| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | ADVANCED SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Biology, Graduate School of Science, and
2 SORST, Japan Science and Technology Agency, Kobe University, Kobe 657-8501, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionExperimental proceduresReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionExperimental proceduresReferences |
|---|
To understand the genetic background of this mechanism, a useful model is the fruit fly Drosophila, which has been used in various studies in developmental biology. In the case of the adult intestine, the existence of stem cells was also recently demonstrated in the posterior midgut (Micchelli & Perrimon 2006; Ohlstein & Spradling 2006). Interestingly, a juxtacrine signaling receptor Notch (N) is involved in regulation of differentiation, as seen in the case of the mammalian intestinal stem cells (ISC). In both animals, a strong N signal directs the fate of the progenitor cells to the absorptive cells (enterocyte, EC), whereas a weak N signal induces the secretory cells (van Es et al. 2005; Ohlstein & Spradling 2007). However, in the case of the mammalian small intestine, an N signal is also used as a proliferation signal in the transiently amplifying cells (Crosnier et al. 2006), a stage that is not present in the Drosophila midgut.
Accordingly, the progenitor cells produced from the Drosophila ISCs activate N to terminate their further cell division and differentiate into the enteroblast (EB), which has a more restricted cell fate. The EB is reproducibly generated after each asymmetric cell division, although the ISCs have the potential to undergo symmetric cell division to produce two daughter stem cells or two progenitor cells, as is the case with most of the tissue stem cells. Therefore, the question of how to guarantee asymmetry in nearly every cell division becomes an important consideration for the sustainable supply of differentiated cells.
Another issue is the timing of N signal activation. Delta (Dl, a ligand)–N is a juxtacrine signaling system that requires attachment of the cells, respectively, expressing Dl and N. However, the progenitor cell migrates apically after asymmetric cell division of the ISC (Ohlstein & Spradling 2007), suggesting that there should be a mechanism to retain a stable attachment between the two cells. Here, we show that the cell adhesion protein E-cadherin (E-cad) (Oda et al. 1998) is responsible for this interaction of the cells for a certain activation of the N signaling. Moreover, the function of E-cad is not restricted to the assistance of N signaling but also extends to the proper cell morphogenesis of the ISC/EB and upward migration of the EB, both of which are expected to be important for appropriate cell turnover in the adult midgut.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionExperimental proceduresReferences |
|---|
It is known that the N signal is activated in one of the daughter cells of the ISC to produce EB by receiving Dl from the ISC (Ohlstein & Spradling 2007). The activation of N signal should occur after completion of the cell division, otherwise the signal might also be transmitted in the ISC itself, which would terminate its proliferation cycle. However, completion of cell division is also the beginning of the separation of the daughter cells, as the ISCs and EBs generally show scattered distribution in the normal midgut (Micchelli & Perrimon 2006). Also, activation of N signaling requires contact of the cells because the N ligand Dl is a transmembrane protein that cannot diffuse from the ISC. Thus these observations raise the question of how the N signal is activated at the moment after cell division and before cell separation.
In order to address this question, we first focused on the distribution pattern of the normal esg-positive cells. Esg is a Snail/Slug family transcription factor required for maintenance of a diploid state in some developmental fields (Hayashi et al. 1993; Fuse et al. 1994). The adult posterior midgut cells expressing esg are known to be restricted to the small cells existing before the differentiation stage (Micchelli & Perrimon 2006). We further observed that the esg-positive cells frequently occur in pairs (Fig. 1A), which are assumed to be the daughter cells derived from a single esg-positive cell. We base this assumption on the fact that we can occasionally find two attached cells, and the EB is presumed to be gradually moving from the ISC (Fig. 1B, B'' and D–D'').
|
E-cad is required for proper cell type, interaction between ISC and EB, and adhesion with ECs
We found that the interface membranes by which the two esg-positive cells are attached are enriched with E-cad and its intracellular effector Armadillo (Arm, the β-Catenin homologue in Drosophila) (Fig. 1B, B', C–C'' and D–D''). This suggests that both cells are connected by E-cad-mediated cell adhesion just after the cell division and then separated by its down-regulation.
To test this possibility, we next tried to inhibit the E-cad function. A prior confirmation trial using the wing imaginal disc harboring mosaic clones with E-cad RNAi actually showed a significant reduction in the levels of E-cad (Fig. 2A and A'), indicating that the RNAi works well with this construct. In the case of the adult posterior midgut, we analyzed the most posterior portion, which has been well-characterized historically (Fig. S1 in the Supporting Information) (Micchelli & Perrimon 2006; Ohlstein & Spradling 2006). The same E-cad RNAi induced significant decreases in the E-cad and Arm levels (Fig. 3A' and E'), indicating that the RNAi also successfully works in the adult posterior midgut. In examining the cell distribution pattern in the posterior midgut, tightly attached esg-positive cells were rarely found (Fig. 2B), suggesting that a decrease in E-cad to lower levels is essential for separation of the newly divided esg-positive cells. Furthermore, most of the esg-positive cells with E-cad RNAi showed a spherical cell shape (Fig. 2B'), unlike the angular shape in wild type cells. In the case where cells homozygous for shotgun2 (shg2, a strong allele of E-cad mutant) are generated by somatic recombination, similar spherical cells can be observed in the ISCs and even in the ECs (Fig. 2C and C'). Some small spherical cells might be the enteroendocrine cells, which normally show a round shape (data not shown). In either case, this feature also indicates that the angular outlines of the ISC and EB seem to be brought by a mechanical tension through E-cad-mediated adhesion to the cells around the ISC and EB (ECs in most cases). Consistently, when E-cad and Arm were doubly over-expressed to reinforce the E-cad-mediated cell adhesion (Sanson et al. 1996), the cells showed unusually irregular protrusions for attachment to the cell membranes of the surrounding ECs (Fig. 2D and D', Supporting Information Fig. S3).
|
|
We examined the activities in N signaling under this E-cad-depleted condition. In contrast to the normal posterior midgut, where one of the paired esg-positive cells activates N (Fig. 1E), only a small portion of cell pairs showed N activation (Fig. 3A–A''). A statistically significant decrease in the N-active cells can be seen even at 5 days after E-cad knockdown, and this was continuously observed thereafter (Fig. 3C). It has recently been shown that the adult aging during of 3–30 days causes increases in the number of the esg-positive cells and the N-active cells of up to 2.1 (122.7/57.5) and 1.6 (112.5/70.2) times, respectively (Choi et al. 2008). This means that the ratio of the N-active cells to the esg-positive cells rather decreases to 0.75 times. Although this aging event needs a long term (30 days), the effects by E-cad knockdown can be detected within a short term (5 days). Consequently, the decrease in the frequency of the N-active cells by E-cad knockdown is not relevant to adult aging.
Furthermore, the efficiency of differentiating the enteroendocrine cells became slightly higher than normal (Fig. 3B, B' and C). Although this tendency is not always statistically significant, the effect is also consistent with the previous knowledge that weak N signaling induces differentiation to the enteroendocrine cells (Ohlstein & Spradling 2007). The fluctuation in the frequency of the enteroendocrine cells is probably derived from other unknown signaling pathways involved in differentiation to the enteroendocrine cells. In fact, it is known that the weak N signaling can also cause maintenance of the ISC, although no other signaling factor for differentiation to the enteroendocrine cells from the ISC has been identified.
We also checked the effect of double over-expression of E-cad and Arm on N signaling (Supporting Information Fig. S3C). Within a short term after over-expression, we found no significant elevation in N signal activities. However, at 25 days after over-expression, we could obtain the obvious effects of an increase in the N-active cells and a decrease in the enteroendocrine cells. These delayed effects may be explained by the general nature of asymmetric cell division, in which the stem cells normally produce an equal number of the progenitor cells in maximum production. Accordingly, elevation of N signaling does not always result in an increase of the N-active progenitor cells. However, the stem cells sometimes undergo symmetric cell division to produce two daughter stem cells or two progenitor cells. Interference of the former type of symmetric cell division may cause a slow increase in the N-active progenitor cells.
Together with these findings, E-cad functions are considered necessary for N signaling to achieve proper cell differentiation in the adult posterior midgut.
Interruption of E-cad-mediated cell adhesion attenuates migration of EB and its differentiation to EC
Although the tightly attached esg-positive cells become rare under the E-cad-depleted condition, we also noticed that the distance between the paired esg-positive cells frequently becomes abnormally short (Fig. 3E). Accordingly, the relocalization of the developing EB among the various gut cells after reduction of highly accumulated E-cad seems to be also dependent on the low levels of residual E-cad. Consistently, forced increases in E-cad and Arm levels in the esg-positive cells also showed a delay in proper separation of the EB from ISC (Supporting Information Fig. S3B–B'' and C–C'').
We then focused on the cell turnover of the posterior midgut in which the E-cad was depleted. In the normal midgut, cell differentiation from the EB to EC can be recognized by a gradual decrease in Arm accumulation at the cell membrane in conjunction with a gradual increase in ploidy of the nuclei (Fig. 3D). Thus, during the active cell turnover in the adult midgut, we can find small cells with higher Arm levels (ISCs and EBs) and large cells with lower Arm levels (ECs), as well as intermediate-sized cells with moderate Arm levels (early ECs). When E-cad was depleted, the cell composition in the posterior midgut was altered. We could only find small cells (ISCs and EBs) and large cells (ECs), both of which had lower Arm levels caused by reduced E-cad function (Fig. 3E). Intermediate-sized early ECs were rarely found, indicating that the cell turnover was attenuated, probably because of insufficient N signaling in the progenitor cells (Fig. 3A–A'').
Cell mass formation in the ISC-tumor by N inactivation is achieved by E-cad-mediated cell adhesion
The above finding led us to consider whether a tumor-like cell sheet with the esg-positive cells in the N-depleted midgut (Micchelli & Perrimon 2006) is caused by a defect of down-regulation of E-cad. In fact, the N-depleted esg-positive cells showed a high level of E-cad and Arm (Fig. 4A, B–B'') in forming a cell sheet. Furthermore, when E-cad was depleted from the N-depleted esg-positive cells, they were dispersed and cell sheet formation was largely blocked (Fig. 4C). Accordingly, one of the N functions in the esg-positive cells is partial destruction of E-cad to separate the ISC and EB.
|
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionExperimental proceduresReferences |
|---|
Formation of the spherical shape in the ISC and EB as a result of E-cad depletion indicates that the edgy cell morphology in the normal ISC and EB as well as their proper attachment to the surrounding ECs are defined by the levels of E-cad. This observation is consistent with the previous knowledge that inactivation of 
-catenin, an intracellular effector of Cadherins, leads to a spherical cell shape in the colon carcinoma cell DLD-1 (Watabe-Uchida et al. 1998; Takeichi & Abe 2005). The edgy cell shape in the normal ISC and EB seems to be important for appropriate positioning of the cells among the gut cell populations. For example, the pairs of E-cad-depleted esg-positive cells do not sufficiently extend the distance between the cells (Fig. 3E) when compared with wild type. It has also been known that the ISCs are located on the basement membrane whereas the EBs are at a slightly more apical position. Moreover, the early ECs reside at a further upper position (Ohlstein & Spradling 2006). In addition to these findings, the apically-directed migration of the EB associated with normal cell differentiation may also depend on sliding on the EC membrane, which requires E-cad-mediated cell adhesion. During this migration of the EB, Cadherin-flow, a recently discovered apicobasal movement of the mammalian Cadherin cluster, may be involved in the movement of the cell position (Kametani & Takeichi 2007).
Cell adhesion allows ISC and EB to interact via Dl-N binding
The juxtacrine N signal is required for the EB establishment at the end of the asymmetric cell division of the ISC. However, the moment of completing the cell division is equal to that of beginning the cell separation. Thus, the period after cell division and before cell separation is eventually absent unless both daughter cells actively adhere. To retain this moment and thus ensure a certain interaction of the two cells, stabilization of cell adhesion should be consistent. Similar reinforcement of N signaling assisted by E-cad-mediated cell adhesion has previously been shown in the case of the developing wing disc, where the DV boundary cells activate N (Sasaki et al. 2007). In this case, N localizes to the subapical complex (SAC) or adherens junction (AJ) on the lateral membrane of the columnar epithelial cell, whereas E-cad localizes to the AJ. As the two junctions are topologically very close, interaction between N and its ligands was actually affected when the E-cad levels were changed. In the case of the ISC and EB, distribution of Dl and N proteins does not seem to depend on their apicobasal polarity but extends throughout the cell surface. In contrast, E-cad is particularly observed at the interface membrane. Furthermore, duration of N signaling in the EBs is temporal, whereas that in the wing DV boundary cells occurs for a long period. Thus, the Drosophila intestinal EB can be interpreted as another example showing the assistance of E-cad to N signaling, although the spatiotemporal pattern for interacting with the adjacent cell is largely different from that in the wing columnar epithelial cells. A similar unconventional distribution pattern of various cell junctional proteins along the apicobasal axis in the ISC and EB was also confirmed by other cell junctional markers such as phosphorylated Moesin (pMoe), aPKC and Coracle (Supporting Information Fig. S2).
E-cad helps N whereas N breaks E-cad: a feedback relationship
Knockdown of E-cad in the N-depleted ISC tumor resulted in a strong suppression of cell mass formation. This suggests that destruction of E-cad-mediated adhesion between ISC and EB may be a novel N role in the normal adult midgut. After activation of N signaling in the EB, the levels in E-cad are reduced to help down-regulate the N signal and to also separate the EB from ISC. In detail, it is assumed that down-regulation of E-cad in the course of differentiation from the progenitor cell to EC should be cell autonomous and composed of two steps. One is a modest and local lowering at the interface membrane during cell separation. The other is a strong and global reduction in the entire cell membrane during differentiation to the EC. However, in the case of the ISC, only a single-step down-regulation would be expected. This down-regulation is parallel with the former down-regulation in the EB but its non-cell autonomous character should be observed. Because of the well-known nature of homophilic binding between the two molecules of the extracellular domains of E-cad, its increase at the interface membrane of one cell may absorb and concentrate the binding partners on the other side. Conversely, down-regulation of the E-cad levels on the EB may also induce diffusion of the partner E-cad molecules on the ISC to its entire cell membrane.
Reduction of E-cad alone is not sufficient for differentiation to the EC, based on the results of double knockdown of N and E-cad (Fig. 4C). As a result of this experiment, the cells can be separated but they cannot differentiate to the EC. Thus, N activation definitely elicits another direct output necessary for differentiation to the EC. Together with these processes, a negative feedback from N to E-cad facilitates progression of the differentiation stage and inhibition of the tumor-like cell sheet formation (Fig. 5). The latter observation is, unexpectedly, the opposite of a well-known condition in which various malignant tumor cells down-regulate E-cad to cause invasion and metastasis (Takeichi 1991; Friedl & Wolf 2003; Guarino 2007).
|
Together with the previously known roles of N in the posterior midgut (Micchelli & Perrimon 2006; Ohlstein & Spradling 2006), we now understand that the N signaling in the EB exhibits various effects on, for example, terminating the ISC proliferation, determining the progenitor cell's fate, and attenuating cell adhesion between the ISC and progenitor cell. Consequently, one of the direct effects of Notch (N) inactivation is the continuation of the E-cad-mediated cell adhesion, which is a temporal event between the ISC and EB under normal conditions. However, inactivation of N might not activate ISC proliferation as previously thought (Micchelli & Perrimon 2006; Ohlstein & Spradling 2006). In fact, it is difficult to detect a significant difference in frequencies of the phospho-Histone H3 (pH3)-positive cells (M-phase cells) between wild type and the N-inactive cell tumor (Fig. 4D and E). This suggests that the continuous adhesion between the ISCs can sufficiently account for the tumor formation without special stimulation of cell proliferation. The tumor can be generated by using normal levels of cell proliferation signal(s) unless the cell proliferation rate is increased. In this case, signal(s) other than N should act as a genuine proliferation signal in the ISC. The possibility of such a proliferation signal could be demonstrated by searching for suppressors against the inactive N-induced ISC tumor.
|
Experimental procedures |
|---|
|
TopAbstractIntroductionResultsDiscussionExperimental proceduresReferences |
|---|
esg-GAL4 (GETDB NP6267 line, (Hayashi et al. 2002)) was derived from the Drosophila Genetic Resource Center (Japan). UAS-E-cadIR was derived from the National Institute of Genetics (Japan). tub-GAL80ts (McGuire et al. 2003), UAS-arm, UAS-GFP(S65T), and UAS-NIR (Presente et al. 2002) were derived from the Bloomington Stock Center (University of Indiana). UAS-E-cad (M. B. O’Connor, University of Minnesota), E(spl)mβ-CD2 (de Celis et al. 1998) (K. Matsuno, Tokyo University of Science, Japan) and GBE + Su(H)m8 (Furriols & Bray 2001) (S. Bray, University of Cambridge) were gifts from each laboratory. All of the RNAi lines were confirmed to show phenotypes that were indistinguishable from their mutant phenotypes. Mouse anti-β-Galactosidase (Promega), rabbit anti-β-Galactosidase (Cappel), mouse anti-Rat CD2 antibodies (Serotec), rabbit anti-phospho-ERM (Cell signaling), rabbit anti-pH3 (Upstate), rabbit anti-aPKC (Santa Cruz) were purchased. Mouse anti-Arm, rat anti-E-cad, mouse anti-Coracle, mouse anti-Delta and mouse anti-Prospero (Pros) antibodies were derived from DSHB (University of Iowa).
GAL80ts technique
As E-cad and N signals are essential throughout the Drosophila lifespan, simple alteration of the gene function results in lethal at earlier stages. Thus, we employed a technique with a temperature-sensitive version of GAL80 (GAL80ts, (McGuire et al. 2003)) which permits us to control gene expression irrespective of developmental stages. Thus, the flies possessing GAL4, UAS-gene, and tub-GAL80ts does not induce expression of UAS-gene at a lower temperature (19 °C), but can express it at a higher temperature (29 °C). Accordingly, we grew the flies with the desired genes at 19 °C and shifted the temperature to 29 °C at the adult stage. After incubation for an appropriate duration (6–30 days), the adult flies were dissected for immunofluorescence staining.
Immunofluorescence
The isolated abdomens were dissected in phosphate buffered saline (PBS) to remove the abdominal cuticle by using ultraprecise scissors. The obtained whole gut was fixed in 4% paraformaldehyde in PEM buffer (0.1 M PIPES pH 6.9/2 mM EGTA/1 mM MgSO4) for 20 min. After fixation, the organs were rinsed thoroughly with PBT (PBS containing Triton X-100), and then soaked in a fluorescence intensifier (Image-iT FX Signal Enhancer, Invitrogen) for 30 min. The first antibodies were diluted in PBT with an appropriate concentration depending on the titer of each antibody (e.g. 1/200 for anti-β-Galactosidase antibody, 1/20 for anti-Arm antibody). The second antibody (labeled with Alexa Fluor 555 (Invitrogen)) was diluted in PBT 1/200. The incubation of the organs with each of the first and second antibody solution was performed at 37 °C for 1 h or at 4 °C overnight. In some cases, nuclei were counterstained with TO-PRO-3 (Invitrogen). After rinsing out the excess second antibody with PBT, the midguts were observed by laser confocal microscope Digital Eclipse C1 (Nikon).
|
Acknowledgements |
|---|
|
Footnotes |
|---|
* Correspondence: yamadach{at}kobe-u.ac.jp
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionExperimental proceduresReferences |
|---|
Crosnier, C., Stamataki, D. & Lewis, J. (2006) Organizing cell renewal in the intestine: stem cells, signals and combinatorial control. Nat. Rev. Genet. 7, 349–359.[CrossRef][Medline]
de Celis, J.F., Tyler, D.M., de Celis, J. & Bray, S.J. (1998) Notch signalling mediates segmentation of the Drosophila leg. Development 125, 4617–4626.[Abstract]
Friedl, P. & Wolf, K. (2003) Tumour-cell invasion and migration: diversity and escape mechanisms. Nat. Rev. Cancer. 3, 362–374.[CrossRef][Medline]
Furriols, M. & Bray, S. (2001) A model Notch response element detects Suppressor of Hairless-dependent molecular switch. Curr. Biol. 11, 60–64.[CrossRef][Medline]
Fuse, N., Hirose, S. & Hayashi, S. (1994) Diploidy of Drosophila imaginal cells is maintained by a transcriptional repressor encoded by escargot. Genes Dev. 8, 2270–2281.
Guarino, M. (2007) Epithelial-mesenchymal transition and tumour invasion. Int. J. Biochem. Cell Biol. 39, 2153–2160.[CrossRef][Medline]
Hayashi, S., Hirose, S., Metcalfe, T. & Shirras, A.D. (1993) Control of imaginal cell development by the escargot gene of Drosophila. Development 118, 105–115.[Abstract]
Hayashi, S., Ito, K., Sado, Y., Taniguchi, M., Akimoto, A., Takeuchi, H., Aigaki, T., Matsuzaki, F., Nakagoshi, H., Tanimura, T., Ueda, R., Uemura, T., Yoshihara, M. & Goto, S. (2002) GETDB, a database compiling expression patterns and molecular locations of a collection of Gal4 enhancer traps. Genesis 34, 58–61.[CrossRef][Medline]
Kametani, Y. & Takeichi, M. (2007) Basal-to-apical cadherin flow at cell junctions. Nat. Cell Biol. 9, 92–98.[CrossRef][Medline]
Lee, T. & Luo, L. (2001) Mosaic analysis with a repressible cell marker (MARCM) for Drosophila neural development. Trends Neurosci. 24, 251–254.[CrossRef][Medline]
McGuire, S.E., Le, P.T., Osborn, A.J., Matsumoto, K. & Davis, R.L. (2003) Spatiotemporal rescue of memory dysfunction in Drosophila. Science 302, 1765–1768.
Micchelli, C.A. & Perrimon, N. (2006) Evidence that stem cells reside in the adult Drosophila midgut epithelium. Nature 439, 475–479.[CrossRef][Medline]
Oda, H., Tsukita, S. & Takeichi, M. (1998) Dynamic behavior of the cadherin-based cell–cell adhesion system during Drosophila gastrulation. Dev. Biol. 203, 435–450.[CrossRef][Medline]
Ohlstein, B. & Spradling, A. (2006) The adult Drosophila posterior midgut is maintained by pluripotent stem cells. Nature 439, 470–474.[CrossRef][Medline]
Ohlstein, B. & Spradling, A. (2007) Multipotent Drosophila intestinal stem cells specify daughter cell fates by differential notch signaling. Science 315, 988–992.
Presente, A., Shaw, S., Nye, J.S. & Andres, A.J. (2002) Transgene-mediated RNA interference defines a novel role for notch in chemosensory startle behavior. Genesis 34, 165–169.[CrossRef][Medline]
Radtke, F. & Clevers, H. (2005) Self-renewal and cancer of the gut: two sides of a coin. Science 307, 1904–1909.
Sanson, B., White, P. & Vincent, J.P. (1996) Uncoupling cadherin-based adhesion from wingless signalling in Drosophila. Nature 383, 627–630.[CrossRef][Medline]
Sasaki, N., Sasamura, T., Ishikawa, H.O., Kanai, M., Ueda, R., Saigo, K. & Matsuno, K. (2007) Polarized exocytosis and transcytosis of Notch during its apical localization in Drosophila epithelial cells. Genes Cells 12, 89–103.
Takeichi, M. (1991) Cadherin cell adhesion receptors as a morphogenetic regulator. Science 251, 1451–1455.
Takeichi, M. & Abe, K. (2005) Synaptic contact dynamics controlled by cadherin and catenins. Trends Cell Biol. 15, 216–221.[CrossRef][Medline]
van Es, J.H., van Gijn, M.E., Riccio, O., van den Born, M., Vooijs, M., Begthel, H., Cozijnsen, M., Robine, S., Winton, D.J., Radtke, F. & Clevers, H. (2005) Notch/
-secretase inhibition turns proliferative cells in intestinal crypts and adenomas into goblet cells. Nature 435, 959–963.[CrossRef][Medline]
Watabe-Uchida, M., Uchida, N., Imamura, Y., Nagafuchi, A., Fujimoto, K., Uemura, T., Vermeulen, S., van Roy, F., Adamson, E.D. & Takeichi, M. (1998) -Catenin–vinculin interaction functions to organize the apical junctional complex in epithelial cells. J. Cell Biol. 142, 847–857.
Received: 16 April 2008
Accepted: 20 September 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | ADVANCED SEARCH | TABLE OF CONTENTS |
