMI, mood or sleep quality. Possibly, sleep duration is a proxy for these measures, as it correlates to each, albeit to varying degrees. Of all the sociodemographic factors we analyzed, race was the strongest correlate of aMT6s measures. This is consistent with results of the analysis of covariance reported in Table 2. Independent of individuals' age and gender, race had significant effects on the melatonin measures. Similarly, race had significant effects on the ophthalmic, illumination, and sleep variables. These findings evidence that race is an important factor when analyzing sleep and circadian rhythm measures. Notwithstanding, it is less robust than the illumination, sleep, and ophthalmic factors in explaining the variance in aMT6s measures. One explanation for the reduced significance of race in the regression models relates to the shared variance in aMT6s measures explained by both race and these other factors. Consistent with previous epidemiological and clinical data, individuals of the Black race showed worse scores on ophthalmic exams [63,64]. A thinner nerve fiber layer, an elevated intraocular pressure, and greater cup-to-disk ratios, as observed among Blacks, are three important indicators of optic nerve loss in glaucoma. One implication of these findings is that since glaucoma is more common among Blacks [65,66], they may be at increased risks of developing circadian abnormalities through reduction of photic transduction to the circadian pacemaker. Since we used a relatively small sample size, we could not assess the overlapping effects of all the independent factors on melatonin rhythms. It was evident that daily illumination, ophthalmic factors, sleep duration, and race each had independent associations with both the mesors and acrophases of melatonin excretion. Although our regression models approximated predictions of the power analysis, they warrant replication in a larger sample. Efforts should be made to provide detailed instructions in gathering melatonin samples among minority groups. The observation that Blacks had lower illumination exposure, greater ophthalmic dysfunction, and higher aMT6s levels merits further empirical study, as these characteristics are suggestive of depressed moods [18,51,58,67]. Competing interests The author(s) declare that they have no competing interests. Authors' contributions GJL supervised volunteer recruitment, data collection, data analysis, and drafting of the manuscript. DFK helped design the study and assisted in data analysis and drafting of the manuscript. JAE performed aMT6s assays and assisted in the drafting of the manuscript. WAH participated in the analysis and interpretation of the ophthalmic data; he also assisted in the drafting of the manuscript. LDR helped with the interpretation of the ophthalmic data and with the drafting of the manuscript. All authors read and approved the final manuscript. Acknowledgements This research was supported by NIA (AG12364-07S1). We thank Dr. E. Leung, Dr. T. Brevetti, and J. Pierre-Louis for their assistance in the study. ==== Refs Hollick MF Jung EG Hollick MF and Jung EG Biologic effects of light 1998 Norwell, MA, Kluwer Academic Publishers Czeisler CA Shanahan TL Klerman EB Martens H Brotman DJ Emens JS Klein T Rizzo JF Suppression of melatonin secretion in some blind patients by exposure to bright light [see comments] N Engl J Med 1995 332 6 11 7990870 10.1056/NEJM199501053320102 Czeisler CA The effect of light on the human circadian pacemaker Ciba Found Symp 1995 183 254 290 7656689 Czeisler CA Allan JS Strogatz SH Ronda JM Sánchez R Ríos CD Freitag WO Richardson GS Kronauer RE Bright light resets the human circadian pacemaker independent of the timing of the sleep-wake cycle Science 1986 8 667 671 3726555 Lewy AJ Wehr TA Goodwin FK Newsome DA Markey SP Light suppresses melatonin secretion in humans Science 1980 210 1267 1269 7434030 Gaddy JR Ruberg FL Brainard GC Rollag MD Jung EG and Holick MF Pupillary modulation of light-induced melatonin suppression Biologic effects of light 1994 Berlin, Walter de Gruyter & Co. 159 168 Panda S Nayak SK Campo B Walker JR Hogenesch JB Jegla T Illumination of the melanopsin signaling pathway Science 2005 307 600 604 15681390 10.1126/science.1105121 Hattar S Lucas RJ Mrosovsky N Thompson S Douglas RH Hankins MW Lem J Biel M Hofmann F Foster RG Yau KW Melanopsin and rod-cone photoreceptive systems account for all major accessory visual functions in mice Nature 2003 424 76 81 12808468 10.1038/nature01761 Berson DM Dunn FA Takao M Phototransduction by retinal ganglion cells that set the circadian clock Science 2002 295 1070 1073 11834835 10.1126/science.1067262 Sakamoto K Liu C Tosini G Classical photoreceptors regulate melanopsin mRNA levels in the rat retina J Neurosci 2004 24 9693 9697 15509757 10.1523/JNEUROSCI.2556-04.2004 Provencio I Rodriguez IR Jiang G Hayes WP Moreira EF Rollag MD A novel human opsin in the inner retina J Neurosci 2000 20 600 605 10632589 Sekaran S Lupi D Jones SL Sheely CJ Hattar S Yau KW Lucas RJ Foster RG Hankins MW Melanopsin-dependent photoreception provides earliest light detection in the mammalian retina Curr Biol 2005 15 1099 1107 15964274 10.1016/j.cub.2005.05.053 Youngstedt SD Kripke DF Elliott JA Circadian phases of illumination and melatonin are associated. Sleep Research 1997 26 760 M D Hollick MF and Jung EG Light exposure and melatonin secretion in shiftworkers. Biologic effects of light 1998 Norwell, MA, Kluwer Academic Publishers 437 446 Kripke DF Jean-Louis G Elliott JA Klauber MR Rex KM Tuunainen A Langer RD Ethnicity, sleep, mood, and illumination in postmenopausal women. BMC Psychiatry 2004 4 8 9 15070419 10.1186/1471-244X-4-8 Jean-Louis G Kripke DF Ancoli-Israel S Klauber M Sepulveda RS Mowen MA Assmus JD Langer RD Circadian sleep, illumination, and activity patterns in women: Influences of aging and time reference. Physiol Behav 2000 68 347 352 10716544 10.1016/S0031-9384(99)00186-9 Espiritu RC Kripke DF Ancoli-Israel S Mowen MA Mason WJ Fell RL Klauber MR Kaplan OJ Low illumination experienced by San Diego adults: association with atypical depressive symptoms Biol Psychiatry 1994 106 780 786 Jean-Louis G Kripke D Cohen C Zizi F Wolintz A Associations of Ambient Illumination With Mood: Contribution of Ophthalmic Dysfunctions. Physiol Behav 2005 84 479 487 15763587 10.1016/j.physbeh.2005.01.011 Jean-Louis G Kripke DF Ancoli-Israel S Klauber M Sepulveda RS Sleep duration, illumination, and activity patterns in a population sample: Effects of gender and ethnicity Biol Psychiatry 2000 47 921 927 10807965 10.1016/S0006-3223(99)00169-9 Cole RJ Kripke DF Wisbey J Mason WJ Gruen W Hauri PJ Juarez S Seasonal variation in human illumination exposure at two different latitudes J Biol Rhythms 1995 10 324 334 8639941 Hebert M Dumont M Paquet J Seasonal and diurnal patterns of human illumination under natural conditions Chronobiol Int 1998 15 59 70 9493715 Guillemette J Hebert M Paquet J Dumont M Natural bright light exposure in the summer and winter in subjects with and without complaints of seasonal mood variations Biol Psychiatry 1998 44 622 628 9787887 10.1016/S0006-3223(97)00543-X Eastman CI Natural summer and winter sunlight exposure patterns in seasonal affective disorder Physiol Behav 1990 48 611 616 2082359 10.1016/0031-9384(90)90199-E Lockley SW Skene DJ Arendt J Tabandeh H Bird AC Defrance R Relationship between melatonin rhythms and visual loss in the blind J Clin Endocrinol Metab 1997 82 3763 3770 9360538 10.1210/jc.82.11.3763 Klerman EB Shanahan TL Brotman DJ Rimmer DW Emens JS Rizzo JFIII Czeisler CA Photic resetting of the human circadian pacemaker in the absence of conscious vision J Biol Rhythms 2002 17 548 555 12465888 10.1177/0748730402238237 Sack RL Lewy AJ Blood ML Keith LD Nakagawa H Circadian rhythm abnormalities in totally blind people: Incidence and clinical significance J Clin Endocrinol Metab 1992 75 127 134 1619000 10.1210/jc.75.1.127 Skene D Lockley S Arendt J Melatonin in Circadian Sleep Disorders in the Blind. Biol Signals 1999 8 90 95 10.1159/000014575 Klein T Martens H Dijk DJ Kronauer RE Seely EW Czeisler CA Circadian sleep regulation in the absence of light perception: chronic non-24-hour circadian rhythm sleep disorder in a blind man with a regular 24-hour sleep-wake schedule Sleep 1993 16 333 343 8341894 Wee R Van Gelder RN Sleep disturbances in young subjects with visual dysfunction Ophthalmology 2004 111 297 302 15019378 10.1016/j.ophtha.2003.05.014 Gaddy JR Rollag MD Brainard GC Pupil size regulation of threshold of light-induced melatonin suppression Journal of Clinical Endocrinolology and Metabolism 1993 77 1398 1401 10.1210/jc.77.5.1398 Brainard GC Gaddy JR Ruberg FM Barker FM Hanifin JP Rollag MD Moller M and Pevet P Ocular mechanisms that regulate the human pineal gland Advances in Pineal Research 1994 London, John Libby & Company Ltd. 415 432 Blumenthal EZ Weinreb RN Assessment of the retinal nerve fiber layer in clinical trials of glaucoma neuroprotection Surv Ophthalmol 2001 45 Suppl 3 S305 S312 11377454 10.1016/S0039-6257(01)00202-8 Hannibal J Hindersson P Ostergaard J Georg B Hegge FW Melanopsin is expressed in PACAp containing retinal ganglion cells of the human retinohypothalamic tract. Social Light Treatment Biol Rhythms Abst 2004 159 Czeisler CA Richardson GS Zimmerman JC Moore-Ede MC Weitzman ED Entrainment of human circadian rhythms by light-dark cycles: a reassessment Photochem Photobiol 1980 34 239 247 7267730 Morin LP The circadian visual system Brain Res Brain Res Rev 1994 19 102 127 7909471 10.1016/0165-0173(94)90005-1 Myers BL Badia P Changes in circadian rhythms and sleep quality with aging: mechanisms and interventions [published erratum appears in Neurosci Biobehav Rev 1996 Summer;20(2):I-IV] Neurosci Biobehav Rev 1995 19 553 571 8684716 10.1016/0149-7634(95)00018-6 Teresi JA Golden RR Gurland BJ Wilder DE Bennett RG Construct validity of indicator-scales developed from the Comprehensive Assessment and Referral Evaluation interview schedule J Gerontol 1984 39 147 157 6699368 Teresi JA Golden RR Gurland BJ Concurrent and predictive validity of indicator scales developed for the Comprehensive Assessment and Referral Evaluation interview schedule J Gerontol 1984 39 158 165 6699369 Lyness JM Noel TK Cox CFAU King DA Conwell Y CRaine ED Screening for depression in elderly primary care patients. A comparison of the Center for Epidemiologic Studies-Depression Scale and the Geriatric Depression Scale Arch Intern Med 1997 157 449 554 9046897 10.1001/archinte.157.4.449 Burke WJ Nitcher RL Roccaforte WH Wengel SP A prospective evaluation of the Geriatric Depression Scale in an outpatient geriatric assessment center J Am Geriatr Soc 1992 40 1227 1230 1447439 Doi Y Minowa M Uchiyama M Okawa M Kim K Shibui K Kamei Y Psychometric assessment of subjective sleep quality using the Japanese version of the Pittsburgh Sleep Quality Index (PSQI-J) in psychiatric disordered and control subjects Psychiatry Res 2000 97 165 172 11166088 10.1016/S0165-1781(00)00232-8 Bass SJ Feldman J Visual-field defects in well-defined retinal lesions using Humphrey and Dicon perimeters [In Process Citation] Optometry 2000 71 643 652 11063269 Johnson CA Keltner JL Cello KE Edwards M Kass MA Gordon MO Budenz DL Gaasterland DE Werner E Baseline visual field characteristics in the ocular hypertension treatment study Ophthalmology 2002 109 432 437 11874743 10.1016/S0161-6420(01)00948-4 Bonomi L Marchini G Marraffa M Morbio R The Relationship between Intraocular Pressure and Glaucoma in a Defined Population. data from the egna-neumarkt glaucoma study Ophthalmologica 2001 215 34 38 11125267 10.1159/000050823 Dandona L Dandona R Naduvilath TJ McCarty CA Mandal P Srinivas M Nanda A Rao GN Population-based assessment of the outcome of cataract surgery in an urban population in southern India Am J Ophthalmol 1999 127 650 658 10372874 10.1016/S0002-9394(99)00044-6 Kinyoun J Barton F Fisher M Hubbard L Aiello L Ferris FIII Detection of diabetic macular edema. Ophthalmoscopy versus photography--Early Treatment Diabetic Retinopathy Study Report Number 5. The ETDRS Research Group Ophthalmology 1989 96 746 750 2740076 Weinreb RN Dreher AW Coleman A Quigley H Shaw B Reiter K Histopathologic validation of Fourier-ellipsometry measurements of retinal nerve fiber layer thickness Arch Ophthalmol 1990 108 557 560 2322159 Medeiros FA Susanna RJ Comparison of algorithms for detection of localised nerve fibre layer defects using scanning laser polarimetry Br J Ophthalmol 2003 87 413 419 12642302 10.1136/bjo.87.4.413 Kushida CA Chang A Gadkary C Guilleminault C Carrillo O Dement WC Comparison of actigraphic, polysomnographic, and subjective assessment of sleep parameters in sleep-disordered patients Sleep Med 2001 2 389 396 14592388 10.1016/S1389-9457(00)00098-8 Tielsch JM Sommer A Witt K Katz J Royall RM Blindness and visual impairment in an American urban population. The Baltimore Eye Survey Arch Ophthalmol 1990 108 286 290 2271016 Tuunainen A Kripke DF Elliott JA Assmus JD Rex KM Klauber MR Langer RD Depression and endogenous melatonin in postmenopausal women J Affect Disord 2002 69 149 158 12103461 10.1016/S0165-0327(01)00303-2 Lockley SW Skene DJ Butler LJ Arendt J Sleep and activity rhythms are related to circadian phase in the blind. Sleep 1999 22 616 623 10450596 IY Y DF K JA E Langer RD Naps and circadian rhythms in postmenopausal women J Gerontol A Biol Sci Med Sci 2004 59 844 848 15345736 Kripke DF Jean-Louis G Assmus JD Low illumination associated with poor mood and disturbed sleep. SLTBR 1999 33 8 8 Reiter RJ Richardson BA Some perturbations that disturb the circadian melatonin rhythm Chronobiol Int 1992 9 314 321 1423740 Siegmund R Tittel M Schiefenhovel W Activity Monitoring of the inhabitants of Tauwema, a traditional Melanesian village: rest/activity behavior of Trobriand Islands Papua New Guinea Biologiacl Rhythm Research 1998 29 49 59 10.1076/brhm.29.1.49.3045 Swaab DF Hofman MA Lucassen PJ Purba JS Raadsheer FC Van de Nes JA Functional neuroanatomy and neuropathology of the human hypothalamus Anat Embryol (Berl) 1993 187 317 330 8512084 Youngstedt SD Kripke DF Elliot JA Baehr EK Sepulveda RS Hollick MF and Jung EG Light exposure, sleep quality, and depression in older adults. Biologic effects of light 1998 Norwell, MA, Kluver Academic Publishers 427 435 Hajak G Rodenbeck A Staedt J Bandelow B Huether G Rüther E Nocturnal plasma melatonin levels in patients suffering from chronic primary insomnia J Pineal Res 1993 15 191 198 8120797 Tzischinsky O Shlitner A Lavie P The association between the nocturnal sleep gate and nocturnal onset of urinary 6-sulfatoxymelatonin Acta Psychiatr Scand 1994 89 1 7 8140901 Leger D Laudon MFAU Zisapel N Nocturnal 6-sulfatoxymelatonin excretion in insomnia and its relation to the response to melatonin replacement therapy Am J Med 2004 116 91 95 14715322 10.1016/j.amjmed.2003.07.017 Wehr TA The durations of human melatonin secretion and sleep respond to changes in daylength (photoperiod) Brain Res 1995 688 77 85 8542325 10.1016/0006-8993(95)96872-I Varma R Tielsch JM Quigley HA Hilton SC Katz J Spaeth GL Sommer A Race-, age-, gender-, and refractive error-related differences in the normal optic disc Arch Ophthalmol 1994 112 1068 1076 8053821 Chi T Ritch R Pitaman B Stickler D Tsai C Hsieh FY Racial differences in optic nerve head parameters Arch Ophthalmol 1989 107 836 839 2730402 National Eye Institute Vision Report. Vision Problems in the U.S.: Prevalence of adult vision impairment and age-related eye disease in America (2002). http://www.nei.nih.gov/eyedata/pdf/VPUS.pdf. Accessed 05/28/02. 2002 Leske MC Connell AM Wu SY Nemesure B Li X Schachat A Hennis A Incidence of open-angle glaucoma: the Barbados Eye Studies. The Barbados Eye Studies Group Arch Ophthalmol 2001 119 89 95 11146731 Tsai SY Cheng CY Hsu WM Su TP Liu JH Chou P Association between visual impairment and depression in the elderly J Formos Med Assoc 2003 102 86 90 12709736	16321164	PMC1325258	CC BY	2021-01-04 16:25:19	no	J Circadian Rhythms. 2005 Dec 1; 3:13	utf-8	J Circadian Rhythms	2,005	10.1186/1740-3391-3-13	oa_comm
==== Front BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2901633665210.1186/1471-2105-6-290SoftwareEfficient analysis and extraction of MS/MS result data from Mascot™ result files Grosse-Coosmann Florian [email protected] Andreas M [email protected] Albert [email protected] Protein Mass Spectrometry and Functional Proteomics Group, Rudolf-Virchow-Center for Experimental Biomedicine, Universitaet Wuerzburg, Versbacher Strasse 9, D-97078 Wuerzburg, Germany2005 7 12 2005 6 290 290 15 7 2005 7 12 2005 Copyright © 2005 Grosse-Coosmann et al; licensee BioMed Central Ltd.2005Grosse-Coosmann 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 Mascot™ is a commonly used protein identification program for MS as well as for tandem MS data. When analyzing huge shotgun proteomics datasets with Mascot™'s native tools, limits of computing resources are easily reached. Up to now no application has been available as open source that is capable of converting the full content of Mascot™ result files from the original MIME format into a database-compatible tabular format, allowing direct import into database management systems and efficient handling of huge datasets analyzed by Mascot™. Results A program called mres2x is presented, which reads Mascot™ result files, analyzes them and extracts either selected or all information in order to store it in a single file or multiple files in formats which are easier to handle downstream of Mascot™. It generates different output formats. The output of mres2x in tab format is especially designed for direct high-performance import into relational database management systems using native tools of these systems. Having the data available in database management systems allows complex queries and extensive analysis. In addition, the original peak lists can be extracted in DTA format suitable for protein identification using the Sequest™ program, and the Mascot™ files can be split, preserving the original data format. During conversion, several consistency checks are performed. mres2x is designed to provide high throughput processing combined with the possibility to be driven by other computer programs. The source code including supplement material and precompiled binaries is available via and . Conclusion The database upload allows regrouping of the MS/MS results using a database management system and complex analyzing queries using SQL without the need to run new Mascot™ searches when changing grouping parameters. ==== Body Background For instance, protein identification via MDLC combined with tandem mass spectrometry techniques or other shotgun approaches usually generate huge data sets and compels application of software programs such as Sequest™ [1], Profound [2] or Mascot™ [3]. This produces peptide sequences that need to be grouped in order to obtain protein identifications with several peptides per hit, which increases reliability of the results. Mascot™ groups the peptide results of a single search run automatically. Recombination and merging of search runs is not supported. The data volume limits of Mascot™'s result display tool defined by the underlying computing resource are easily reached and exceeded when applied to a shotgun approach, excluding the opportunity to analyze a huge MDLC experiment at once. Generally, scientists require their protein identification results in tabular format in order to visualize, filter or sort them by several criteria. Concerning Sequest™, some open source tools for extracting data from its result files already exist, such as Out2Summary from the SASHIMI Project[4] or Sequest Browser™ [1]. For Mascot™, which produces text files in MIME format [[5-10]], such a tool is currently not available as open source. Tools like the ExtParser module integrated in Phenyx [11] convert the preprocessed HTML output of Mascot™'s result display tool rather than the original result file. The parser Mascot2XML of SASHIMI project[4] reads original Mascot™ data and converts into pepXML [12]. This program is available as open source, but does not export all information contained in the Mascot™ file. For efficient import in spread sheet applications and relational database systems, a straight-forward format is needed, in order to achieve the best performance. The MIME format of Mascot™ result files looks as shown in figure 1. Obviously, this format cannot be imported into spread sheet applications or database programs because it contains internal references. Figure 1 An example of the MIME format of Mascot™ result files is shown in this figure. Wrapped lines are indented. Some lines are removed due to space savings, marked by [...]. The original example file contains 322 lines. Cross-reference links are marked in red. Here, we present the command line tool mres2x that is capable of converting results from original, unprocessed MIME formatted Mascot™ output files (extension .DAT) into a comprehensive tabular format. Extraction of included peak lists into Sequest™'s DTA format is supported, too. Another option allows splitting the original Mascot™ output into several files in Mascot™'s native format according to the number of series of measurements. An example of running mres2x on Unix/Linux producing tab format output in mascot.tab of the file mascotresult.res stored in /tmp is the following command line: ./mres2x -d ./mascot.tab -o tab /tmp/mascotresult.res Implementation mres2x is implemented in C [13], and therefore is portable to several platforms, most notably Windows™ and Unix™. It offers a command line interface. All functionality is controlled by command line parameters that are shown in table 1. A detailed documentation of all command line switches is given in the file Overview.html (see additional file 1) included in the source code package. Table 1 The command line options of mres2x. Parameters for setting the Mascot™'s username, changing line break characters as well as debugging mode exist, too. The usage of mres2x is: mres2x -d destination -o type [-rvfpSuU] filemask_of_input_files, where the last parameter defines the input file(s) including the path and can even be a single file. The input must be in original Mascot™ format, not HTML. The files from the file mask must be in the same directory if the output format is not tab. In case of tab format output, the destination must be a single file, otherwise a folder. mrex2x explicitly expands input file masks. A description of the parameters also can be found in the file Overview.html (see additional file 1) included in the source code package. Option Parameters Explanation type s_dta m_dat Tab Describes the output format. Supported types are: s_dta Sequest™'s dta format. Only spectra data will be exported. m_dat split the input into several output files in Mascot™'s output format, one for each query. Tab write out a tabbed format for direct database upload. -r Use CR LF instead of LF as linefeed in data blocks. Some OS need special line feed characters in text files. -v Increase verbosity mode by one per occurrence of -v. A maximum of two -v is allowed. -f Overwrite files/allow usage of non-empty directories. Usually, the destination directory must be empty. -p Preserves files on unsuccessful program termination. Useful for debugging purposes. -S Show message indicator even if stderr is a terminal. -u name Set the username to name, if no entry is present and if the tab output format is selected. -U name Set the username to name in all cases if the tab output format is selected. This allows changing the username in the result files of Mascot™. The program uses one or more Mascot™ result files as input for processing.Its output can be directed to the program's standard output or to a file in case of the tabular output format. Otherwise, an existing directory must be specified as output destination. The converted tabular format is up to 40 percent smaller in size than the original data without any loss of information. It is designed for direct import into relational database management systems, but also can be used with spread sheet applications or other programs for further processing and validation. The tabular format is documented extensively in the file tabformat.html (see additional file 2), where the format of the original Mascot™ result files is implicitly documented, too. mres2x can be used to split huge Mascot™ result files into single files using the -o s_dat switch, each containing a single query and its corresponding results. This increases performance of reusing the separated results. Typical examples of use are display, analysis or validation by standard tools, such as the bundled result browser of Mascot™. Nevertheless, the main purpose of mres2x is conversion of huge MIME formatted files into a more readable and compact format for efficient direct import into database management systems, using their native import tools. Several data analysis steps are performed in order to check the validity of Mascot™ files even while processing the input data. Values are checked for their range at this stage. The most detailed validation is performed when producing the tabular format. A full cross-reference check is performed here. Thereby, it is assured that the output is fully consistent. The cross-referenced structure is shown in figure 1. In case of errors, a cleanup is performed which removes any result files produced so far and the OS is informed by a non-zero return code of the program. It depends on the error whether further analysis of the input file is performed by mres2x. If possible, the algorithm collects errors and prints them out before termination. If included in the input file, Mascot™ warnings are printed to standard error and are available by the calling program. On success, the message "Operation ended successfully." is written to the standard output. Wrapping programs can easily test for this message or for a return code of zero. An example of the output in tabular format is displayed in figure 2. The success codes (in this case E1, O0) at the end of a query section (B to E) or a file section (I to O) allow usage of database transaction rollback in case of errors. Figure 2 An example of the output of mres2x in tabular format, one record per line. The lines begin with a prefix, indicating the line type. Lower case letters indicate description lines; the corresponding data has upper case letters as prefix which may be directly followed by running numbers. The format is semicolon separated. Chapters are marked and commented in red. The format is described in the file Tabformat.html (see Additional file 2), included in the source code package. mres2x has thoroughly been tested with several thousands of data sets produced by Mascot™ version 2.1.0 and earlier. Results We compared the performance of mres2x with the result viewer named master_results.pl that comes with Mascot™ version 2.0.04, using a 368.74 megabytes large MIME formatted Mascot™ result file, containing 1,565,945 lines obtained from 60,000 MS/MS spectra. Conversion of this file with mres2x lasts 1 minute, 20 seconds, whereas display of this file with master_results.pl using the binary library msparser coming with Mascot™ takes more than 15 minutes on the same computer; the version fully implemented in Perl of master_results.pl would be even slower. Conclusion We introduced a tool capable of converting Mascot™ result files efficiently into other formats, most notable the one designed for direct database import. mres2x is designed to provide high throughput processing combined with extensive error checking and the possibility to be driven by other computer programs. Therefore, mres2x is suitable for integration into computer automated high throughput environments, using direct import into database management systems. mres2x reads Mascot™ result files and extracts all information in order to store it to another file or files. It currently supports three output formats: First, the original Mascot™ output file can be split into several files with the same format according to the number of series of measurements. Second, the original input peak lists can be extracted into DTA format. Third, a file in tabular format for direct bulk database upload can be created. In contrast to other formats, such as pepXML [12], protXML [12] and mzIdent [12], mres2x avoids the overhead implied by the need of interpreting the intermediate XML format over again. This allows for importing data directly in a relational database system or spread sheet applications. XML is a storage space consuming format[14] and parsing and interpretation of XML is a time consuming task, decreasing performance of the whole process [15]. Same with other intermediate formats, such as SQT [14]. The tab output format of mres2x is not intended to meet all requirements of the currently discussed file format standardization [[16-19]] and is not designed as a substitute of either XML format mentioned before. mres2x is designed to be used for direct bulk database uploads of Mascot™ results by means of the corresponding database management system, such as SQL*Loader of Oracle™ or bcp of SQL Server™. However, it creates an easy to parse tabular format which makes the creation of translating software to produce other formats nearly trivial. This allows export to any other industry standard. Storing the results in a database management system allows efficient complex queries on the data such as regrouping of peptide results to protein results without need to research the MS/MS data again and yields time and resource savings as well as increased flexibility. As the tab output format contains one result record per line, filtering and processing directly after conversion is easily possible, such as for false positives as well as allowing for assembling identifications. The records of protein and peptide results can be distinguished after processing, as the first character of each line indicates the record type. Availability and requirements For compilation a standard C compiler is needed. mres2x can be compiled and run on Windows™ and Unix/Linux. The program is freely available via and for download. Authors' contributions FGC implemented the program and made a draft of the manuscript. AB initiated the development. AB and AS contributed with ideas and proofread the manuscript. AB supervised the final testing. All authors have read and approved the final manuscript. List of abbreviations used HTML hypertext markup language MDLC multidimensional liquid chromatography MIME multipurpose internet mail extensions MS mass spectrometry OS operating system of a computer SQL structured query language XML extended markup language Supplementary Material Additional File 1 Documentation of mres2x, this document describes mres2x and how to use it. Click here for file Additional File 2 Format of the tab output format, this document describes the output of mres2x when tab format is selected. It implicitly documents the format of Mascot™'s result files. Click here for file Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft (SI 835/3-1; FZT82). ==== Refs Eng JK McCormack AL Yates JR An Approach to Correlate Tandem Mass Spectral Data of Peptides with Amino Acid Sequences in a Protein Database Journal of the American Society for Mass Spectrometry 1994 5 976 989 10.1016/1044-0305(94)80016-2 Zhang W Chait BT ProFound: An Expert System for Protein Identification Using Mass Spectrometric Peptide Mapping Information Analytical Chemistry 2000 72 2482 2489 10857624 10.1021/ac991363o Perkins DN Pappin DJC Creasy DM Cottrell JS Probability-Based Protein Identification by Searching Sequence Databases Using Mass Spectrometry Data Electrophoresis 1999 20 3551 3567 10612281 10.1002/(SICI)1522-2683(19991201)20:18<3551::AID-ELPS3551>3.0.CO;2-2 Pedrioli PGA The SASHIMI Project Freed N Borenstein N RFC 2045 - Multipurpose Internet Mail Extensions (MIME) Part One: Format of Internet Message Bodies Freed N Borenstein N RFC 2046 - Multipurpose Internet Mail Extensions (MIME) Part Two: Multipurpose Internet Mail Extensions Freed N Borenstein N RFC 2047 - Multipurpose Internet Mail Extensions (MIME) Part Three: Message Header Extensions for Non-ASCII Text Freed N Borenstein N RFC 2048 - Multipurpose Internet Mail Extensions (MIME) Part Four: Registration Procedures Freed N Borenstein N RFC 2049 - Multipurpose Internet Mail Extensions (MIME) Part Five: Conformance Criteria and Examples Masinter L RFC 2388 - Returning Values from Forms: multipart/form-data GeneBio GeneBio Phenyx Keller A Eng J Zhang N Li X Aebersold R A Uniform Proteomics MS/MS Analysis Platform Utilizing Open XML File Formats Molecular Systems Biology 2005 Kernighan BW Ritchie DM The C Programming Language 1990 2 , Prentice-Hall Int. McDonald WH Tabb DL Sadygov RG MacCoss MJ Venable J Graumann J Johnson JR Cociorva D Yates JR MS1, MS2, and SQT - three Unified, Compact, and Easily Parsed File Formats for the Storage of Shotgun Proteomic Spectra and Identifications Rapid Communications in Mass Spectrometry 2004 18 2162 2168 15317041 10.1002/rcm.1603 Boehm AM Galvin RP Sickmann A Extractor for ESI Quadrupole TOF Tandem MS Data Enabled for High Throughput Batch Processing BMC Bioinformatics 2004 5 15507135 Orchard S Hermjakob H Binz PA Hoogland C Taylor CF Zhu W Jr. RKJ Apweiler R Further Steps Towards Data Standardisation: The Proteomic Standards Initiative HUPO 3rd Annual Congress, Beijing 25-27th October, 2004 Proteomics 2005 5 337 3339 15700244 10.1002/pmic.200401158 Orchard S Hermjakob H Julian RK Runte K Sherman D Wojcik J Zhu W Apweiler R Common Interchange Standards for Proteomics Data: Public Availability of Tools and Schema Report on the Proteomic Standards Initiative Workshop, 2nd Annual HUPO Congress, Montreal, Canada, 8-11th October 2003 Proteomics 2004 4 490 4491 14760721 10.1002/pmic.200300694 Orchard S Zhu W Julian RK Hermjakob H Apweiler R Further Advances in the Development of a Data Interchange Standard for Proteomics Data Proteomics 2003 3 2065 2066 14625869 10.1002/pmic.200300588 Taylor CF Paton NW Garwood KL Kirby PD Stead DA Yin Z Deutsch EW Selway L Walker J Riba-Garcia I Mohammed S Deery MJ Howard JA Dunkley T Aebersold R Kell DB Lilley KS Roepstorff P Yates JR Brass A Brown AJP Cash P Gaskell SJ Hubbard SJ Oliver SG A Systematic Approach to Modeling, Capturing, and Disseminating Proteomics Experimental Data Nature Biotechnology 2003 21 247 2254 12610571 10.1038/nbt0303-247	16336652	PMC1325259	CC BY	2021-01-04 16:27:47	no	BMC Bioinformatics. 2005 Dec 7; 6:290	utf-8	BMC Bioinformatics	2,005	10.1186/1471-2105-6-290	oa_comm
==== Front BMC Evol BiolBMC Evolutionary Biology1471-2148BioMed Central London 1471-2148-5-701633225610.1186/1471-2148-5-70Research ArticleMolecular dating of caprines using ancient DNA sequences of Myotragus balearicus, an extinct endemic Balearic mammal Lalueza-Fox Carles [email protected] Jose [email protected] Lourdes [email protected]ès-Bonet Tomàs [email protected] Josep Antoni [email protected] Jaume [email protected] Unitat de Biologia Evolutiva, Facultat de Ciències de la Salut i de la Vida, Universitat Pompeu Fabra, Doctor Aiguader 80, 08003 Barcelona, Spain2 Unitat d'Antropologia, Dept. Biologia Animal, Facultat de Biologia, Universitat de Barcelona, Avda. Diagonal 645, 08028 Barcelona, Spain3 Department of Physiology and Molecular Biodiversity, Institut de Biologia Molecular de Barcelona, CSIC, 08034 Barcelona, Spain4 Institut Mediterrani d'Estudis Avançats de les Illes Balears (CSIC-UIB), Cta. de Valldemosa km 7.5, 07071 Ciutat de Mallorca, Spain2005 6 12 2005 5 70 70 11 6 2005 6 12 2005 Copyright © 2005 Lalueza-Fox et al; licensee BioMed Central Ltd.2005Lalueza-Fox 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 Myotragus balearicus was an endemic bovid from the Balearic Islands (Western Mediterranean) that became extinct around 6,000-4,000 years ago. The Myotragus evolutionary lineage became isolated in the islands most probably at the end of the Messinian crisis, when the desiccation of the Mediterranean ended, in a geological date established at 5.35 Mya. Thus, the sequences of Myotragus could be very valuable for calibrating the mammalian mitochondrial DNA clock and, in particular, the tree of the Caprinae subfamily, to which Myotragus belongs. Results We have retrieved the complete mitochondrial cytochrome b gene (1,143 base pairs), plus fragments of the mitochondrial 12S gene and the nuclear 28S rDNA multi-copy gene from a well preserved Myotragus subfossil bone. The best resolved phylogenetic trees, obtained with the cytochrome b gene, placed Myotragus in a position basal to the Ovis group. Using the calibration provided by the isolation of Balearic Islands, we calculated that the initial radiation of caprines can be dated at 6.2 ± 0.4 Mya. In addition, alpine and southern chamois, considered until recently the same species, split around 1.6 ± 0.3 Mya, indicating that the two chamois species have been separated much longer than previously thought. Conclusion Since there are almost no extant endemic mammals in Mediterranean islands, the sequence of the extinct Balearic endemic Myotragus has been crucial for allowing us to use the Messinian crisis calibration point for dating the caprines phylogenetic tree. ==== Body Background Myotragus balearicus [1] was an extremely modified caprine, endemic of the Balearic Islands, characterized by a series of unusual apomorphies developed throughout more than five million years of insular evolution [2-7]. Myotragus disappeared from Mallorca between 3,700 and 2,040 years BC [8], probably after the arrival of the first humans to the Islands. The morphological peculiarities of Myotragus, including extreme size reduction, a single ever growing lower incisor, modified limb bones and frontal eyes [2], makes it difficult to clarify its taxonomic position [9]. In two previous studies [10,11], we obtained bits of the mtDNA cytochrome b gene from two different Myotragus bone specimens from Mallorca island, one found in cova Estreta (Pollença) and the other found in Cova des Gorgs (Escorca). Although the quick radiation of all the Caprinae [12] and the short cytochrome b fragment retrieved (338 bp) made the previous analysis difficult, our data indicated that Myotragus was genetically close to the sheep (Ovis) group. There is a general agreement that the continental ancestor of the Balearic bovid possibly settled in Mallorca and Menorca during the Messinian period. The Balearic Islands were last connected to the continent during the Messinian regression, when the Mediterranean basin was dried out, allowing faunal exchanges between the islands and the continental lands [13,14]. The opening of the Strait of Gibraltar 5.35 Mya definitely isolated the Balearic Islands and promoted the beginning of the independent evolution of Myotragus [2]. The Myotragus lineage spreads from the Pliocene to the Holocene, and includes in Mallorca five chronospecies: M. pepgonellae, M. antiquus, M. kopperi, M. batei and M. balearicus. No absolute datings are available for the oldest species, M. pepgonellae, although a chronology of Lower Pliocene has been proposed; a paleomagnetic data of about 2.6 My is available for M. antiquus remains [2]. Apart from two shrews from the islands of Crete and Sicily, respectively, no other Pliocene endemic mammal persists in Mediterranean islands [15,16]. Thus, the accurately dated vicariant event that isolated Myotragus from their continental relatives constitutes a unique opportunity to use the sequences of Myotragus for calibrating the DNA molecular clock of caprines. Previously, only dates outside caprines (the emergence of the family Bovidae 18.5 Mya), or dates that are too recent for accurate estimations (common ancestry of some domestic sheep breeds a few thousand years ago) could be used [17,18]. In addition, since the molecular clock is not perfect, it is necessary to use as much sequence information as possible to obtain reliable date estimates. For this purpose, we decided to retrieve the complete cytochrome b gene and some other genetic markers of Myotragus; to do this, we used a newly excavated Myotragus bone, obtained in 2002 from Cova des Gorgs, that looked macroscopically very well preserved and therefore was suggestive of DNA survival. We designed different sets of overlapping primers for retrieving the complete cytochrome b gene, as well as a 305 bp fragment from another mtDNA gene, the 12S. In addition, we retrieved a short fragment of a nuclear gene, 28S rDNA, from the same extract. The only precedent of nuclear DNA retrieval from warm climates is, to our knowledge, the analysis of a ground sloth coprolite from the south of United States [19]. Our results confirm that Ovis is the sister group of Myotragus. We also show that the initial radiation of caprines occurred 6.2 ± 0.4 Mya, more recent than the date obtained in other analyses that used very different calibration points [17,18]. Finally, we show that the cytochrome b sequences of alpine and southern chamois, until recently considered to be subspecies, have been separated 1.6 ± 0.3 Mya, much longer than previously suspected. Results and discussion Ancient DNA sequences The complete 1,143 nucleotides of the cytochrome b gene were retrieved in seven overlapping fragments (see primers in Table 1), including the 338 base pairs section between positions 14,312 and 14,649 (using Ovis as reference sequence) already retrieved in a previous work [11] and a section of 85 bp replicated in Oxford (between positions 14,399 and 14,483) (Table 2). Two discrepancies with the previous consensus sequences were observed, a T in position 14,635 and a T in 14,638. The first one was already reported as heteroplasmic in the cloned PCR product in [11], where it was present in five up to ten clones sequenced. The latter was not found in our previous study; therefore, it can correspond to a silent polymorphism within the Myotragus population or to a DNA damage involving a rare A to T change in that particular PCR reaction. All cytochrome b fragments were routinely cloned (not shown), although very few heteroplasmies were detected in the direct sequencing, attributable to both the exceptional preservation of the sample and the short lengths of the fragments; consequently, the error rate in the clones (number of nucleotide differences/1,000 base pairs) was very low (< 2 per thousand base pairs). Considering the overlapping of the primers and the fragment replicated, 404 bp (around 35% of the cytochrome b gene) have been determined from more than one PCR. It cannot be discarded that DNA damage could have affected some positions, although it is unlikely that this would significantly alter the phylogenetic inferences. Table 1 Primer sequences used in this study for different mtDNA and nuclear genes; L and H refers to light and heavy strand, respectively, and numbers refers to the 3' position of the Ovis mtDNA sequence. Primer sequence Amplicon length Cytochrome b gene L14136 5'-GCTTGATATGAAAAACCATCGTTG-3' H14313 5'-TGTGTCGGATGTATAGTGTATTG-3' 176 bp L14310 5'-ATCCTAACAGGCCTATTCCT-3' H14481 5'-CCGATGTTTCATGTTTCTAGGA-3' 170 bp L14475 5'-CGAGGCCTGTACTACGGATC-3' H14650 5'-AACTGAGAATCCGCCTCAG-3' 174 bp L14631 5'-GCTATCCCATACATTGGAAC-3' H14813 5'-GTATARTARGGGTGAAATGG-3' 181 bp L14792 5'-TCCAACAACCCCTCAGGAATTC-3' H14987 5'-TTGATCGTARGATTGCGTATGC-3' 194 bp L14973 5'-CCTCACATCAAACCCGAATG-3' H15159 5'-TCCTCCAATTCATGTGAGTG-3' 185 bp L15152 5'TTCTGAATCCTAGTAGCCGACC-3' H15327 5'-TGCAGTCATCTCCGGTTTACAAGAC-3' 174 bp 12 S gene L599 5'-CTCAAAGGACTTGGCGGTGC-3' H673 5'-GAAGATGGCGGTATATAGAC-3' 73 bp L671 5'-TCACCAATCCTTGCTAATAC-3' H805 5'-AATGGCTTTCGTATTAAATT-3' 133 bp L733 5'-AACAAGAGTAAGCTCAATCA-3' H891 5'-CGGTGTGTGCGTGCTTCATG-3' 157 bp 28 S gene L28S 5'-GGTCGTCCGACCTGGGTATA-3' H28S 5'-TCTAATCATTCGCTTTACCGGAT-3' 96 bp Table 2 Myotragus sequence of the complete mtDNA cytochrome b (between positions 14,159 and 15,301), aligned with the Ovis sequence. Dots indicate sequence identity. OVIS MYOTRAGUS 14159 ATG ATC AAC ATC CGA AAA ACC CAC CCA CTA ATA AAA ATT GTA AAC AAC GCA TTC ATT GAT ... .C. ... T.. ... ... ... A.. ... ... ... ... ... ... ... ... ... ... ... ..C OVIS MYOTRAGUS 14219 CTC CCA GCT CCA TCA AAT ATT TCA TCA TGA TGA AAC TTT GGC TCT CTC CTA GGC ATT TGC ... ... ..C ... ... ..C ..C ... ... ... ... ... ..C ... ..C ... ... ... G.C ... OVIS MYOTRAGUS 14279 TTA ATT TTA CAG ATT CTA ACA GGC CTA TTC CTA GCA ATA CAC TAT ACA CCC GAC ACA ACA ... ..C ... ..A ..C ... ... ... ... ... ... ... ... ... ... ... T.. ... ... ... OVIS MYOTRAGUS 14339 ACA GCA TTC TCC TCT GTA ACC CAC ATT TGC CGA GAC GTG AAC TAT GGC TGA ATT ATC CGA ... ... .G. ... ... ..C G.. ..T ... ... ... ... ..A ... ... ... ... ... ... ... OVIS MYOTRAGUS 14399 TAT ATA CAC GCA AAC GGG GCA TCA ATA TTT TTT ATC TGC CTA TTT ATG CAT GTA GGA CGA ... ... ..T ... ... ..A ... ..C ... ... ..C G.. ... ... ... ... ..C ..G ... ..G OVIS MYOTRAGUS 144590 GGC CTA TAT TAT GGA TCA TAT ACC TTC CTA GAA ACA TGA AAC ATC GGA GTA ATC CTC CTA ... ... ..C ..C ... ... ..C ..T ... ... ... ... ... ... ... ... A.. ... ... ... OVIS MYOTRAGUS 14519 TTT GCG ACA ATA GCC ACA GCA TTC ATA GGC TAT GTC TTA CCA TGA GGA CAA ATA TCA TTC ..C A.A ... ... ..T ... ... ... ... ..T ..C ..T ... ... ... ... ... ... ..T ..T OVIS MYOTRAGUS 14579 TGA GGA GCA ACA GTT ATT ACC AAC CTC CTT TCA GCA ATT CCA TAT ATT GGC ACA AAC CTA ... ... ... ..C ... ..C ... ... ... ..C ... ..T ..C ... ..C ... ..A ..C ..T ..T OVIS MYOTRAGUS 14639 GTC GAA TGA ATC TGG GGA GGA TTC TCA GTA GAC AAA GCT ACC CTC ACC CGA TTT TTC GCC ..A ... ... ... ..A ... ... ... ... ... ... ..G ..C ... ... ..A ... ..C ... ..T OVIS MYOTRAGUS 14699 TTT CAC TTT ATT TTC CCA TTC ATC ATC GCA GCC CTC GCC ATA GTT CAC CTA CTC TTC CTC ... ... ..C ..C C.. ... ..T ..T ... ... ... ... ... ... ..C ... ... ..A ... ... OVIS MYOTRAGUS 14759 CAC GAA ACA GGA TCC AAC AAC CCC ACA GGA ATT CCA TCG GAC ACA GAT AAA ATT CCC TTC ... ... ... ... ... ... ... ... T.. ... ... ... ..A ... G.. ..C ... ..C ..A ..T OVIS MYOTRAGUS 14819 CAC CCT TAT TAC ACC ATT AAA GAC ATC CTA GGT GCT ATC CTA CTA ATC CTC ATC CTC ATG ... ..C ... ... ... ... ... ... ... ... ..C ATG ..A ... ... ..T T.A ... ... ... OVIS MYOTRAGUS 14879 CTA CTA GTA CTA TTC ACG CCT GAC TTA CTC GGA GAC CCA GAC AAC TAC ACC CCA GCA AAC ... ... ... ... ... ..A ..A ... C.. ... ... ... ... ..T ..T ..T ..A ... ..C ... OVIS MYOTRAGUS 14939 CCA CTT AAC ACT CCC CCT CAC ATC AAA CCT GAA TGA TAC TTC CTA TTT GCG TAC GCA ATC ... ..C ... ..A ... ... ... ... ... ..C ... ... ... ... ... ..C ..A ... ... ..T OVIS MYOTRAGUS 14999 TTA CGA TCA ATC CCT AAT AAA CTA GGA GGA GTC CTC GCC CTA ATC CTC TCA ATC CTA GTC C.. ... ... ..T ..C ..A ... ... ... ... ..T ..A ... ... G.. ... ... ..T ..G A.. OVIS MYOTRAGUS 15059 CTA GTA ATT ATA CCC CTC CTC CAT ACA TCA AAG CAA CGG AGC ATA ATA TTC CGA CCA ATC ... ..G C.. ... ..T ..A ... ..C .A. ..C ..A ... T.A ... ... ... ... .AG ... ..T OVIS MYOTRAGUS 15119 AGT CAA TGC ATA TTC TGA ATC CTA GTA GCC GAC CTA TTA ACA CTC ACA TGA ATT GGA GGA ... ... ..T C.. G.. ... ..T T.. ... ..A ... ... C.. ... ... ... ... ... ... ..C OVIS MYOTRAGUS 15179 CAG CCA GTT GAA CAC CCC TAC ATC ATT ATT GGA CAA CTA GCA TCT ATT ATA TAT TTC CTT ... ... ... ... ... ... ..T ..T ... ... ... ..G ... ... ... ... ... ... ... ... OVIS MYOTRAGUS 15239 ATC ATT CTA GTC ATA ATA CCA GTA GCT AGC ATC ATC GAA AAC AAC CTC CTA AAA TGA AGA ... ... ... ... ... ... ... ..G ..G ... .C. ... ... ... ... ... ... ... ... ... OVIS MYOTRAGUS 15299 CAA ... The putative presence of nuclear mtDNA insertions is very unlikely, since we proceed designing the L primers for the next fragment in the sequence already retrieved from the previous fragment, and the H primers in a consensus Caprinae sequence. The 12S gene sequence (305 base pairs) was retrieved in three overlapping fragments (see primers in Table 1). The sequences found in 12S showed several differences from those of extant Caprinae, a fact that points to its authenticity. The PCR of the 28S nuclear gene yielded a very faint band around 140 base pairs and was subsequently cloned and sequenced. In addition, DNA from tissue samples from domestic goat (Capra aegagrus), southern chamois (Rupicapra pyrenaica) and domestic sheep (Ovis aries) was extracted in a separated laboratory, and the same 28S fragment was determined. The 96 base pairs sequence obtained from Myotragus is similar to that found in other Caprinae (Table 3), and clearly different to the human one and to the cow, a suggested source of contamination because of the use of BSA (bovine serum albumine) in ancient DNA amplifications. Moreover, no other Bovids, extinct or living, had been analyzed in the same laboratory when the extraction, amplification, cloning and sequencing of the Myotragus genes was undertaken; therefore, the 28S sequence seems to be endogenous of Myotragus. The cloning of the 28S PCR product (Table 3) showed that two of the sequences (about 20% of the clones) were human contaminants; this accounts for the noise observed in the direct sequencing of the PCR product. Most likely, the human DNA comes from handling of the Myotragus bones during its excavation and posterior morphological study. Nevertheless, the error rate in the clones is relatively low (3.07 per thousand base pairs), a figure within the range found in some modern specimens and well preserved ancient remains, like the moas [20]. Table 3 Myotragus clones and consensus sequence of the nuclear gene 28S rDNA, aligned with sequences of Capra, Ovis and Rupicapra obtained in this study (including Bos as a reference sequence). BOS GGGGCGAAAGACTAATCGAACCATCTAGTAGCTGGTTCCCTCCGAAGTTTCCCTCAGGATAGCTGGCGCTCTCGCA-AACGC-ACAG--ACCC-ACGCAGTTTT M clone1 ............................................................................---...----.-A....-.......... M clone2 ............................................................................---...----.-A....-.......... M Clone3 ..........................................................................T.---...----.-A....-.......... M Clone4 ............................................................................---...----.-A....-.......... M Clone5 ............................................................................---...----.-A....-.......... M Clone6 .......................................T..................................T.---...----.-A....-.......... M Clone7 ............................................................................---...----.-A....-.......... M Clone8 ................................................C...........................---...----.-A....-.......... M CONSENSE ............................................................................---...----.-A....-.......... CAPRA ............................................................................-...C--------....-.......... OVIS ............................................................................-------------....-.......... RUPICAPRA ............................................................................G.-..-G...AG-....-.......... Variation in length is characteristic of 28S gene; size variations are due to expansions or contractions of variable segments in 10–12 positions within the gene, where variation does not interfere with ribosomal function [21]. The sequences were aligned by eye; due to the high interspecific variation in 28S, no definite alignment is possible and therefore, the phylogenetic information obtained from this gene is limited. However, the retrieval of a nuclear gene indicates the quality of the DNA of the sample used. Furthermore, it opens new possibilities of research for assessing the phylogenetic relationships of Myotragus and for retrieving nuclear genes with phenotypic implications. Phylogeny of the Caprinae subfamily and the position of Myotragus in the tree We included Pantholops hodgsoni in the phylogenetic analysis since previous analysis as well as ours indicate that this species is clearly associated to other members of the subfamily Caprinae. As an outgroup, we used 14 members of the tribes Alcephalini and Hippotragini (see Methods) previously detected as the closest members within the Bovidae [17,22]. The cytochrome b and cytochrome b + 12S Bayesian and maximum-likelihood trees (Figure 1) showed a topology roughly similar to that found in previous studies on Caprinae phylogeny (e.g., [22]). Recently, it has been reported that the Budorcas cytochrome b sequence available at GenBank was in fact a chimaera that included a fragment of Ovis sequence, and a new sequence for this species was provided [17]. Crucially, in previous works Myotragus formed a quite stable clade with Budorcas and Ovis that is not maintained anymore, with Budorcas now clustering in a different position. Figure 1 Bayesian tree (a) and maximum-likelihood tree (b) of the entire cytochrome b of all Caprinae species in databases. Bayesian tree (c) and maximum-likelihood tree (d) of the concatenated cytochrome b plus a 12S fragment of all Caprinae species in databases with both sequences. Bootstrap per cent values higher than 40% (maximum-likelihood tree) and support values bigger than 0.4 (Bayesian tree) are depicted in the nodes. The scale bar represents 0.1 substitutions/site. Examination of the cytochrome b trees (Figure 1a and 1b) shows that some of the clades previously found are present in these trees, specially Myotragus+Ovis (67.5% in the maximum-likelihood trees), Capricornis+Ovibos+Naemorhedus (100% in the maximum-likelihood tree) and Hemitragus+Pseudois+Capra (94.6% in the maximum-likelihood tree). The position of Hemitragus, that precludes the monophyly of the genus Capra, is problematic, as some authors have already noticed [18]. Some species show unstable positions in the trees, particularly the genera Budorcas, Ammotragus, Pantholops, Rupicapra and Oreamnos. The problematic position of these taxa has been already reported on extant species studies (see, for instance [22,23]). However, our tree, although still with low bootstrap values, has been able to better resolve the phylogenetic position of Myotragus, which is basal to the Ovis clade; the support value for the Myotragus+Ovis grouping is now 0.46 in the Bayesian tree and the bootstrap value 67.5% in the maximum-likelihood tree, respectively [11]. The trees with the cytochrome b + 12S fragment (305 base pairs) did not significantly improve the phylogeny (Figure 1c and 1d); the Capricornis+Ovibos and Hemitragus+Pseudois+Capra clades are well supported by bootstrap analysis, while the support of the cluster of Myotragus+Ovis is reduced with respect to the cytochrome b sequence alone. This discrepancy can be attributed to the small number of informative positions added by the 12S fragment together with the reduced number of species that could be used due to the unavailability of the 12S sequence in many species. Evolutionary rates and diversification of Caprinae The complete resolution of the Caprinae phylogeny cannot be achieved with the cytochrome b gene alone, as some species of the subfamily (specially of the genera Budorcas, Oreamnos, Rupicapra, Ammotragus and Pantholops) still have an unstable position. It is likely that the lack of resolution in the basal branches is due to the existence of a very quick initial radiation in the Caprinae subfamily. To test this, we performed a likelihood ratio test for zero branch-length of all branches of the tree, that showed that there are nine branches with p > 0.05 (and therefore, not significantly different from 0). Five of them belong to the basal branches of the Caprinae tree, thus supporting the hypothesis of a very fast initial radiation of this group. Additionally, the plot of grow of lineages of an ultrametric tree (see below) also supports this fast initial radiation. Being Myotragus a small caprine, its long branch in the maximum-likelihood tree previously detected with a cytochrome b fragment [11] was attributed to an earlier age of first reproduction and a shorter generation time in Myotragus than in other wild caprines [11]. This long branch is also present in the trees of the complete cytochrome b (Figure 1a and 1b). However, recent estimates indicate that Myotragus was actually bigger than previously believed, with a weight ranging between 15 and 25 kg for the smallest adult specimens and between 40 and 60 kg for the larger specimens [24]. Therefore the reason for this long branch cannot be due to its presumed extreme reduced size (and the consequences that this involves). On the other hand, the hypothesis that tries to explain differences in evolutionary rates as due to differences in metabolic rate or generation time (both correlated to body mass) was not supported in the analysis of cytochrome b evolutionary rates of a wide set of mammalian species [25]. It is thus more likely that this relatively long branch of Myotragus in the Caprinae tree is mainly a stochastic phenomenon. Dating of the Caprinae phylogeny Due to the endemicity of Myotragus [2], it can be hypothesized that this lineage diverged from the continental species 5.35 million years ago (Mya), when the desiccation of the Mediterranean ended [13,14] and the Myotragus ancestors became isolated in the Balearic Islands. In principle, this would be a minimum age for the node in the tree, but we consider likely that the isolation of the Balearic Islands was a vicariant event responsible for the split of the Myotragus and continental lineages, so that we can use this date as a fixed age in the Caprinae cytochrome b tree (Figure 2). The dating results for the maximum-likelihood and Bayesian trees are very similar and only the former are shown. We have calculated ultrametric trees using the Langley-Fitch method [26] as well as the penalyzed likelihood method with several smoothing factors [27] implemented in the r8s program. The cross-validation procedure to explore the fidelity with which these different methods explain the branch length variation in the tree indicates that the Langley-Fitch method is slightly better for our data set and therefore we used this method in the calculations described next (although dates obtained were very similar using penalyzed likelihood with a log10 of the smoothing factor of 2; not shown). In addition, we have performed a parametric bootstrap procedure, in which we simulated alignments from the maximum-likelihood tree and re-calculated dates from the same maximum-likelihood topology with branch lengths optimized for each simulated alignment. This allowed us to evaluate the stochastic errors of date estimates associated to sampling a finite number of alignment positions [28]. In a different parametric bootstrap procedure, we reconstructed new trees from each simulated alignment and used them for time estimation (instead of only the original topology) and therefore errors associated with finite number of positions as well as with tree reconstruction imperfections are estimated [28]. The standard deviations and 95% confidence intervals for most date estimates were very similar for both parametric bootstrap procedures (Table 4), indicating that most of the errors in the present data set are associated to the finite sample sequence rather than to tree errors. In the following, standard deviations given refer to the parametric bootstrap using only the original topology for calculating dates. Table 4 Dates in Mya, standard deviations, and 95% intervals (in parenthesis) obtained with parametric bootstrap using the original topology (Single topology) or new tree calculations (Several topologies) for each simulation of the mtDNA cytochrome b of Caprinae. The calibration is based on the isolation of Myotragus in the Balearic Islands at 5.35 million years ago. Lineage Single topology Several topologies Capra 1.5 ± 0.2 (1.2 – 1.9) 1.5 ± 0.2 (1.1 – 1.9) Capra ibex/C. pyrenaica 0.6 ± 0.1 (0.4 – 0.9) 0.6 ± 0.1 (0.3 – 0.8) Rupicapra pyrenaica/R. rupicapra 1.6 ± 0.3 (1.1 – 2.1) 1.5 ± 0.3 (1.1 – 2.1) Ovis 2.2 ± 0.3 (1.7 – 2.8) 2.1 ± 0.3 (1.6 – 2.8) Root 6.2 ± 0.4 (5.5 – 7) 6.1 ± 0.5 (5.4 – 7.2) Figure 2 Ultrametric tree of Caprinae derived from the cytochrome b maximum-likelihood tree. The geological date of the isolation of the Myotragus lineage in the Balearic islands 5.35 Mya, that was used as a calibration point, is highlighted. Calculated dates and standard deviations are given for some important splits in the tree. A lineages through time plot is shown under the tree. The number of lineages in each node is show in logarithmic scale. The horizontal scale bar represents 1 Mya for both the tree and the lineages through time plot. The ultrametric trees calculated by the Langley-Fitch method applied to the cytochrome b maximum-likelihood tree showed that the separation of the most divergent lineages within Caprinae (Pantholops and the rest of species) occurred 6.2 ± 0.4 Mya. These dates are more recent than a proposed Caprinae radiation 11 Mya based on different molecular calibrations (see Introduction) [17,18]. There is no general consensus from fossil data on the date of the start of the Caprinae radiation, that ranges from around 14 Mya [12] to a more recent Late Miocene origin [29], which would be more consistent with our results. After the first split, the separation of the most basal lineages took place around or before 5.35 Mya (Figure 2), and therefore we can estimate that the main radiation of caprines occurred in less than one million years of evolution. A lineages through time plot, that reflects the number of lineages in each split point of the tree (bottom of Figure 2), indicates a fast growing of lineages during the first million years of evolution in comparison with the flatter slope in the rest of the tree. Although a few species are missing in this tree, all genera are covered and it is likely that the tree with all species will reveal the same clear steep slope at its base. This truly supports the existence of a quite fast initial radiation that led to the differentiation of nine clades (including the Myotragus lineage) that are at the base of all extant genera. In addition, this explains the difficulties we face in reconstructing the phylogenetic tree of this group. Interestingly, the separation of the two chamois species (Rupicapra) and the beginning of the radiation of the main clade of goats (separation of the Capra ibex and Capra pyrenaica clade from the rest of species) occurred at 1.6 ± 0.3 Mya and 1.5 ± 0.2, respectively, according to the calibrated cytochrome b tree. This would imply that the beginning of the Pleistocene glaciations promoted the diversification of the main lineages of Caprinae species living in the Alps and other Eurasian mountainous regions. The deep split observed between the alpine and southern chamois, at 1.6 ± 0.3 Mya, is quite remarkable, since these two species have been considered subspecies until recently [30], and other molecular studies have given much more recent dates for the separation of these two species using restriction fragment length polymorphism of mitochondrial DNA [31] or microsatellites [32]. Their split is even older than that between Capra pyrenaica and Capra ibex, a pair of species with a similar geographic disjunction than the two chamois and traditionally recognized as different species, that separated 0.6 ± 0.1 Mya according to our data. To test whether the deep split observed between the two chamois was caused by particularly divergent cytochrome b haplotypes of chamois used in this study, we performed the calibrations with all sequences of chamois in databases, including shorter ones (five for R. pyrenaica and six for R. rupicapra), obtaining very similar dates (not shown). To analyze if the date of this split depended on the particularly long branch of Myotragus, we manually reduced this branch to the half, but we obtained again similar dates for all splits in the trees (1.7 Mya for the split between chamois). This is due to the smoothing of rates performed by the calibration methods used, that correct for the higher evolutionary rate of the Myotragus lineage. To account for the possibility that Myotragus could be more basal in the tree, we manually positioned Myotragus just after the separation of Pantholops. Then the split date of the two chamois was reduced, but only up to 1.4 Mya. This is in agreement with the parametric bootstrap analyses (Table 4), that indicate that the tree topology is not crucial in this calculation. This is surely due to the small internal branches involved in all different topologies arranging the basal species of the tree. Other works where several topologies were used for calculating dates have also shown that age estimates were largely insensitive to different phylogenies [28]. Finally, if Myotragus and Ovis had diverged in fact more recently and the ancestors of Myotragus had reached the Balearic Islands via a transmarine colonization after these islands became disconnected from the continent, as proposed for other vertebrates [32], all the dates in the tree could be more recent; however, this hypothesis is extremely unlikely in Myotragus from the paleontological evidence [24]. On the other hand, the use of evolutionary rates calculated from other sources, for example, the separation of humans and chimpanzees 5 – 7 Mya [34,35] combined with the genetic distance between their cytochrome b sequence, would lead to a separation of Myotragus and Ovis clades at 7.2 Mya. Although the use of an external calibration could be problematic, this reference would not support a transmarine colonization of Myotragus ancestors. Thus, the calibrated cytochrome b trees indicate that alpine and southern chamois have been separated much longer than previously thought, which would help explain the anti-hybridization mechanisms currently operating [30]. Conclusion In conclusion, this study is among the first ones in which mitochondrial as well as nuclear DNA has been retrieved from an extinct species from a warm location, in this case the Mediterranean area. Also, our research demonstrates the importance of ancient DNA in phylogenetics, since we have been able to use the isolation of Myotragus in the Baleric Islands as a vicariant event for dating the Caprinae phylogenetic tree. This is quite a unique opportunity for mammals because, contrary to other zoological groups such as insects, reptiles or amphibians, almost no extant endemic mammal remains in these Islands. Methods A left tibiae (MNIB 60176) bone from Myotragus balearicus from Cova des Gorgs (Escorca, Mallorca), dated to 6,010-5,830 cal BC 2σ (Beta-177239) was chosen for DNA analysis because of its excellent external preservation. The remainder bone not used for analysis is held in the vertebrate collection in the Museu de la Naturalesa de les Illes Balears (MNIB, Palma de Mallorca); the present sample is approximately between 500 and 1,850 years younger than the other one from the same site previously analyzed [11]. DNA was extracted following procedures described elsewhere [11]. The sample was demineralized overnight with 0.5 M EDTA, incubated overnight with proteinase K and SDS, extracted with phenol-chloroform and concentrated and desalinized with centricon columns. Finally, the extract was purified with silica. The sample was extracted with blank extractions to monitor against contamination. Several fragments of mtDNA cytochrome b gene were amplified and sequenced until completing the whole gene (Table 1). The cytochrome b sequences matched the partial sequence previously retrieved from another Myotragus bone from the same site [10]; multiple controls, such as independent replication of the results in two different laboratories and cloning of different overlapping PCR products, contributed to the authentication of the sequences. All the work was carried out in a dedicated ancient DNA laboratory, with UV lights, air-positive pressure and regular cleaning of surfaces with bleach. No signs of contamination were observed along the study in the extraction and PCR blanks; the sample was analyzed along with an Upper Paleolithic human remain [36] that yielded a human sequence. In addition, no other Bovid had ever been extracted in the same laboratory. One fragment of the cytochrome b was previously replicated in the Ancient Biomolecules Centre (Oxford), yielding the same sequence as in the present study. For the 12S gene, some primers were designed (Table 1) to match previous Caprinae sequences published [37]. Due to the exceptional quality of this extract, a pilot project was launched for trying to retrieve a fragment of a nuclear gene. The 28S rDNA was chosen because it is present in some hundreds of copies in the genome. The primers used to amplify the 28S rDNA were those designed [38] to be vertebrate-specific (Table 1). Five microliters of extract were added to 20 μl PCR reactions, containing 1× reaction buffer, 1 unit of taq DNA polymerase, 2.5 mM of MgCl2, 25 pmol of each primer. Forty cycles of 1 min at 94°C, 1 min at 50°C and 1 min at 72°C were performed. PCR products were resolved in 1% low-melting agarose gels in a TA buffer; bands were excised from the gel and subjected to a second 30 cycles of PCR with limiting reagents. PCR products were cloned with the SureClone Ligation Kit (Pharmacia, Upsala, Sweden); inserts were sequenced with 3100 Gene Analyzer (Applied Biosystems). Sequences have been deposited at the GenBank under accession numbers: AY380560, AY380561, AY380562, AY380563, AY894418, AY894419 and AY894420. Several phylogenetic analyses were performed with the Myotragus sequences. We used its cytochrome b together with the whole cytochrome b of all Caprinae species available in databases (Capra hircus, Capra aegagrus, Capra falconeri, Capra caucasica, Capra cylindricornis, Capra nubiana, Capra ibex, Capra sibirica, Hemitragus jemlahicus, Pseudois nayaur, Pseudois schaeferi, Ovis ammon, Ovis aries, Ovis dalli, Budorcas taxicolor, Myotragus balearicus, Rupicapra pyrenaica, Rupicapra rupicapra, Capricornis crispus, Ovibos moschatus, Naemorhedus caudatus, Pantholops hodgsoni, Oreamnos americanus), plus the 987 base pairs sequence of Capra pyrenaica and the complete sequence of 14 members of Alcephalini and Hippotragini used as outgroups (Addax nasomaculatus, Alcelaphus buselaphus, Alcelaphus lichtensteini, Beatragus hunteri, Connochaetes gnou, Connochaetes taurinus, Damaliscus lunatus, Damaliscus pygargus, Hippotragus equinus, Hippotragus niger, Oryx dammah, Oryx gazella, Oryx leucoryx and Sigmoceros lichtensteinii); this high number of outgroups eliminates the possible biases of choosing a single one at random and partitions the longest branches, thus reducing the problem of long branch attraction [39]. The analyses considered a general time reversible (GTR) model with a proportion of invariable sites and a discrete gamma model with 6 categories to account for among site rate variation. This model was selected upon checking with Modeltest [40] that the set of all positions as well as the 1st, 2nd and 3rd codon positions favored the GTR model or a model close to it (GTR being the closest model in the phylogeny programs used here) and gamma plus invariable rates. Phylogenies were performed using Bayesian analysis implemented in the program MrBayes v. 3.0. [41]. For the calculations, independent models of sequence evolution were used for each codon position. Four chains of 4,000,000 trees were generated, sampling every 100th tree, with burning completed by the 20,000th tree; thus, 20,000 trees were used to estimate topologies and posterior probabilities of parameters. It was checked that the likelihood and all parameters were stationary after this burn-in. Maximum-likelihood analyses were performed with the program Phyml, version 2.4.4 [42]. The gamma parameter, the proportion of invariable sites and the nucleotide frequencies were estimated from the data. Two maximum-likelihood tree searches were made, one starting from a neighbor joining tree and another from the previously generated Bayesian tree, selecting the tree with the best likelihood (almost always the one starting from the Bayesian tree). To test the robustness of the tree clades, 1,000 bootstrap samples were performed. Maximum-likelihood and Bayesian trees were similarly generated in a combination of the cytochrome b sequence with a 12S fragment, using the only 20 caprine species in which both 12S and cytochrome b were available at GenBank, and an outgroup. For the Bayesian tree, an independent GTR model was considered for the 12S partition. No tree was generated for the 28S fragment, due to irresolvable difficulties with the alignment. To test the possibility of an explosive diversification of the Caprinae subfamily, we performed the likelihood ratio test for zero-length branches [43] from the cytochrome b data using PAUP 4,0 [44] and the same maximum-likelihood model as above. To estimate absolute rates of molecular evolution with a relaxed molecular clock on the cytochrome b phylogenetic tree (and divergence times using the geological date of the isolation of the Myotragus lineage), the r8s v1.7 program was used [45]. Outgroup sequences were eliminated from the tree previous to the analysis of caprines. Smoothing of rate variation along the tree was performed with the Langley-Fitch [26] and penalized likelihood [27] methods. Sixteen smoothing factors with log10 from -2 to 5.5 were used for the penalized likelihood method. The lowest χ2 cross-validation score, as calculated by r8s, was used to select the best method. To perform parametric bootstraps, 1,000 Monte Carlo simulations of alignments of the same length than the complete cytochrome b were generated with SeqGen [46] using the phylogenetic trees and model parameters previously obtained. A lineages through time plot of an ultrametric tree obtained with these methods was calculated with the program GENIE [47]. Authors' contributions CL-F and LS did the experimental work, JC carried out the phylogenetic and dating analyses, TM-B carried out some of the analyses, JAA provided the sample and critical feedback to CL-F, JB provided laboratory facilities. CL-F, JC and JB contributed to the writing of the manuscript. Acknowledgements This research was supported by the Dirección General de Investigación, Ministerio de Ciencia y Tecnología of Spain (grants BFU2004-02002 and BOS2003-08070), by Departament d'Universitats, Recerca i Societat de la Informació, Generalitat de Catalunya (grant 2001SGR00285) and a fellowship to LS. We are grateful to Alan Cooper and Beth Saphiro (Ancient Biomolecules Center, Oxford), for the partial replication of some ancient sequences, to Alexandre Hassanin (Muséum National d'Histoire Naturelle, Paris) for advice in the designing of the 12S Myotragus primers and to Salvador Carranza (Universitat de Barcelona) for critical comments on the manuscript. ==== Refs Bate DMA A new artiodactyle from Majorca Geol Mag NS 1909 6 385 388 Alcover JA Moyà-Solà S Pons-Moyà J Les quimeres del passat Els vertebrats fòssils del Plio-Quaternari de les Balears i les Pitiüses 1981 Palma de Mallorca: Ed. Moll Alcover JA Perez-Obiol R Yll E-I Bover P The diet of Myotragus balearicus Bate 1909 (Artiodactyla: Caprinae), an extinct bovid from the Balearic Islands: evidence from coprolites Biol J Linn Soc 1999 66 57 74 10.1006/bijl.1998.0260 Alcover JA Seguí B Bover P MacPhee RDE Extinctions and Local Disappearances of Vertebrates in the Western Mediterranean Islands Extinctions in Near Time 1999 New York: Kluwer Academic/Plenum Publishers Bover P Alcover JA Estimating physical characteristics of neonate Myotragus balearicus (Artiodactyla, Caprinae) Deinsea – Annual of the Natural History Museum Rotterdam 1999 7 33 54 Bover P Alcover JA The evolution and ontogeny of the dentition of Myotragus balearicus Bate 1909 (Artiodactyla, Caprinae): evidence from new fossil data Biol J Linnean Soc 1999 68 401 428 10.1006/bijl.1998.0294 Ramis D Alcover JA Revisiting de earliest human presence in Mallorca, Western Mediterranean Proc Prehistoric Soc 2001 67 261 269 Bover P Alcover JA Understanding Late Quaternary Extinctions: the case of Myotragus balearicus Bate 1909 J Biogeograph 2003 30 771 782 10.1046/j.1365-2699.2003.00872.x Wilson DE Reeder DM Mammal Species of the World: A Taxonomic and Geographic Reference 1993 2 Washington: Smithsonian Institution Press Lalueza-Fox C Bertranpetit J Alcover JA Shailer N Hagelberg E Mitochondrial DNA from Myotragus balearicus, an extinct Bovid from the Balearic Islands J Exp Zool (Mol Dev Evol) 2000 288 56 62 10.1002/(SICI)1097-010X(20000415)288:1<56::AID-JEZ6>3.0.CO;2-X Lalueza-Fox C Shapiro B Bover P Alcover JA Bertranpetit J Molecular phylogeny and evolution of the extinct bovid Myotragus balearicus Mol Phylogenet Evol 2002 25 501 510 12450754 10.1016/S1055-7903(02)00290-7 Vrba ES Africa Bovidae: evolutionary events since the Miocene S Afr J Sci 1985 81 263 266 Krijgsman W Hilgen FJ Raffi I Sierro FJ Wilson DS Chronology, causes and progression of the Messinian salinity crisis Nature 1999 400 652 654 10.1038/23231 Duggen S Hoernle K van den Bogaard P Rupke L Morgan JP Deep roots of the Messinian salinity crisis Nature 2003 422 602 606 12686997 10.1038/nature01553 Vogel P Cosson JF Lopez Jurado LF Taxonomic status and origin of the shrews (Soricidae) from the Canary islands inferred from a mtDNA comparison with the European Crocidura species Mol Phylogenet Evol 2003 27 271 82 12695091 10.1016/S1055-7903(02)00403-7 Mitchell-Jones A J Amori G Bogdanowicz W Krystufek B Reijnders PJH Spitzenberg F Stubbe M Thissen JBM Vohralik V Zima J The atlas of European mammals 1999 London: T & AD Poyser Ropiquet A Hassanin A Molecular phylogeny of caprines (Bovidae, Antilopinae): the question of their origin and diversification during the Miocene J Zoolog Syst Evol Res 2005 43 49 60 10.1111/j.1439-0469.2004.00290.x Ropiquet A Hassanin A Molecular evidence for the polyphyly of the genus Hemitragus (Mammalia Bovidae) Mol Phylogenet Evol 2005 36 154 168 15904863 10.1016/j.ympev.2005.01.002 Poinar H Kuch M McDonald G Martin P Pääbo S Nuclear gene sequences from a late pleistocene sloth coprolite Curr Biol 2003 13 1150 1152 12842016 10.1016/S0960-9822(03)00450-0 Cooper A Lalueza-Fox C Anderson S Rambaut A Austin J Ward R Complete mitochondrial genome sequences of two extinct moas clarify ratite evolution Nature 2001 409 704 707 11217857 10.1038/35055536 Laudien Gonzalez I Gorski JL Campen TJ Dorney DJ Erickson JM Sylvester JE Schmickel RD Variation among human 28S ribosomal RNA genes Proc Natl Acad Sci USA 1985 82 7666 7670 3865188 Hassanin A Douzery EJP The tribal radiation of the family Bovidae (Artiodactyla) and the evolution of the mitochondrial Cytochrome b gene Mol Phylogenet Evol 1999 13 227 243 10603253 10.1006/mpev.1999.0619 Groves P Shields GF Phylogenetics of the Caprinae based on cytochrome b sequence Mol Phylogenet Evol 1996 5 467 476 8744761 10.1006/mpev.1996.0043 Bover P Noves aportacions al coneixement del gènere Myotragus Bate 1909 (Artiodactyla, Caprinae) de les Illes Balears 2004 PhD thesis, University of the Balearic Islands Castresana J Cytochrome b phylogeny and the taxonomy of Great Apes and Mammals Mol Biol Evol 2001 18 465 471 11264397 Langley CH Fitch WM An examination of the constancy of the rate of molecular evolution J Mol Evol 1974 3 161 177 4368400 10.1007/BF01797451 Sanderson MJ Estimating absolute rates of molecular evolution and divergence times: a penalized likelihood approach Mol Biol Evol 2002 19 101 109 11752195 Sanderson MJ Doyle JA Sources of error and confidence intervals in estimating the age of angiosperms from rbcL and 18S rDNA data Amer J Bot 2001 88 1499 1516 Gentry AW The Miocene differentiation of Old World Pecora (Mammalia) Historical Biol 1994 7 115 158 Masini F Lovari S Systematics, phylogenetic relationships, and dispersal of the chamois (Rupicapra spp.) Quaternary Research 1988 30 339 349 Hammer S Nadlinger K Hartl GB Mitochondrial DNA differentiation in chamois (genus Rupicapra implications for taxonomy, conservation, and management Acta Theriologica Suppl 1995 3 145 155 Perez T Albornoz J Dominguez A Phylogeography of chamois (Rupicapra spp.) inferred from microsatellites Mol Phylogenet Evol 2002 25 524 534 12450756 10.1016/S1055-7903(02)00296-8 Martinez-Solano I Gonzalbes HA Arntzen JW Garcia-Paris M Phylogenetic relationships and biogeography of midwife toads (Discoglossidae: Alytes J Biogeogr 2004 31 603 618 Yoder AD Yang Z Estimation of primate speciation dates using local molecular clocks Mol Biol Evol 2000 17 1081 1090 10889221 Glazko GV Nei M Estimation of divergence times for major lineages of primate species Mol Biol Evol 2003 20 424 434 12644563 10.1093/molbev/msg050 Caramelli D Lalueza-Fox C Vernesi C Lari M Casoli A Mallegni F Chiarelli B Dupanloup I Bertranpetit J Barbujani G Bertorelle G Evidence for a genetic discontinuity between Neandertals and 24,000-year-old anatomically modern Europeans Proc Natl Acad Sci USA 2003 100 6593 6597 12743370 10.1073/pnas.1130343100 Gatesy J Amato G Vrba E Schaller G DeSalle R A cladistic analysis of mitochondrial ribosomal DNA from the Bovidae Mol Phylogenet Evol 1997 7 303 319 9187090 10.1006/mpev.1997.0402 Greenwood A Capelli C Possnert G Pääbo S Nuclear DNA sequences from Late Pleistocene megafauna Mol Biol Evol 1999 16 1466 1473 10555277 Sanderson MJ Shaffer HB Troubleshooting molecular phylogenetic analyses Annu Rev Ecol Syst 2002 33 49 72 10.1146/annurev.ecolsys.33.010802.150509 Posada D Crandall KA MODELTEST: testing the model of DNA substitution Bioinformatics 1998 14 817 818 9918953 10.1093/bioinformatics/14.9.817 Huelsenbeck JP Ronquist F MRBAYES: Bayesian inference of phylogenetic trees Bioinformatics Application Notes 2001 17 754 755 Guindon S Gascuel O A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood Syst Biol 2003 52 696 704 14530136 10.1080/10635150390235520 Slowinski JB Molecular polytomies Mol Phylogenet Evol 2001 19 114 120 11286496 10.1006/mpev.2000.0897 Swofford DL PAUP: phylogenetic analysis using parsimony (and other methods) 1998 Version 4. Sunderland, Mass: Sinauer Sanderson MJ r8s: inferring absolute rates of molecular evolution and divergence times in the absence of a molecular clock Bioinformatics 2003 19 301 302 12538260 10.1093/bioinformatics/19.2.301 Rambaut A Grassly NC Seq-Gen: an application for the Monte Carlo simulation of DNA sequence evolution along phylogenetic trees Comput Appl Biosci 1997 13 235 238 9183526 Pybus OG Rambaut A GENIE: estimating demographic history from molecular phylogenies Bioinformatics 2002 18 1404 1405 12376389 10.1093/bioinformatics/18.10.1404	16332256	PMC1325260	CC BY	2021-01-04 16:29:17	no	BMC Evol Biol. 2005 Dec 6; 5:70	utf-8	BMC Evol Biol	2,005	10.1186/1471-2148-5-70	oa_comm
==== Front BMC Evol BiolBMC Evolutionary Biology1471-2148BioMed Central London 1471-2148-5-701633225610.1186/1471-2148-5-70Research ArticleMolecular dating of caprines using ancient DNA sequences of Myotragus balearicus, an extinct endemic Balearic mammal Lalueza-Fox Carles [email protected] Jose [email protected] Lourdes [email protected]ès-Bonet Tomàs [email protected] Josep Antoni [email protected] Jaume [email protected] Unitat de Biologia Evolutiva, Facultat de Ciències de la Salut i de la Vida, Universitat Pompeu Fabra, Doctor Aiguader 80, 08003 Barcelona, Spain2 Unitat d'Antropologia, Dept. Biologia Animal, Facultat de Biologia, Universitat de Barcelona, Avda. Diagonal 645, 08028 Barcelona, Spain3 Department of Physiology and Molecular Biodiversity, Institut de Biologia Molecular de Barcelona, CSIC, 08034 Barcelona, Spain4 Institut Mediterrani d'Estudis Avançats de les Illes Balears (CSIC-UIB), Cta. de Valldemosa km 7.5, 07071 Ciutat de Mallorca, Spain2005 6 12 2005 5 70 70 11 6 2005 6 12 2005 Copyright © 2005 Lalueza-Fox et al; licensee BioMed Central Ltd.2005Lalueza-Fox 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 Myotragus balearicus was an endemic bovid from the Balearic Islands (Western Mediterranean) that became extinct around 6,000-4,000 years ago. The Myotragus evolutionary lineage became isolated in the islands most probably at the end of the Messinian crisis, when the desiccation of the Mediterranean ended, in a geological date established at 5.35 Mya. Thus, the sequences of Myotragus could be very valuable for calibrating the mammalian mitochondrial DNA clock and, in particular, the tree of the Caprinae subfamily, to which Myotragus belongs. Results We have retrieved the complete mitochondrial cytochrome b gene (1,143 base pairs), plus fragments of the mitochondrial 12S gene and the nuclear 28S rDNA multi-copy gene from a well preserved Myotragus subfossil bone. The best resolved phylogenetic trees, obtained with the cytochrome b gene, placed Myotragus in a position basal to the Ovis group. Using the calibration provided by the isolation of Balearic Islands, we calculated that the initial radiation of caprines can be dated at 6.2 ± 0.4 Mya. In addition, alpine and southern chamois, considered until recently the same species, split around 1.6 ± 0.3 Mya, indicating that the two chamois species have been separated much longer than previously thought. Conclusion Since there are almost no extant endemic mammals in Mediterranean islands, the sequence of the extinct Balearic endemic Myotragus has been crucial for allowing us to use the Messinian crisis calibration point for dating the caprines phylogenetic tree. ==== Body Background Myotragus balearicus [1] was an extremely modified caprine, endemic of the Balearic Islands, characterized by a series of unusual apomorphies developed throughout more than five million years of insular evolution [2-7]. Myotragus disappeared from Mallorca between 3,700 and 2,040 years BC [8], probably after the arrival of the first humans to the Islands. The morphological peculiarities of Myotragus, including extreme size reduction, a single ever growing lower incisor, modified limb bones and frontal eyes [2], makes it difficult to clarify its taxonomic position [9]. In two previous studies [10,11], we obtained bits of the mtDNA cytochrome b gene from two different Myotragus bone specimens from Mallorca island, one found in cova Estreta (Pollença) and the other found in Cova des Gorgs (Escorca). Although the quick radiation of all the Caprinae [12] and the short cytochrome b fragment retrieved (338 bp) made the previous analysis difficult, our data indicated that Myotragus was genetically close to the sheep (Ovis) group. There is a general agreement that the continental ancestor of the Balearic bovid possibly settled in Mallorca and Menorca during the Messinian period. The Balearic Islands were last connected to the continent during the Messinian regression, when the Mediterranean basin was dried out, allowing faunal exchanges between the islands and the continental lands [13,14]. The opening of the Strait of Gibraltar 5.35 Mya definitely isolated the Balearic Islands and promoted the beginning of the independent evolution of Myotragus [2]. The Myotragus lineage spreads from the Pliocene to the Holocene, and includes in Mallorca five chronospecies: M. pepgonellae, M. antiquus, M. kopperi, M. batei and M. balearicus. No absolute datings are available for the oldest species, M. pepgonellae, although a chronology of Lower Pliocene has been proposed; a paleomagnetic data of about 2.6 My is available for M. antiquus remains [2]. Apart from two shrews from the islands of Crete and Sicily, respectively, no other Pliocene endemic mammal persists in Mediterranean islands [15,16]. Thus, the accurately dated vicariant event that isolated Myotragus from their continental relatives constitutes a unique opportunity to use the sequences of Myotragus for calibrating the DNA molecular clock of caprines. Previously, only dates outside caprines (the emergence of the family Bovidae 18.5 Mya), or dates that are too recent for accurate estimations (common ancestry of some domestic sheep breeds a few thousand years ago) could be used [17,18]. In addition, since the molecular clock is not perfect, it is necessary to use as much sequence information as possible to obtain reliable date estimates. For this purpose, we decided to retrieve the complete cytochrome b gene and some other genetic markers of Myotragus; to do this, we used a newly excavated Myotragus bone, obtained in 2002 from Cova des Gorgs, that looked macroscopically very well preserved and therefore was suggestive of DNA survival. We designed different sets of overlapping primers for retrieving the complete cytochrome b gene, as well as a 305 bp fragment from another mtDNA gene, the 12S. In addition, we retrieved a short fragment of a nuclear gene, 28S rDNA, from the same extract. The only precedent of nuclear DNA retrieval from warm climates is, to our knowledge, the analysis of a ground sloth coprolite from the south of United States [19]. Our results confirm that Ovis is the sister group of Myotragus. We also show that the initial radiation of caprines occurred 6.2 ± 0.4 Mya, more recent than the date obtained in other analyses that used very different calibration points [17,18]. Finally, we show that the cytochrome b sequences of alpine and southern chamois, until recently considered to be subspecies, have been separated 1.6 ± 0.3 Mya, much longer than previously suspected. Results and discussion Ancient DNA sequences The complete 1,143 nucleotides of the cytochrome b gene were retrieved in seven overlapping fragments (see primers in Table 1), including the 338 base pairs section between positions 14,312 and 14,649 (using Ovis as reference sequence) already retrieved in a previous work [11] and a section of 85 bp replicated in Oxford (between positions 14,399 and 14,483) (Table 2). Two discrepancies with the previous consensus sequences were observed, a T in position 14,635 and a T in 14,638. The first one was already reported as heteroplasmic in the cloned PCR product in [11], where it was present in five up to ten clones sequenced. The latter was not found in our previous study; therefore, it can correspond to a silent polymorphism within the Myotragus population or to a DNA damage involving a rare A to T change in that particular PCR reaction. All cytochrome b fragments were routinely cloned (not shown), although very few heteroplasmies were detected in the direct sequencing, attributable to both the exceptional preservation of the sample and the short lengths of the fragments; consequently, the error rate in the clones (number of nucleotide differences/1,000 base pairs) was very low (< 2 per thousand base pairs). Considering the overlapping of the primers and the fragment replicated, 404 bp (around 35% of the cytochrome b gene) have been determined from more than one PCR. It cannot be discarded that DNA damage could have affected some positions, although it is unlikely that this would significantly alter the phylogenetic inferences. Table 1 Primer sequences used in this study for different mtDNA and nuclear genes; L and H refers to light and heavy strand, respectively, and numbers refers to the 3' position of the Ovis mtDNA sequence. Primer sequence Amplicon length Cytochrome b gene L14136 5'-GCTTGATATGAAAAACCATCGTTG-3' H14313 5'-TGTGTCGGATGTATAGTGTATTG-3' 176 bp L14310 5'-ATCCTAACAGGCCTATTCCT-3' H14481 5'-CCGATGTTTCATGTTTCTAGGA-3' 170 bp L14475 5'-CGAGGCCTGTACTACGGATC-3' H14650 5'-AACTGAGAATCCGCCTCAG-3' 174 bp L14631 5'-GCTATCCCATACATTGGAAC-3' H14813 5'-GTATARTARGGGTGAAATGG-3' 181 bp L14792 5'-TCCAACAACCCCTCAGGAATTC-3' H14987 5'-TTGATCGTARGATTGCGTATGC-3' 194 bp L14973 5'-CCTCACATCAAACCCGAATG-3' H15159 5'-TCCTCCAATTCATGTGAGTG-3' 185 bp L15152 5'TTCTGAATCCTAGTAGCCGACC-3' H15327 5'-TGCAGTCATCTCCGGTTTACAAGAC-3' 174 bp 12 S gene L599 5'-CTCAAAGGACTTGGCGGTGC-3' H673 5'-GAAGATGGCGGTATATAGAC-3' 73 bp L671 5'-TCACCAATCCTTGCTAATAC-3' H805 5'-AATGGCTTTCGTATTAAATT-3' 133 bp L733 5'-AACAAGAGTAAGCTCAATCA-3' H891 5'-CGGTGTGTGCGTGCTTCATG-3' 157 bp 28 S gene L28S 5'-GGTCGTCCGACCTGGGTATA-3' H28S 5'-TCTAATCATTCGCTTTACCGGAT-3' 96 bp Table 2 Myotragus sequence of the complete mtDNA cytochrome b (between positions 14,159 and 15,301), aligned with the Ovis sequence. Dots indicate sequence identity. OVIS MYOTRAGUS 14159 ATG ATC AAC ATC CGA AAA ACC CAC CCA CTA ATA AAA ATT GTA AAC AAC GCA TTC ATT GAT ... .C. ... T.. ... ... ... A.. ... ... ... ... ... ... ... ... ... ... ... ..C OVIS MYOTRAGUS 14219 CTC CCA GCT CCA TCA AAT ATT TCA TCA TGA TGA AAC TTT GGC TCT CTC CTA GGC ATT TGC ... ... ..C ... ... ..C ..C ... ... ... ... ... ..C ... ..C ... ... ... G.C ... OVIS MYOTRAGUS 14279 TTA ATT TTA CAG ATT CTA ACA GGC CTA TTC CTA GCA ATA CAC TAT ACA CCC GAC ACA ACA ... ..C ... ..A ..C ... ... ... ... ... ... ... ... ... ... ... T.. ... ... ... OVIS MYOTRAGUS 14339 ACA GCA TTC TCC TCT GTA ACC CAC ATT TGC CGA GAC GTG AAC TAT GGC TGA ATT ATC CGA ... ... .G. ... ... ..C G.. ..T ... ... ... ... ..A ... ... ... ... ... ... ... OVIS MYOTRAGUS 14399 TAT ATA CAC GCA AAC GGG GCA TCA ATA TTT TTT ATC TGC CTA TTT ATG CAT GTA GGA CGA ... ... ..T ... ... ..A ... ..C ... ... ..C G.. ... ... ... ... ..C ..G ... ..G OVIS MYOTRAGUS 144590 GGC CTA TAT TAT GGA TCA TAT ACC TTC CTA GAA ACA TGA AAC ATC GGA GTA ATC CTC CTA ... ... ..C ..C ... ... ..C ..T ... ... ... ... ... ... ... ... A.. ... ... ... OVIS MYOTRAGUS 14519 TTT GCG ACA ATA GCC ACA GCA TTC ATA GGC TAT GTC TTA CCA TGA GGA CAA ATA TCA TTC ..C A.A ... ... ..T ... ... ... ... ..T ..C ..T ... ... ... ... ... ... ..T ..T OVIS MYOTRAGUS 14579 TGA GGA GCA ACA GTT ATT ACC AAC CTC CTT TCA GCA ATT CCA TAT ATT GGC ACA AAC CTA ... ... ... ..C ... ..C ... ... ... ..C ... ..T ..C ... ..C ... ..A ..C ..T ..T OVIS MYOTRAGUS 14639 GTC GAA TGA ATC TGG GGA GGA TTC TCA GTA GAC AAA GCT ACC CTC ACC CGA TTT TTC GCC ..A ... ... ... ..A ... ... ... ... ... ... ..G ..C ... ... ..A ... ..C ... ..T OVIS MYOTRAGUS 14699 TTT CAC TTT ATT TTC CCA TTC ATC ATC GCA GCC CTC GCC ATA GTT CAC CTA CTC TTC CTC ... ... ..C ..C C.. ... ..T ..T ... ... ... ... ... ... ..C ... ... ..A ... ... OVIS MYOTRAGUS 14759 CAC GAA ACA GGA TCC AAC AAC CCC ACA GGA ATT CCA TCG GAC ACA GAT AAA ATT CCC TTC ... ... ... ... ... ... ... ... T.. ... ... ... ..A ... G.. ..C ... ..C ..A ..T OVIS MYOTRAGUS 14819 CAC CCT TAT TAC ACC ATT AAA GAC ATC CTA GGT GCT ATC CTA CTA ATC CTC ATC CTC ATG ... ..C ... ... ... ... ... ... ... ... ..C ATG ..A ... ... ..T T.A ... ... ... OVIS MYOTRAGUS 14879 CTA CTA GTA CTA TTC ACG CCT GAC TTA CTC GGA GAC CCA GAC AAC TAC ACC CCA GCA AAC ... ... ... ... ... ..A ..A ... C.. ... ... ... ... ..T ..T ..T ..A ... ..C ... OVIS MYOTRAGUS 14939 CCA CTT AAC ACT CCC CCT CAC ATC AAA CCT GAA TGA TAC TTC CTA TTT GCG TAC GCA ATC ... ..C ... ..A ... ... ... ... ... ..C ... ... ... ... ... ..C ..A ... ... ..T OVIS MYOTRAGUS 14999 TTA CGA TCA ATC CCT AAT AAA CTA GGA GGA GTC CTC GCC CTA ATC CTC TCA ATC CTA GTC C.. ... ... ..T ..C ..A ... ... ... ... ..T ..A ... ... G.. ... ... ..T ..G A.. OVIS MYOTRAGUS 15059 CTA GTA ATT ATA CCC CTC CTC CAT ACA TCA AAG CAA CGG AGC ATA ATA TTC CGA CCA ATC ... ..G C.. ... ..T ..A ... ..C .A. ..C ..A ... T.A ... ... ... ... .AG ... ..T OVIS MYOTRAGUS 15119 AGT CAA TGC ATA TTC TGA ATC CTA GTA GCC GAC CTA TTA ACA CTC ACA TGA ATT GGA GGA ... ... ..T C.. G.. ... ..T T.. ... ..A ... ... C.. ... ... ... ... ... ... ..C OVIS MYOTRAGUS 15179 CAG CCA GTT GAA CAC CCC TAC ATC ATT ATT GGA CAA CTA GCA TCT ATT ATA TAT TTC CTT ... ... ... ... ... ... ..T ..T ... ... ... ..G ... ... ... ... ... ... ... ... OVIS MYOTRAGUS 15239 ATC ATT CTA GTC ATA ATA CCA GTA GCT AGC ATC ATC GAA AAC AAC CTC CTA AAA TGA AGA ... ... ... ... ... ... ... ..G ..G ... .C. ... ... ... ... ... ... ... ... ... OVIS MYOTRAGUS 15299 CAA ... The putative presence of nuclear mtDNA insertions is very unlikely, since we proceed designing the L primers for the next fragment in the sequence already retrieved from the previous fragment, and the H primers in a consensus Caprinae sequence. The 12S gene sequence (305 base pairs) was retrieved in three overlapping fragments (see primers in Table 1). The sequences found in 12S showed several differences from those of extant Caprinae, a fact that points to its authenticity. The PCR of the 28S nuclear gene yielded a very faint band around 140 base pairs and was subsequently cloned and sequenced. In addition, DNA from tissue samples from domestic goat (Capra aegagrus), southern chamois (Rupicapra pyrenaica) and domestic sheep (Ovis aries) was extracted in a separated laboratory, and the same 28S fragment was determined. The 96 base pairs sequence obtained from Myotragus is similar to that found in other Caprinae (Table 3), and clearly different to the human one and to the cow, a suggested source of contamination because of the use of BSA (bovine serum albumine) in ancient DNA amplifications. Moreover, no other Bovids, extinct or living, had been analyzed in the same laboratory when the extraction, amplification, cloning and sequencing of the Myotragus genes was undertaken; therefore, the 28S sequence seems to be endogenous of Myotragus. The cloning of the 28S PCR product (Table 3) showed that two of the sequences (about 20% of the clones) were human contaminants; this accounts for the noise observed in the direct sequencing of the PCR product. Most likely, the human DNA comes from handling of the Myotragus bones during its excavation and posterior morphological study. Nevertheless, the error rate in the clones is relatively low (3.07 per thousand base pairs), a figure within the range found in some modern specimens and well preserved ancient remains, like the moas [20]. Table 3 Myotragus clones and consensus sequence of the nuclear gene 28S rDNA, aligned with sequences of Capra, Ovis and Rupicapra obtained in this study (including Bos as a reference sequence). BOS GGGGCGAAAGACTAATCGAACCATCTAGTAGCTGGTTCCCTCCGAAGTTTCCCTCAGGATAGCTGGCGCTCTCGCA-AACGC-ACAG--ACCC-ACGCAGTTTT M clone1 ............................................................................---...----.-A....-.......... M clone2 ............................................................................---...----.-A....-.......... M Clone3 ..........................................................................T.---...----.-A....-.......... M Clone4 ............................................................................---...----.-A....-.......... M Clone5 ............................................................................---...----.-A....-.......... M Clone6 .......................................T..................................T.---...----.-A....-.......... M Clone7 ............................................................................---...----.-A....-.......... M Clone8 ................................................C...........................---...----.-A....-.......... M CONSENSE ............................................................................---...----.-A....-.......... CAPRA ............................................................................-...C--------....-.......... OVIS ............................................................................-------------....-.......... RUPICAPRA ............................................................................G.-..-G...AG-....-.......... Variation in length is characteristic of 28S gene; size variations are due to expansions or contractions of variable segments in 10–12 positions within the gene, where variation does not interfere with ribosomal function [21]. The sequences were aligned by eye; due to the high interspecific variation in 28S, no definite alignment is possible and therefore, the phylogenetic information obtained from this gene is limited. However, the retrieval of a nuclear gene indicates the quality of the DNA of the sample used. Furthermore, it opens new possibilities of research for assessing the phylogenetic relationships of Myotragus and for retrieving nuclear genes with phenotypic implications. Phylogeny of the Caprinae subfamily and the position of Myotragus in the tree We included Pantholops hodgsoni in the phylogenetic analysis since previous analysis as well as ours indicate that this species is clearly associated to other members of the subfamily Caprinae. As an outgroup, we used 14 members of the tribes Alcephalini and Hippotragini (see Methods) previously detected as the closest members within the Bovidae [17,22]. The cytochrome b and cytochrome b + 12S Bayesian and maximum-likelihood trees (Figure 1) showed a topology roughly similar to that found in previous studies on Caprinae phylogeny (e.g., [22]). Recently, it has been reported that the Budorcas cytochrome b sequence available at GenBank was in fact a chimaera that included a fragment of Ovis sequence, and a new sequence for this species was provided [17]. Crucially, in previous works Myotragus formed a quite stable clade with Budorcas and Ovis that is not maintained anymore, with Budorcas now clustering in a different position. Figure 1 Bayesian tree (a) and maximum-likelihood tree (b) of the entire cytochrome b of all Caprinae species in databases. Bayesian tree (c) and maximum-likelihood tree (d) of the concatenated cytochrome b plus a 12S fragment of all Caprinae species in databases with both sequences. Bootstrap per cent values higher than 40% (maximum-likelihood tree) and support values bigger than 0.4 (Bayesian tree) are depicted in the nodes. The scale bar represents 0.1 substitutions/site. Examination of the cytochrome b trees (Figure 1a and 1b) shows that some of the clades previously found are present in these trees, specially Myotragus+Ovis (67.5% in the maximum-likelihood trees), Capricornis+Ovibos+Naemorhedus (100% in the maximum-likelihood tree) and Hemitragus+Pseudois+Capra (94.6% in the maximum-likelihood tree). The position of Hemitragus, that precludes the monophyly of the genus Capra, is problematic, as some authors have already noticed [18]. Some species show unstable positions in the trees, particularly the genera Budorcas, Ammotragus, Pantholops, Rupicapra and Oreamnos. The problematic position of these taxa has been already reported on extant species studies (see, for instance [22,23]). However, our tree, although still with low bootstrap values, has been able to better resolve the phylogenetic position of Myotragus, which is basal to the Ovis clade; the support value for the Myotragus+Ovis grouping is now 0.46 in the Bayesian tree and the bootstrap value 67.5% in the maximum-likelihood tree, respectively [11]. The trees with the cytochrome b + 12S fragment (305 base pairs) did not significantly improve the phylogeny (Figure 1c and 1d); the Capricornis+Ovibos and Hemitragus+Pseudois+Capra clades are well supported by bootstrap analysis, while the support of the cluster of Myotragus+Ovis is reduced with respect to the cytochrome b sequence alone. This discrepancy can be attributed to the small number of informative positions added by the 12S fragment together with the reduced number of species that could be used due to the unavailability of the 12S sequence in many species. Evolutionary rates and diversification of Caprinae The complete resolution of the Caprinae phylogeny cannot be achieved with the cytochrome b gene alone, as some species of the subfamily (specially of the genera Budorcas, Oreamnos, Rupicapra, Ammotragus and Pantholops) still have an unstable position. It is likely that the lack of resolution in the basal branches is due to the existence of a very quick initial radiation in the Caprinae subfamily. To test this, we performed a likelihood ratio test for zero branch-length of all branches of the tree, that showed that there are nine branches with p > 0.05 (and therefore, not significantly different from 0). Five of them belong to the basal branches of the Caprinae tree, thus supporting the hypothesis of a very fast initial radiation of this group. Additionally, the plot of grow of lineages of an ultrametric tree (see below) also supports this fast initial radiation. Being Myotragus a small caprine, its long branch in the maximum-likelihood tree previously detected with a cytochrome b fragment [11] was attributed to an earlier age of first reproduction and a shorter generation time in Myotragus than in other wild caprines [11]. This long branch is also present in the trees of the complete cytochrome b (Figure 1a and 1b). However, recent estimates indicate that Myotragus was actually bigger than previously believed, with a weight ranging between 15 and 25 kg for the smallest adult specimens and between 40 and 60 kg for the larger specimens [24]. Therefore the reason for this long branch cannot be due to its presumed extreme reduced size (and the consequences that this involves). On the other hand, the hypothesis that tries to explain differences in evolutionary rates as due to differences in metabolic rate or generation time (both correlated to body mass) was not supported in the analysis of cytochrome b evolutionary rates of a wide set of mammalian species [25]. It is thus more likely that this relatively long branch of Myotragus in the Caprinae tree is mainly a stochastic phenomenon. Dating of the Caprinae phylogeny Due to the endemicity of Myotragus [2], it can be hypothesized that this lineage diverged from the continental species 5.35 million years ago (Mya), when the desiccation of the Mediterranean ended [13,14] and the Myotragus ancestors became isolated in the Balearic Islands. In principle, this would be a minimum age for the node in the tree, but we consider likely that the isolation of the Balearic Islands was a vicariant event responsible for the split of the Myotragus and continental lineages, so that we can use this date as a fixed age in the Caprinae cytochrome b tree (Figure 2). The dating results for the maximum-likelihood and Bayesian trees are very similar and only the former are shown. We have calculated ultrametric trees using the Langley-Fitch method [26] as well as the penalyzed likelihood method with several smoothing factors [27] implemented in the r8s program. The cross-validation procedure to explore the fidelity with which these different methods explain the branch length variation in the tree indicates that the Langley-Fitch method is slightly better for our data set and therefore we used this method in the calculations described next (although dates obtained were very similar using penalyzed likelihood with a log10 of the smoothing factor of 2; not shown). In addition, we have performed a parametric bootstrap procedure, in which we simulated alignments from the maximum-likelihood tree and re-calculated dates from the same maximum-likelihood topology with branch lengths optimized for each simulated alignment. This allowed us to evaluate the stochastic errors of date estimates associated to sampling a finite number of alignment positions [28]. In a different parametric bootstrap procedure, we reconstructed new trees from each simulated alignment and used them for time estimation (instead of only the original topology) and therefore errors associated with finite number of positions as well as with tree reconstruction imperfections are estimated [28]. The standard deviations and 95% confidence intervals for most date estimates were very similar for both parametric bootstrap procedures (Table 4), indicating that most of the errors in the present data set are associated to the finite sample sequence rather than to tree errors. In the following, standard deviations given refer to the parametric bootstrap using only the original topology for calculating dates. Table 4 Dates in Mya, standard deviations, and 95% intervals (in parenthesis) obtained with parametric bootstrap using the original topology (Single topology) or new tree calculations (Several topologies) for each simulation of the mtDNA cytochrome b of Caprinae. The calibration is based on the isolation of Myotragus in the Balearic Islands at 5.35 million years ago. Lineage Single topology Several topologies Capra 1.5 ± 0.2 (1.2 – 1.9) 1.5 ± 0.2 (1.1 – 1.9) Capra ibex/C. pyrenaica 0.6 ± 0.1 (0.4 – 0.9) 0.6 ± 0.1 (0.3 – 0.8) Rupicapra pyrenaica/R. rupicapra 1.6 ± 0.3 (1.1 – 2.1) 1.5 ± 0.3 (1.1 – 2.1) Ovis 2.2 ± 0.3 (1.7 – 2.8) 2.1 ± 0.3 (1.6 – 2.8) Root 6.2 ± 0.4 (5.5 – 7) 6.1 ± 0.5 (5.4 – 7.2) Figure 2 Ultrametric tree of Caprinae derived from the cytochrome b maximum-likelihood tree. The geological date of the isolation of the Myotragus lineage in the Balearic islands 5.35 Mya, that was used as a calibration point, is highlighted. Calculated dates and standard deviations are given for some important splits in the tree. A lineages through time plot is shown under the tree. The number of lineages in each node is show in logarithmic scale. The horizontal scale bar represents 1 Mya for both the tree and the lineages through time plot. The ultrametric trees calculated by the Langley-Fitch method applied to the cytochrome b maximum-likelihood tree showed that the separation of the most divergent lineages within Caprinae (Pantholops and the rest of species) occurred 6.2 ± 0.4 Mya. These dates are more recent than a proposed Caprinae radiation 11 Mya based on different molecular calibrations (see Introduction) [17,18]. There is no general consensus from fossil data on the date of the start of the Caprinae radiation, that ranges from around 14 Mya [12] to a more recent Late Miocene origin [29], which would be more consistent with our results. After the first split, the separation of the most basal lineages took place around or before 5.35 Mya (Figure 2), and therefore we can estimate that the main radiation of caprines occurred in less than one million years of evolution. A lineages through time plot, that reflects the number of lineages in each split point of the tree (bottom of Figure 2), indicates a fast growing of lineages during the first million years of evolution in comparison with the flatter slope in the rest of the tree. Although a few species are missing in this tree, all genera are covered and it is likely that the tree with all species will reveal the same clear steep slope at its base. This truly supports the existence of a quite fast initial radiation that led to the differentiation of nine clades (including the Myotragus lineage) that are at the base of all extant genera. In addition, this explains the difficulties we face in reconstructing the phylogenetic tree of this group. Interestingly, the separation of the two chamois species (Rupicapra) and the beginning of the radiation of the main clade of goats (separation of the Capra ibex and Capra pyrenaica clade from the rest of species) occurred at 1.6 ± 0.3 Mya and 1.5 ± 0.2, respectively, according to the calibrated cytochrome b tree. This would imply that the beginning of the Pleistocene glaciations promoted the diversification of the main lineages of Caprinae species living in the Alps and other Eurasian mountainous regions. The deep split observed between the alpine and southern chamois, at 1.6 ± 0.3 Mya, is quite remarkable, since these two species have been considered subspecies until recently [30], and other molecular studies have given much more recent dates for the separation of these two species using restriction fragment length polymorphism of mitochondrial DNA [31] or microsatellites [32]. Their split is even older than that between Capra pyrenaica and Capra ibex, a pair of species with a similar geographic disjunction than the two chamois and traditionally recognized as different species, that separated 0.6 ± 0.1 Mya according to our data. To test whether the deep split observed between the two chamois was caused by particularly divergent cytochrome b haplotypes of chamois used in this study, we performed the calibrations with all sequences of chamois in databases, including shorter ones (five for R. pyrenaica and six for R. rupicapra), obtaining very similar dates (not shown). To analyze if the date of this split depended on the particularly long branch of Myotragus, we manually reduced this branch to the half, but we obtained again similar dates for all splits in the trees (1.7 Mya for the split between chamois). This is due to the smoothing of rates performed by the calibration methods used, that correct for the higher evolutionary rate of the Myotragus lineage. To account for the possibility that Myotragus could be more basal in the tree, we manually positioned Myotragus just after the separation of Pantholops. Then the split date of the two chamois was reduced, but only up to 1.4 Mya. This is in agreement with the parametric bootstrap analyses (Table 4), that indicate that the tree topology is not crucial in this calculation. This is surely due to the small internal branches involved in all different topologies arranging the basal species of the tree. Other works where several topologies were used for calculating dates have also shown that age estimates were largely insensitive to different phylogenies [28]. Finally, if Myotragus and Ovis had diverged in fact more recently and the ancestors of Myotragus had reached the Balearic Islands via a transmarine colonization after these islands became disconnected from the continent, as proposed for other vertebrates [32], all the dates in the tree could be more recent; however, this hypothesis is extremely unlikely in Myotragus from the paleontological evidence [24]. On the other hand, the use of evolutionary rates calculated from other sources, for example, the separation of humans and chimpanzees 5 – 7 Mya [34,35] combined with the genetic distance between their cytochrome b sequence, would lead to a separation of Myotragus and Ovis clades at 7.2 Mya. Although the use of an external calibration could be problematic, this reference would not support a transmarine colonization of Myotragus ancestors. Thus, the calibrated cytochrome b trees indicate that alpine and southern chamois have been separated much longer than previously thought, which would help explain the anti-hybridization mechanisms currently operating [30]. Conclusion In conclusion, this study is among the first ones in which mitochondrial as well as nuclear DNA has been retrieved from an extinct species from a warm location, in this case the Mediterranean area. Also, our research demonstrates the importance of ancient DNA in phylogenetics, since we have been able to use the isolation of Myotragus in the Baleric Islands as a vicariant event for dating the Caprinae phylogenetic tree. This is quite a unique opportunity for mammals because, contrary to other zoological groups such as insects, reptiles or amphibians, almost no extant endemic mammal remains in these Islands. Methods A left tibiae (MNIB 60176) bone from Myotragus balearicus from Cova des Gorgs (Escorca, Mallorca), dated to 6,010-5,830 cal BC 2σ (Beta-177239) was chosen for DNA analysis because of its excellent external preservation. The remainder bone not used for analysis is held in the vertebrate collection in the Museu de la Naturalesa de les Illes Balears (MNIB, Palma de Mallorca); the present sample is approximately between 500 and 1,850 years younger than the other one from the same site previously analyzed [11]. DNA was extracted following procedures described elsewhere [11]. The sample was demineralized overnight with 0.5 M EDTA, incubated overnight with proteinase K and SDS, extracted with phenol-chloroform and concentrated and desalinized with centricon columns. Finally, the extract was purified with silica. The sample was extracted with blank extractions to monitor against contamination. Several fragments of mtDNA cytochrome b gene were amplified and sequenced until completing the whole gene (Table 1). The cytochrome b sequences matched the partial sequence previously retrieved from another Myotragus bone from the same site [10]; multiple controls, such as independent replication of the results in two different laboratories and cloning of different overlapping PCR products, contributed to the authentication of the sequences. All the work was carried out in a dedicated ancient DNA laboratory, with UV lights, air-positive pressure and regular cleaning of surfaces with bleach. No signs of contamination were observed along the study in the extraction and PCR blanks; the sample was analyzed along with an Upper Paleolithic human remain [36] that yielded a human sequence. In addition, no other Bovid had ever been extracted in the same laboratory. One fragment of the cytochrome b was previously replicated in the Ancient Biomolecules Centre (Oxford), yielding the same sequence as in the present study. For the 12S gene, some primers were designed (Table 1) to match previous Caprinae sequences published [37]. Due to the exceptional quality of this extract, a pilot project was launched for trying to retrieve a fragment of a nuclear gene. The 28S rDNA was chosen because it is present in some hundreds of copies in the genome. The primers used to amplify the 28S rDNA were those designed [38] to be vertebrate-specific (Table 1). Five microliters of extract were added to 20 μl PCR reactions, containing 1× reaction buffer, 1 unit of taq DNA polymerase, 2.5 mM of MgCl2, 25 pmol of each primer. Forty cycles of 1 min at 94°C, 1 min at 50°C and 1 min at 72°C were performed. PCR products were resolved in 1% low-melting agarose gels in a TA buffer; bands were excised from the gel and subjected to a second 30 cycles of PCR with limiting reagents. PCR products were cloned with the SureClone Ligation Kit (Pharmacia, Upsala, Sweden); inserts were sequenced with 3100 Gene Analyzer (Applied Biosystems). Sequences have been deposited at the GenBank under accession numbers: AY380560, AY380561, AY380562, AY380563, AY894418, AY894419 and AY894420. Several phylogenetic analyses were performed with the Myotragus sequences. We used its cytochrome b together with the whole cytochrome b of all Caprinae species available in databases (Capra hircus, Capra aegagrus, Capra falconeri, Capra caucasica, Capra cylindricornis, Capra nubiana, Capra ibex, Capra sibirica, Hemitragus jemlahicus, Pseudois nayaur, Pseudois schaeferi, Ovis ammon, Ovis aries, Ovis dalli, Budorcas taxicolor, Myotragus balearicus, Rupicapra pyrenaica, Rupicapra rupicapra, Capricornis crispus, Ovibos moschatus, Naemorhedus caudatus, Pantholops hodgsoni, Oreamnos americanus), plus the 987 base pairs sequence of Capra pyrenaica and the complete sequence of 14 members of Alcephalini and Hippotragini used as outgroups (Addax nasomaculatus, Alcelaphus buselaphus, Alcelaphus lichtensteini, Beatragus hunteri, Connochaetes gnou, Connochaetes taurinus, Damaliscus lunatus, Damaliscus pygargus, Hippotragus equinus, Hippotragus niger, Oryx dammah, Oryx gazella, Oryx leucoryx and Sigmoceros lichtensteinii); this high number of outgroups eliminates the possible biases of choosing a single one at random and partitions the longest branches, thus reducing the problem of long branch attraction [39]. The analyses considered a general time reversible (GTR) model with a proportion of invariable sites and a discrete gamma model with 6 categories to account for among site rate variation. This model was selected upon checking with Modeltest [40] that the set of all positions as well as the 1st, 2nd and 3rd codon positions favored the GTR model or a model close to it (GTR being the closest model in the phylogeny programs used here) and gamma plus invariable rates. Phylogenies were performed using Bayesian analysis implemented in the program MrBayes v. 3.0. [41]. For the calculations, independent models of sequence evolution were used for each codon position. Four chains of 4,000,000 trees were generated, sampling every 100th tree, with burning completed by the 20,000th tree; thus, 20,000 trees were used to estimate topologies and posterior probabilities of parameters. It was checked that the likelihood and all parameters were stationary after this burn-in. Maximum-likelihood analyses were performed with the program Phyml, version 2.4.4 [42]. The gamma parameter, the proportion of invariable sites and the nucleotide frequencies were estimated from the data. Two maximum-likelihood tree searches were made, one starting from a neighbor joining tree and another from the previously generated Bayesian tree, selecting the tree with the best likelihood (almost always the one starting from the Bayesian tree). To test the robustness of the tree clades, 1,000 bootstrap samples were performed. Maximum-likelihood and Bayesian trees were similarly generated in a combination of the cytochrome b sequence with a 12S fragment, using the only 20 caprine species in which both 12S and cytochrome b were available at GenBank, and an outgroup. For the Bayesian tree, an independent GTR model was considered for the 12S partition. No tree was generated for the 28S fragment, due to irresolvable difficulties with the alignment. To test the possibility of an explosive diversification of the Caprinae subfamily, we performed the likelihood ratio test for zero-length branches [43] from the cytochrome b data using PAUP 4,0 [44] and the same maximum-likelihood model as above. To estimate absolute rates of molecular evolution with a relaxed molecular clock on the cytochrome b phylogenetic tree (and divergence times using the geological date of the isolation of the Myotragus lineage), the r8s v1.7 program was used [45]. Outgroup sequences were eliminated from the tree previous to the analysis of caprines. Smoothing of rate variation along the tree was performed with the Langley-Fitch [26] and penalized likelihood [27] methods. Sixteen smoothing factors with log10 from -2 to 5.5 were used for the penalized likelihood method. The lowest χ2 cross-validation score, as calculated by r8s, was used to select the best method. To perform parametric bootstraps, 1,000 Monte Carlo simulations of alignments of the same length than the complete cytochrome b were generated with SeqGen [46] using the phylogenetic trees and model parameters previously obtained. A lineages through time plot of an ultrametric tree obtained with these methods was calculated with the program GENIE [47]. Authors' contributions CL-F and LS did the experimental work, JC carried out the phylogenetic and dating analyses, TM-B carried out some of the analyses, JAA provided the sample and critical feedback to CL-F, JB provided laboratory facilities. CL-F, JC and JB contributed to the writing of the manuscript. Acknowledgements This research was supported by the Dirección General de Investigación, Ministerio de Ciencia y Tecnología of Spain (grants BFU2004-02002 and BOS2003-08070), by Departament d'Universitats, Recerca i Societat de la Informació, Generalitat de Catalunya (grant 2001SGR00285) and a fellowship to LS. We are grateful to Alan Cooper and Beth Saphiro (Ancient Biomolecules Center, Oxford), for the partial replication of some ancient sequences, to Alexandre Hassanin (Muséum National d'Histoire Naturelle, Paris) for advice in the designing of the 12S Myotragus primers and to Salvador Carranza (Universitat de Barcelona) for critical comments on the manuscript. ==== Refs Bate DMA A new artiodactyle from Majorca Geol Mag NS 1909 6 385 388 Alcover JA Moyà-Solà S Pons-Moyà J Les quimeres del passat Els vertebrats fòssils del Plio-Quaternari de les Balears i les Pitiüses 1981 Palma de Mallorca: Ed. Moll Alcover JA Perez-Obiol R Yll E-I Bover P The diet of Myotragus balearicus Bate 1909 (Artiodactyla: Caprinae), an extinct bovid from the Balearic Islands: evidence from coprolites Biol J Linn Soc 1999 66 57 74 10.1006/bijl.1998.0260 Alcover JA Seguí B Bover P MacPhee RDE Extinctions and Local Disappearances of Vertebrates in the Western Mediterranean Islands Extinctions in Near Time 1999 New York: Kluwer Academic/Plenum Publishers Bover P Alcover JA Estimating physical characteristics of neonate Myotragus balearicus (Artiodactyla, Caprinae) Deinsea – Annual of the Natural History Museum Rotterdam 1999 7 33 54 Bover P Alcover JA The evolution and ontogeny of the dentition of Myotragus balearicus Bate 1909 (Artiodactyla, Caprinae): evidence from new fossil data Biol J Linnean Soc 1999 68 401 428 10.1006/bijl.1998.0294 Ramis D Alcover JA Revisiting de earliest human presence in Mallorca, Western Mediterranean Proc Prehistoric Soc 2001 67 261 269 Bover P Alcover JA Understanding Late Quaternary Extinctions: the case of Myotragus balearicus Bate 1909 J Biogeograph 2003 30 771 782 10.1046/j.1365-2699.2003.00872.x Wilson DE Reeder DM Mammal Species of the World: A Taxonomic and Geographic Reference 1993 2 Washington: Smithsonian Institution Press Lalueza-Fox C Bertranpetit J Alcover JA Shailer N Hagelberg E Mitochondrial DNA from Myotragus balearicus, an extinct Bovid from the Balearic Islands J Exp Zool (Mol Dev Evol) 2000 288 56 62 10.1002/(SICI)1097-010X(20000415)288:1<56::AID-JEZ6>3.0.CO;2-X Lalueza-Fox C Shapiro B Bover P Alcover JA Bertranpetit J Molecular phylogeny and evolution of the extinct bovid Myotragus balearicus Mol Phylogenet Evol 2002 25 501 510 12450754 10.1016/S1055-7903(02)00290-7 Vrba ES Africa Bovidae: evolutionary events since the Miocene S Afr J Sci 1985 81 263 266 Krijgsman W Hilgen FJ Raffi I Sierro FJ Wilson DS Chronology, causes and progression of the Messinian salinity crisis Nature 1999 400 652 654 10.1038/23231 Duggen S Hoernle K van den Bogaard P Rupke L Morgan JP Deep roots of the Messinian salinity crisis Nature 2003 422 602 606 12686997 10.1038/nature01553 Vogel P Cosson JF Lopez Jurado LF Taxonomic status and origin of the shrews (Soricidae) from the Canary islands inferred from a mtDNA comparison with the European Crocidura species Mol Phylogenet Evol 2003 27 271 82 12695091 10.1016/S1055-7903(02)00403-7 Mitchell-Jones A J Amori G Bogdanowicz W Krystufek B Reijnders PJH Spitzenberg F Stubbe M Thissen JBM Vohralik V Zima J The atlas of European mammals 1999 London: T & AD Poyser Ropiquet A Hassanin A Molecular phylogeny of caprines (Bovidae, Antilopinae): the question of their origin and diversification during the Miocene J Zoolog Syst Evol Res 2005 43 49 60 10.1111/j.1439-0469.2004.00290.x Ropiquet A Hassanin A Molecular evidence for the polyphyly of the genus Hemitragus (Mammalia Bovidae) Mol Phylogenet Evol 2005 36 154 168 15904863 10.1016/j.ympev.2005.01.002 Poinar H Kuch M McDonald G Martin P Pääbo S Nuclear gene sequences from a late pleistocene sloth coprolite Curr Biol 2003 13 1150 1152 12842016 10.1016/S0960-9822(03)00450-0 Cooper A Lalueza-Fox C Anderson S Rambaut A Austin J Ward R Complete mitochondrial genome sequences of two extinct moas clarify ratite evolution Nature 2001 409 704 707 11217857 10.1038/35055536 Laudien Gonzalez I Gorski JL Campen TJ Dorney DJ Erickson JM Sylvester JE Schmickel RD Variation among human 28S ribosomal RNA genes Proc Natl Acad Sci USA 1985 82 7666 7670 3865188 Hassanin A Douzery EJP The tribal radiation of the family Bovidae (Artiodactyla) and the evolution of the mitochondrial Cytochrome b gene Mol Phylogenet Evol 1999 13 227 243 10603253 10.1006/mpev.1999.0619 Groves P Shields GF Phylogenetics of the Caprinae based on cytochrome b sequence Mol Phylogenet Evol 1996 5 467 476 8744761 10.1006/mpev.1996.0043 Bover P Noves aportacions al coneixement del gènere Myotragus Bate 1909 (Artiodactyla, Caprinae) de les Illes Balears 2004 PhD thesis, University of the Balearic Islands Castresana J Cytochrome b phylogeny and the taxonomy of Great Apes and Mammals Mol Biol Evol 2001 18 465 471 11264397 Langley CH Fitch WM An examination of the constancy of the rate of molecular evolution J Mol Evol 1974 3 161 177 4368400 10.1007/BF01797451 Sanderson MJ Estimating absolute rates of molecular evolution and divergence times: a penalized likelihood approach Mol Biol Evol 2002 19 101 109 11752195 Sanderson MJ Doyle JA Sources of error and confidence intervals in estimating the age of angiosperms from rbcL and 18S rDNA data Amer J Bot 2001 88 1499 1516 Gentry AW The Miocene differentiation of Old World Pecora (Mammalia) Historical Biol 1994 7 115 158 Masini F Lovari S Systematics, phylogenetic relationships, and dispersal of the chamois (Rupicapra spp.) Quaternary Research 1988 30 339 349 Hammer S Nadlinger K Hartl GB Mitochondrial DNA differentiation in chamois (genus Rupicapra implications for taxonomy, conservation, and management Acta Theriologica Suppl 1995 3 145 155 Perez T Albornoz J Dominguez A Phylogeography of chamois (Rupicapra spp.) inferred from microsatellites Mol Phylogenet Evol 2002 25 524 534 12450756 10.1016/S1055-7903(02)00296-8 Martinez-Solano I Gonzalbes HA Arntzen JW Garcia-Paris M Phylogenetic relationships and biogeography of midwife toads (Discoglossidae: Alytes J Biogeogr 2004 31 603 618 Yoder AD Yang Z Estimation of primate speciation dates using local molecular clocks Mol Biol Evol 2000 17 1081 1090 10889221 Glazko GV Nei M Estimation of divergence times for major lineages of primate species Mol Biol Evol 2003 20 424 434 12644563 10.1093/molbev/msg050 Caramelli D Lalueza-Fox C Vernesi C Lari M Casoli A Mallegni F Chiarelli B Dupanloup I Bertranpetit J Barbujani G Bertorelle G Evidence for a genetic discontinuity between Neandertals and 24,000-year-old anatomically modern Europeans Proc Natl Acad Sci USA 2003 100 6593 6597 12743370 10.1073/pnas.1130343100 Gatesy J Amato G Vrba E Schaller G DeSalle R A cladistic analysis of mitochondrial ribosomal DNA from the Bovidae Mol Phylogenet Evol 1997 7 303 319 9187090 10.1006/mpev.1997.0402 Greenwood A Capelli C Possnert G Pääbo S Nuclear DNA sequences from Late Pleistocene megafauna Mol Biol Evol 1999 16 1466 1473 10555277 Sanderson MJ Shaffer HB Troubleshooting molecular phylogenetic analyses Annu Rev Ecol Syst 2002 33 49 72 10.1146/annurev.ecolsys.33.010802.150509 Posada D Crandall KA MODELTEST: testing the model of DNA substitution Bioinformatics 1998 14 817 818 9918953 10.1093/bioinformatics/14.9.817 Huelsenbeck JP Ronquist F MRBAYES: Bayesian inference of phylogenetic trees Bioinformatics Application Notes 2001 17 754 755 Guindon S Gascuel O A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood Syst Biol 2003 52 696 704 14530136 10.1080/10635150390235520 Slowinski JB Molecular polytomies Mol Phylogenet Evol 2001 19 114 120 11286496 10.1006/mpev.2000.0897 Swofford DL PAUP: phylogenetic analysis using parsimony (and other methods) 1998 Version 4. Sunderland, Mass: Sinauer Sanderson MJ r8s: inferring absolute rates of molecular evolution and divergence times in the absence of a molecular clock Bioinformatics 2003 19 301 302 12538260 10.1093/bioinformatics/19.2.301 Rambaut A Grassly NC Seq-Gen: an application for the Monte Carlo simulation of DNA sequence evolution along phylogenetic trees Comput Appl Biosci 1997 13 235 238 9183526 Pybus OG Rambaut A GENIE: estimating demographic history from molecular phylogenies Bioinformatics 2002 18 1404 1405 12376389 10.1093/bioinformatics/18.10.1404	16356181	PMC1325261	CC BY	2021-01-04 16:28:04	no	BMC Surg. 2005 Dec 15; 5:23	latin-1	BMC Surg	2,005	10.1186/1471-2482-5-23	oa_comm
==== Front Virol JVirology Journal1743-422XBioMed Central London 1743-422X-2-881632114910.1186/1743-422X-2-88MethodologyA general method for nested RT-PCR amplification and sequencing the complete HCV genotype 1 open reading frame Yao Ermei [email protected] John E [email protected] Virahep-C Study Group [email protected] Department of Molecular Microbiology and Immunology, Saint Louis University School of Medicine, Saint Louis, Missouri 63104, USA2 Saint Louis University Liver Center, Saint Louis University School of Medicine, Saint Louis, Missouri 63104, USA2005 1 12 2005 2 88 88 17 6 2005 1 12 2005 Copyright © 2005 Yao et al; licensee BioMed Central Ltd.2005Yao 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 Hepatitis C virus (HCV) is a pathogenic hepatic flavivirus with a single stranded RNA genome. It has a high genetic variability and is classified into six major genotypes. Genotype 1a and 1b cause the majority of infections in the USA. Viral genomic sequence information is needed to correlate viral variation with pathology or response to therapy. However, reverse transcription-polymerase chain reaction (RT-PCR) of the HCV genome must overcome low template concentration and high target sequence diversity. Amplification conditions must hence have both high sensitivity and specificity yet recognize a heterogeneous target population to permit general amplification with minimal bias. This places divergent demands of the amplification conditions that can be very difficult to reconcile. Results RT and nested PCR conditions were optimized independently and systematically for amplifying the complete open reading frame (ORF) from HCV genotype 1a and 1b using several overlapping amplicons. For each amplicon, multiple pairs of nested PCR primers were optimized. Using these primers, the success rate (defined as the rate of production of sufficient DNA for sequencing with any one of the primer pairs for a given amplicon) for amplification of 72 genotype 1a and 1b patient plasma samples averaged over 95% for all amplicons. In addition, two sets of sequencing primers were optimized for each genotype 1a and 1b. Viral consensus sequences were determined by directly sequencing the amplicons. HCV ORFs from 72 patients have been sequenced using these primers. Sequencing errors were negligible because sequencing depth was over 4-fold and both strands were sequenced. Primer bias was controlled and monitored through careful primer design and control experiments. Conclusion Optimized RT-PCR and sequencing conditions are useful for rapid and reliable amplification and sequencing of HCV genotype 1a and 1b ORFs. ==== Body Background Hepatitis C virus (HCV) is a human hepatotropic flavivirus. It is the major cause of non-A, non-B hepatitis, infecting about 3% of people world-wide [1]. Nearly 4 million people in the United States are infected with HCV [2], predominantly with genotypes 1a and 1b. HCV infection becomes chronic in about 80% of infected individuals. These chronically infected patients are at high risk of developing serious liver disease, including cirrhosis and hepatocellular carcinoma [3]. No effective vaccine has been developed to prevent HCV infection. The best available therapy for HCV infection is a combination of pegylated interferon α and ribavirin, an oral guanosine analogue [4]. The response rate to therapy varies depending on HCV genotype, viral load, patient sex, patient age, and the stage of liver fibrosis [5]. The HCV genome is a positive polarity, single-stranded RNA about 9600 nucleotides long. It contains one long ORF flanked by 5' and 3' untranslated regions (UTR). The genome is highly variable due to the poor fidelity of the viral RNA dependent RNA polymerase (RdRp) and the lack of genome repair mechanisms. HCV genomic variability is not uniform throughout the genome. The 5'UTR and the terminal 98 nucleotides of the 3'UTR are conserved, but the region of the 3'UTR immediately downstream of the open reading frame and the adjacent U-rich sequence are highly variable [6]. Significant sequence variation is also present in the ORF at both the nucleotide and the amino acid level, especially in hypervariable regions (HVR1 and HVR2) within the E2 region [7,8]. Analysis of the NS5B region encoding the viral RNA polymerase from a wide range of HCV isolates led to the classification of HCV into six major genotypes and a series of subtypes [9,10]. Genotypes share less than 72% nucleotide homology. Within genotypes, subtypes have homologies of 75%–86%. HCV sequences within an infected individual exist as a group of related but distinct variants [11,12]. This distribution of sequences is common among RNA viruses and is referred to as "quasispecies". Quasispecies variation can lead to significant amino acid variation of the encoded proteins [11,13]. The distribution of sequences in a quasispecies clusters around a master sequence, and the "center" of the genetic distribution can be described either as the dominant quasispecies (the single most common sequence in the viral population) or as the consensus sequence (an "average" sequence comprised of the predominant sequence at each nucleotide position). This protocol is designed to yield the consensus sequence. The high genomic heterogeneity of HCV may contribute to viral immune evasion [9], promote chronicity [14], and may influence the outcome of interferon α therapy in HCV-infected individuals [11,15,16]. Therefore, systematic examination of HCV sequence variation has important implications in understanding HCV biology and could open novel avenues for anti-viral therapy. HCV viremia is relatively low compared to many other viruses, rarely exceeding 106–107 genomes per milliliter. Therefore, reverse transcription-polymerase chain reaction (RT-PCR) of the HCV genome must overcome not only high target sequence diversity, but also low template concentration. Hence, the amplification conditions must have high sensitivity and specificity yet recognize a heterogeneous target population. These divergent demands are difficult to reconcile. In this paper, we report a general method to amplify and sequence the whole ORF of HCV genotypes 1a and 1b. We systematically optimized all steps in the process, including isolation of HCV RNA from patient plasma or serum, RT, PCR primer sequences, PCR conditions, template preparation, sequencing and assembly. We have a success rate of over 95% in RT-PCR amplification and have successfully sequenced HCV ORFs from over 72 patients using this system. Results and discussion Amplification strategy The HCV ORF is over 9 kb long, so long range PCR was initially attempted to amplify partial or full HCV ORFs. Its success frequency was inadequate for large-scale HCV genome sequencing projects, so this approach was abandoned. Efficient amplification with regular PCR is limited to 3 kb. Therefore, to maximize PCR sensitivity, we divided the genome into four amplicons that were numbered sequentially as amplicon 1 to 4 starting from the 5' end of the genome, with each amplicon being less than 3 kb and overlapping with the adjacent amplicon(s). This strategy was effective for amplifying all amplicons except for amplicon 4 for both genotype 1a and 1b and amplicon 1 for genotype 1b. To increase the sensitivity of amplification for these regions, they were subdivided, which resulted in efficient amplification. The HCV ORF was therefore partitioned into amplicons 1, 2, 3, 4x and 4y for genotype 1a and amplicons 1x, 1y, 2, 3, 4x and 4y for genotype 1b. Figure 1 shows the amplicon partition for genotype 1b. Figure 1 HCV genotype 1b amplicons. Amplicons are numbered sequentially as amplicon 1 to 4 starting from 5' of the genome. Amplicon 1 and 4 are divided into halves named 1x, 1y and 4x, 4y. The amplicon boundaries indicate the 5' ends of the innermost amplification primers against the genome of strain J4. Optimization of RNA extraction RNA isolation must be suitable for extracting HCV RNA from both patient plasma and serum because these are common sources of HCV. RNA isolation must be efficient to yield adequate amounts of high purity template due to the limited amount of patient plasma or serum that is often available and the relatively low titer of the virus. We tried three RNA isolation protocols to isolate RNA from plasma/serum samples including guanidine thiocyanate denaturation plus phenol/chloroform extraction, the ZR Viral RNA Kit (ZYMO Research) and the QIAamp Viral RNA Mini Kit (Qiagen). The QIAamp Viral RNA Mini Kit (Qiagen) worked best. The manufacturer's protocol was followed without modification. Processing 140 μl plasma sample routinely yielded about 60 μl viral RNA solution, of which 15 μl was sufficient for an RT reaction. RNA isolation was equally efficient using this kit with either serum or plasma. Optimization of cDNA synthesis The reverse transcriptases tested include Cloned AMV Reverse Transcriptase (Invitrogen), AMV Reverse Transcriptase (Promega), Moloney Murine Leukemia Virus Reverse Transcriptase (M-MLV RT; Promega) and Enhanced Avian Reverse Transcriptase (AMV-RT; Sigma), an enhanced avian myeloblastosis virus reverse transcriptase. Reactions were assembled per manufacturer's instructions employing a constant amount of HCV RNA (15 μl for a 50 μl reaction). Because HCV RNA has a relatively high GC percentage and has many secondary structures that may interfere with RT, incubation temperatures between 30°C – 50°C were tested at 5°C intervals for each enzyme. After RT, nested PCR was performed to test the RT efficiency. Figure 2 shows part of the optimization of RT conditions for genotype 1b amplicon 2. Different sets of PCR primers were used for odd and even numbered lanes. RNA isolated by the Viral RNA Mini Kit (R2V2) was much more efficient than RNA processed through guanidine thiocyanate and phenol/chloroform extraction (R2V1) (compare lanes 1 and 2 versus 3 and 4). Random hexamers (Rndm) were more efficient than B4R1, a primer specific to the 3'-UTR (lanes 3 and 4 versus 5 and 6, or lanes 7 and 8 versus 9 and 10). For amplicon 2, AMV-RT and M-MLV RT worked equally well (lanes 3 and 4 versus 7 and 8, or lanes 5 and 6 versus 9 and 10). Lanes 11 and 12 are negative controls in which template RNA was omitted. Figure 2 Optimization of RT conditions for genotype 1b amplicon 2. R2V1 and R2V2 were RNAs isolated from the same aliquot of a patient plasma; R2V1 employed guanidine thiocyanate and phenol/chloroform extraction and R2V2 employed the Viral RNA Mini Kit. RT primer B4R1 is a specific primer targeted to the 3'UTR. Rndm, random hexamers; M-MLV, Murine Leukemia Virus Reverse Transcriptase; AMV, Enhanced Avian Reverse Transcriptase. Different PCR primers were used for odd or even numbered lanes. Lanes 11 and 12 are negative controls in which template RNA was omitted. M-MLV RT and AMV-RT both worked very well for amplicons 1, 2, 3 and 4x. For amplicon 4y, AMV-RT worked much better, especially if the enzyme was stored at -75°C or lower (data not shown). RT reactions were suitable for amplicons 1, 2, 3 and 4x when stored at -20°C for several months, but for amplicon 4y, fresh RT reactions worked much better. Optimization of nested PCR We optimized nested PCR conditions for each amplicon independently. The process is summarized in Figure 3. Figure 3 Amplification optimization process. First, we designed primers for nested PCR. Because viral genetic heterogeneity will prevent a given primer from working well on all isolates, we optimized three sense and three anti-sense primers for each amplicon as shown in Figure 4. The three anti-sense primers must reside 3' to all three sense primers for the downstream amplicon to prevent gaps between the amplicons. Primers were targeted to relatively conserved regions of the genome to maximize the number of isolates they recognize. We employed Oligo Explorer 1.2 [17] to guide primer design. The software considers melting temperature and length of primers while avoiding sequences prone to dimer or hairpin formation or self-complementary primers. To use the program, a reference sequence must be provided. We used consensus sequences generated by aligning all full length HCV 1a or 1b genome sequences available in Genbank because these consensus sequences represent "average" 1a or 1b isolates. We first chose the rough boundaries of the amplicons, and then designed primers within 500 nucleotides at both ends of each amplicon. Candidate primers of 20–25 nucleotides were designed and compared to the 1a or 1b alignment from which the reference sequence was generated. For positions with unavoidable variability within the primer, degenerated bases were used. Generally no more than 5 mixed bases per primer were employed because we found that primers with more mixed bases were less sensitive. However, a few primers have 6 degenerate bases because the heterogeneity in the target region was unavoidable. Universal bases deoxyinosine (dI) and deoxyuridine (dU) were used in initial optimizations, but the amplification sensitivity with these primers was insufficient, possibly due to dI's less discriminate base pairing and wide range of melting temperatures [18]. Figure 4 Relative position of amplicon amplification primers. Three pairs of amplification primers and their relative positions are shown. The red regions overlap with adjacent amplicon(s). Then we optimized all nine primer permutations (three sense versus three anti-sense primers) for each of the amplicons for primer concentration, Mg++ concentration, and annealing temperature against cloned HCV DNA. For genotype 1a, we optimized our amplification primers against strain H77 [GenBank: AF009606] [19]. For genotype 1b we used plasmid pHCV-CG1b [GenBank: AF333324] [20], which has the HCV 1b strain J structural region, the 1b strain BK non-structural region and the HCV 1a strain H 3' poly (UC) and X regions. Three Taq polymerases were tested against the cloned HCV cDNA using selected primer permutations. The enzymes were Taq DNA Polymerase in Storage Buffer B (Promega), Taq DNA Polymerase (Fisher) and Expand High Fidelity PCR System (Roche). Expand High Fidelity PCR System (Roche) was tested since it has a proofreading polymerase with high fidelity, but it was rejected due to insufficient sensitivity and excessive cost for a large-scale sequencing project. Taq DNA Polymerase from Fisher was chosen for all PCR reactions because it was the most efficient of the three. Because our goal was to directly sequence the RT-PCR products, its lower fidelity was not critical (see "Accuracy of the sequences"). Finally, we tested the optimized primers on several patient plasma samples. If the success rate for a given primer pair on clinical isolates was over 80%, we kept the primer pair. If not, we designed new primers and repeated the optimization process until at least three pairs of optimized primers were available for each amplicon. Table 1 (see additional file 1: HCVMethodPaperTable1.xls) and Table 2 (see additional file 2: HCVMethodPaperTable2.xls) list amplification and sequencing primers for genotypes 1a and 1b. Table 3 (see additional file 3: HCVMethodPaperTable3.xls) and Table 4 (see additional file 4: HCVMethodPaperTable4.xls) list optimized PCR conditions for each primer pair for genotypes 1a and 1b. Table 5 (see additional file 5: HCVMethodPaperTable5.xls) and Table 6 (see additional file 6: HCVMethodPaperTable6.xls) list genotype 1a and 1b primer permutations that worked well on patient samples. Amplification efficiency We amplified 72 genotype 1 patients (44 genotype 1a, 28 genotype 1b) ORFs using these primers and PCR conditions. The overall success rate for amplicons averaged over 95%. Table 7 lists amplification efficiency for each amplicon. The few amplicons that could not be generated by these optimized primers were easily amplified by designing custom primers derived from sequences obtained from the neighboring amplicon(s) for that isolate. Table 7 Amplification efficiency for patients' amplicons Genotype 1a Amplicon A1 A2 A3 A4x A4y Amplification efficiency 95a 98 93 100 95 Average efficiency 96.2 Genotype 1b Amplicon A1x A1y A2 A3 A4x A4y Amplification efficiency 100 100 93 93 100 100 Average efficiency 97.7 a Amplification efficiencies are shown as percentage. Sequencing RT-PCR often yields minor amounts of primer dimers or truncated products that can interfere with sequencing. Therefore, DNA templates were purified by gel extraction using QIAquick Gel Extraction Kit (Qiagen) following manufacturer's protocol. DNA concentration was determined by agarose gel electrophoresis comparing band intensity to the Hyperladder I (Bioline) marker. Two sets of DNA sequencing primers were designed and validated for each genotype 1a and 1b (table 1 and 2). Each set of primers contains both sense and anti-sense primers to obtain complete coverage of both strands. In the primary set of primers, the distance between adjacent primers is 150–300 bp. HCV sequences are very heterogeneous, so not all primers will work for all patients due to mismatches between the primers and templates. Because a typical sequencing read-length is over 600 bp, placing the primers this close together allows each to reach the position of the second primer downstream of it. This yields a sequencing depth of 4- to 5-fold when both strands are sequenced, which maximizes coverage and sequencing quality. The backup set of primers was used to fill in gaps in the rare cases when the primary set failed to completely cover an amplicon. Sequencing employed the ABI automated dye-terminator system. It was performed at a contract sequencing facility (Macrogen, Inc. Seoul, South Korea). For each sequencing reaction, 50 ng template and 3.2 pmol primer were used. Consensus sequences were obtained through assembling and editing the sequencing traces using Vector NTI (Informax). This program automatically assembles overlapping sequencing traces and identifies nucleotide positions with discrepancies between the traces. Computer base-calling errors were corrected following inspection of the sequence chromatograms. Mixed-base positions from the HCV quasispecies were resolved by manually identifying the predominant base at each position. Where necessary, additional sequencing reactions were performed to confirm the identity of a base or its predominance in the quasispecies spectrum. For accuracy, we require that each nucleotide be present in at least two unambiguous sequencing reactions, preferably of opposite polarity. Figure 5 shows an example with six overlapping sequencing traces. Two of the reactions revealed a mixture of G and A at position 1270 and the four other traces clearly indicated that G was dominant at this position; this base was manually identified as G. Figure 5 Resolving discordant sequencing traces. A section of six overlapping primary sequencing traces is shown. Traces 1–4 clearly indicate nt 1270 (shaded) is a G, whereas traces 5 and 6 are ambiguous at this position because both G and A were detected. The nucleotide was manually identified as a G due to the predominance of G's among the six traces. Accuracy of the sequences Errors in sequencing HCV genomes arise from three major sources: sequencing errors, enzymatic errors during RT-PCR and primer bias during PCR. Our sequencing depth averages over 4-fold and both strands are sequenced, so error from sequencing mistakes is negligible. Base changes are certainly introduced into the template DNAs during RT-PCR. However, determining consensus sequence by directly sequencing uncloned templates greatly reduces the impact of this type of error because for an enzymatically-derived error to be detected, the error would have to have become the predominant sequence in the template molecule population. This is rare with direct sequencing of PCR products, in contrast to using cloned templates such as are used for quasispecies analysis, where these errors are very significant. Quality control experiments with templates from a HCV donor-recipient set indicate that the rate of enzymatically-derived errors is less than 0.012% when a common set of RT-PCR primers are used [21]. The largest (and often least-appreciated) source of error in sequencing is due to primer bias. Primer bias is selective amplification of a portion of the sequences in the target population and is a result of varying primer affinities for the heterogeneous template molecules during PCR. Primer bias is unavoidable in HCV genetic analyses due to the extreme genetic heterogeneity of the virus. This bias cannot be eliminated, but it can be quantitated and minimized through careful primer design and conscientious control experiments. To measure our net sequencing reproducibility, we sequenced a HCV 1b ORF from two aliquots of plasma from a single blood draw. The experiment was done in a blinded manner and the primers used to amplify the two genomes were independently chosen. The identity of the two sequences was 99.1% at the nucleotide level and 99.4% at the amino acid level (compared to 91.2% nucleotide and 94.3% amino acid identity between these sequences and HCV J4 [GenBank: AF054247], another 1b isolate). Because these differences are primarily due to primer bias, they are not truly "errors". Rather, they represent alternate samplings of sequences within the viral quasispecies population. Record keeping Record keeping and storage of samples and reagents must be meticulous to avoid costly and time-consuming errors. To assist tracking of samples and data, we developed a custom relational MySQL database into which are entered the identity, source, and location of all PCR primers, sequencing primers, patient samples, RNAs, and PCR products. The database is web-enabled to permit remote access, it is secured behind a fire-wall, and access is limited to authorized users with valid passwords. The database and all sequence data are backed up to a secure tape-backup system in a different building three times a week. The database will be made available free of charge to interested parties. Conclusion Despite the high degree of genomic heterogeneity and relatively low viral titres, efficient amplification and sequencing of the HCV ORF is possible. We report optimized amplification and sequencing conditions for the complete HCV genotype 1a and 1b ORFs. This will facilitate large-scale HCV genome sequencing and greatly ease systematic genetic analyses of the virus. This method was developed to yield the viral consensus sequence through direct sequencing RT-PCR products. However, it should be easily adaptable to quasispecies analysis by replacing the Taq polymerase with a high fidelity thermostable DNA polymerase and sequencing cloned templates rather than uncloned PCR products. Materials and methods Primer naming convention Due to the large number of primers, we chose primer names to include information indicating genotype, amplicon number, polarity, purpose(amplification or sequencing), relative position on the amplicon, and version number. For primer "B2R.3-AP3", "B" stands for genotype 1b, the "2" means amplicon 2, "R" represents anti-sense (reverse) polarity. ".3" means it is from the third set of primers designed. The AP suffix stands for "amplification primer" and indicates the primer is suitable for PCR, and the final "3" means it is the innermost primer compared to the other PCR primers for the amplicon in the same set. Primer "A1L3.2" is a sequencing primer for genotype 1a, amplicon 1, of sense polarity, "3" indicates it is the third sequencing primer for the strand, and the final "2" indicates it is from sequencing primer set 2. cDNA synthesis cDNA was synthesized using random hexamers (Promega) and M-MLV RT or AMV-RT. For a 50 μl RT reaction, 15 μl viral RNA was mixed with 1 μg random primers in a sterile RNase-free 250 μl PCR tube, heated to 70°C for 5 minutes for M-MLV RT or 10 minutes for AMV-RT to melt secondary structures within the template and cooled immediately on ice. For the M-MLV RT, 10 μl M-MLV 5 × Reaction Buffer, 10 μl nucleotide mix (2.5 mM each dNTP), 1 μl RNasin (40 U/μl) (Promega) and 2 μl M-MLV reverse transcriptase were mixed in 50 μl. The reaction was incubated at 37°C for 1 hour followed by 94°C for 5 minutes to inactivate the reverse transcriptase. For AMV-RT, 5 μl AMV-RT 10 × Reaction Buffer, 20 μl nucleotide mix (2.5 mM each dNTP), 1 μl RNasin (40 U/μl) and 2.5 μl reverse transcriptase were used. The reaction was incubated in 50 μl at 25°C for 15 minutes, 42°C for 1 hour followed by 94°C for 5 minutes. All reactions were assembled in PCR hood using aerosol-barrier tips to avoid contamination. Nested – PCR Nested PCR reactions were all assembled in 50 μl, including 5 μl cDNA from the RT reaction as template for the first round PCR or 5 μl first round PCR product as template for the second PCR, 3 μl 10 μM sense primer, 3 μl 10 μM anti-sense primer, 4 μl nucleotide mix (2.5 mM each dNTP), 5 μl 10 × Taq polymerase buffer, 2 units Taq polymerase and MgCl2. The amount of MgCl2 used varied with primer set. Table 2 lists the final Mg++ concentration for every pair of primers. The PCR program is (95°C, 1 min---T°, 1 min---72°C, 2.5 min or 2 min) × 5 cycles --- (95°C, 30 sec---T°, 1 min---72°C, 2.5 min or 2 min) × 30 cycles, where T represents the annealing temperature in Table 2. An extension time of 2.5 min was used for amplicons over 2 kb (amplicons 1, 2 and 3), and extension time of 2 min was used for amplicons less than 2 kb (amplicons 1x, 1y, 4x and 4y). A PCR hood and aerosol-barrier tips were used for assembly of all reactions to avoid contamination. Negative controls lacking template were included for each pair of primers. If any negative control was positive, all PCR reactions in that set were deemed to be contaminated and were discarded. Competing interests The author(s) declare that they have no competing interests. Authors' contributions EY performed the optimizations. JT conceived the study and participated in the design. All authors read and approved the final manuscript. Supplementary Material Additional File 1 Primers for amplification and sequencing the HCV genotype 1a ORF Click here for file Additional File 2 Primers for amplification and sequencing the HCV genotype 1b ORF Click here for file Additional File 3 Optimized PCR conditions for amplifying HCV 1a ORF Click here for file Additional File 4 Optimized PCR conditions for amplifying HCV 1b ORF Click here for file Additional File 5 Optimized nested PCR primer permutations for genotype 1a Click here for file Additional File 6 Optimized nested PCR primer permutations for genotype 1b Click here for file Acknowledgements The Virahep-C clinical study was a cooperative agreement funded by the NIDDK and co-funded by the National Center on Minority Health and Health Disparities (NCMHD), with a Cooperative Research and Development Agreement (CRADA) with Roche Laboratories, Inc. Grant numbers: U01 DK60329, U01 DK 60340, U01 DK60324, U01 DK60344, U01 DK60327, U01 DK60335, U01 DK60352, U01 DK60342, U01 DK60345, U01 DK60309, U01 DK60346, U01 DK60349, U01 DK60341. Other support: National Center for Research Resources (NCRR) General Clinical Research Centers Program grants: M01 RR00645 (New York Presbyterian), M02 RR000079 (University of California, San Francisco), M01 RR16500 (University of Maryland), M01 RR000042 (University of Michigan), M01 RR00046 (University of North Carolina). The participation of the Virahep-C patients is gratefully acknowledged. We thank Ping Wang, Maureen Donlin, Brandon Steel, and Nathan Cannon for technical assistance. We thank Adrian Di Bisceglie and Xiaofeng Fan for helpful discussions. ==== Refs The Global Burden Of Hepatitis C Working Group. Global burden of disease (GBD) for hepatitis C J Clin Pharmacol 2004 44 20 29 14681338 10.1177/0091270003258669 Kim WR The burden of hepatitis C in the United States Hepatology 2002 36 S30 S34 12407574 10.1053/jhep.2002.36791 Major EM Rehermann B M.Stephen F Knipe MD and Howley MP Hepatitis C Virus Fields Virology 2001 Philadelphia: Lippincott Williams & Wilkins 1127 1161 Poynard T Treatment of hepatitis C virus: the first decade Semin Liver Dis 2004 24 Suppl 2 19 24 15346242 10.1055/s-2004-832924 Ferenci P Predictors of response to therapy for chronic hepatitis C Semin Liver Dis 2004 24 Suppl 2 25 31 15346243 10.1055/s-2004-832925 Han JH Houghton M Group specific sequences and conserved secondary structures at the 3' end of HCV genome and its implication for viral replication Nucleic Acids Res 1992 20 3520 1321416 Hijikata M Kato N Ootsuyama Y Nakagawa M Ohkoshi S Shimotohno K Hypervariable regions in the putative glycoprotein of hepatitis C virus Biochem Biophys Res Commun 1991 175 220 228 1847805 10.1016/S0006-291X(05)81223-9 Kato N Ootsuyama Y Tanaka T Nakagawa M Nakazawa T Muraiso K Ohkoshi S Hijikata M Shimotohno K Marked sequence diversity in the putative envelope proteins of hepatitis C viruses Virus Res 1992 22 107 123 1314471 10.1016/0168-1702(92)90038-B Bukh J Miller RH Purcell RH Genetic heterogeneity of hepatitis C virus: quasispecies and genotypes Semin Liver Dis 1995 15 41 63 7597443 Simmonds P Holmes EC Cha TA Chan SW McOmish F Irvine B Beall E Yap PL Kolberg J Urdea MS Classification of hepatitis C virus into six major genotypes and a series of subtypes by phylogenetic analysis of the NS-5 region J Gen Virol 1993 74 ( Pt 11) 2391 2399 8245854 Martell M Esteban JI Quer J Genesca J Weiner A Esteban R Guardia J Gomez J Hepatitis C virus (HCV) circulates as a population of different but closely related genomes: quasispecies nature of HCV genome distribution J Virol 1992 66 3225 3229 1313927 Murakawa K Esumi M Kato T Kambara H Shikata T Heterogeneity within the nonstructural protein 5-encoding region of hepatitis C viruses from a single patient Gene 1992 117 229 232 1322345 10.1016/0378-1119(92)90733-6 Higashi Y Kakumu S Yoshioka K Wakita T Mizokami M Ohba K Ito Y Ishikawa T Takayanagi M Nagai Y Dynamics of genome change in the E2/NS1 region of hepatitis C virus in vivo Virology 1993 197 659 668 8249288 10.1006/viro.1993.1641 Farci P Shimoda A Coiana A Diaz G Peddis G Melpolder JC Strazzera A Chien DY Munoz SJ Balestrieri A Purcell RH Alter HJ The outcome of acute hepatitis C predicted by the evolution of the viral quasispecies Science 2000 288 339 344 10764648 10.1126/science.288.5464.339 Omata M Kato N Recent advances in hepatitis C virus research J Gastroenterol 1994 29 377 382 8061810 Brechot C Hepatitis C virus genetic variability: clinical implications Am J Gastroenterol 1994 89 S41 S47 8048413 Oligo Explorer 1.2 2005 Loakes D Survey and summary: The applications of universal DNA base analogues Nucleic Acids Res 2001 29 2437 2447 11410649 10.1093/nar/29.12.2437 Kolykhalov AA Agapov EV Blight KJ Mihalik K Feinstone SM Rice CM Transmission of hepatitis C by intrahepatic inoculation with transcribed RNA Science 1997 277 570 574 9228008 10.1126/science.277.5325.570 Thomson M Nascimbeni M Gonzales S Murthy KK Rehermann B Liang TJ Emergence of a distinct pattern of viral mutations in chimpanzees infected with a homogeneous inoculum of hepatitis C virus Gastroenterology 2001 121 1226 1233 11677216 10.1053/gast.2001.28669 Tester I Smyk-Pearson S Wang P Wertheimer A Yao E Lewinsohn DM Tavis JE Rosen HR Immune evasion versus recovery following acute hepatitis c virus infection from a shared source. J Exp Med 2005 201 1725 1731 15939788 10.1084/jem.20042284	16321149	PMC1325262	CC BY	2021-01-04 16:25:32	no	Virol J. 2005 Dec 1; 2:88	utf-8	Virol J	2,005	10.1186/1743-422X-2-88	oa_comm

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.