| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 281, Issue 45, 34716-34724, November 10, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Center for Cardiovascular Sciences, Albany Medical College, Albany, New York 12208
Received for publication, August 11, 2006 , and in revised form, August 29, 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
The dynamic characteristics of the intermediate filament network may be regulated by protein phosphorylation. Vimentin is the most abundant intermediate filament protein in various cell types, including smooth muscle cells (2, 5, 6). Vimentin phosphorylation in association with vimentin disassembly and spatial reorganization occurs during mitosis or in response to extracellular stimulation (8, 9). In cultured smooth muscle cells, contractile stimulation triggers vimentin phosphorylation at Ser-56 concurrently with vimentin partial disassembly and spatial reorientation (6).
The disassembly and spatial reorganization of the vimentin network may regulate the translocation of certain molecules (7, 10, 11). The adapter protein p130 Crk-associated substrate (CAS)2 has been shown to participate in the signaling processes that regulate smooth muscle contraction and cell migration (12-14). Our recent study has suggested that vimentin phosphorylation and disassembly are related to CAS redistribution during contractile activation of smooth muscle (10). In addition, external stress initiates Rho kinase redistribution associated with vimentin depolymerization in fibroblasts and the translocation of Ca2+/calmodulin-dependent protein kinase II in differentiated smooth muscle cells, which may be an important event for cell signaling (7, 11).
p21-activated kinase (PAK) may be an upstream regulator of the vimentin network (6, 9). In cultured smooth muscle cells, agonist-mediated vimentin phosphorylation at Ser-56 and spatial reorientation of the vimentin network are inhibited by silencing of PAK1, a dominant isoform in smooth muscle (6, 15). Additionally, PAK has been implicated in modulating smooth muscle contraction; introduction of an active PAK isoform into smooth muscle potentiates force development at constant intracellular calcium (16). Expression of an inactive PAK1 mutant attenuates migration of cultured smooth muscle cells in response to platelet-derived growth factor (15).
In response to external stimulation, PAK undergoes autophosphorylation at Thr-423, which increases PAK activity for substrates (17, 18). In addition to the small GTPases Cdc42 and Rac1, the activity of PAK may be regulated by the paxillin kinase linker/PIX (PAK-interacting exchange factor; guanine nucleotide exchange factor) (19-21). CAS has been shown to interact with the paxillin kinase linker/PIX via CrkII and paxillin (21-23). Thus, CAS could be involved in the regulation of PAK activation.
The aim of this study was to test the hypothesis that vimentin phosphorylation at Ser-56 may play a critical role in regulating PAK activation and the spatial reorientation of the vimentin network. Our results demonstrate that PAK-mediated vimentin phosphorylation at Ser-56 leads to increases in vimentin disassembly and CAS dissociation from the vimentin network during 5-hydroxytryptamine (5-HT; serotonin) stimulation. Expression of a non-phosphorylatable vimentin mutant (S56A) attenuates the increase in PAK activation and the structural rearrangement of vimentin filaments upon agonist stimulation.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-actin as determined by immunoblot analysis. Assessment of Protein Phosphorylation—Smooth muscle cells were treated with SDS sample buffer containing 1.5% dithiothreitol, 2% SDS, 80 mM Tris-HCl (pH 6.8), 10% (v/v) glycerol, and 0.01% bromphenol blue. Cells were boiled in the buffer for 4 min and separated by SDS-PAGE. Proteins were transferred to nitrocellulose membranes, after which the membranes were blocked with 2% bovine serum albumin. To analyze PAK phosphorylation, membranes were reacted with anti-phospho-PAK1 (Thr-423)/PAK2 (Thr-403) antibody (Cell Signaling Biotechnology, Inc.), followed by horseradish peroxidase-conjugated anti-rabbit Ig (ICN Biomedicals). Proteins were visualized by enhanced chemiluminescence (SuperSignal, Pierce) using a Fuji LAS-3000 imaging system. The membranes were stripped and reprobed with anti-PAK1 antibody (Cell Signaling Biotechnology, Inc.). For determination of vimentin phosphorylation, membranes were probed with a phosphorylation site-specific antibody for Ser-56 (anti-phosphovimentin (Ser-56) antibody), stripped, and reprobed with anti-vimentin monoclonal antibody (clone RV202, BD Biosciences), followed by incubation with horseradish peroxidase-conjugated anti-mouse Ig (Amersham Biosciences) (6). Phosphoprotein and total protein were quantified by scanning densitometry of immunoblots (Fuji MultiGauge software). Changes in protein phosphorylation were expressed as the magnitude increase over the phosphorylation levels in unstimulated cells. The luminescent signals from all immunoblots were within the linear range.
Preparation of Recombinant Proteins—PCR-mediated mutagenesis was carried out on pEGFP-vimentin (kindly provided by Dr. Robert D. Goldman, Northwestern University, Chicago, IL) to generate mutant S56A vimentin. DNA sequencing was used to confirm the mutation of vimentin. The PCR product and wild-type vimentin cDNA were subcloned into pGEX-4T at the BamHI and EcoRI sites, followed by transformation of Escherichia coli BL21. Recombinant proteins were harvested and purified according to the manual accompanying the glutathione S-transferase gene fusion system of Amersham Biosciences. Thrombin (Amersham Biosciences) was used to release the recombinant protein from glutathione-Sepharose beads and was then removed by incubation with p-aminobenzamidine-agarose beads (Sigma).
In Vitro Kinase Assay—Purified wild-type or mutant S56A vimentin (0.1 mg/ml) was incubated with 2 µg/ml activated PAK (Upstate) for 10 or 30 min in kinase buffer containing 20 mM HEPES (pH 7.5), 60 mM NaCl, 2 mM MgCl2, 5 mM EGTA, and 100 µM ATP. The reaction mixture was boiled in SDS sample buffer for 5 min, and proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes. The membranes were probed with anti-phosphovimentin (Ser-56) antibody, stripped, and reprobed with anti-vimentin antibody.
Immunofluorescence Molecular Imaging—We developed a molecular immunostaining technique to visualize vimentin filament assembly in vitro. Briefly, purified wild-type and mutant S56A vimentin (0.2 mg/ml) were placed in kinase buffer in the presence or absence of PAK for 30 min. NaCl (150 mM) was then added to the reaction mixture, followed by incubation at 37°C for 1 h to initiate filament formation. The mixture was treated with 0.1% glutaraldehyde at room temperature for 30 min to cross-link vimentin filaments and plated on slides, followed by fixation with 4% paraformaldehyde for 15 min. The purified molecules were reacted with anti-vimentin monoclonal antibody, followed by incubation with Alexa 488-conjugated anti-mouse IgG (Molecular Probes, Eugene, OR). Vimentin filament formation was examined and analyzed under a Zeiss laser scanning confocal microscope.
Analysis of the Ratios of Soluble to Insoluble Protein in Cells— The amounts of soluble and insoluble vimentin were evaluated by the method described previously (4, 6). Smooth muscle cells were washed with ice-cold Hanks' balanced salt solution and then scraped, collected, and centrifuged at 8000 x g for 5 min. The resulting pellets were incubated at 37 °C for 30 min in buffer containing 1% Nonidet P-40, 10% (v/v) glycerol, 20 mM HEPES (pH 7.6), 150 mM NaCl, 2 mM sodium orthovanadate, 2 mM molybdate, 2 mM sodium pyrophosphate, and protease inhibitors (2 mM benzamidine, 0.5 mM aprotinin, and 1 mM phenylmethylsulfonyl fluoride). The soluble (disassembled) and insoluble (assembled) fractions were collected after centrifugation at 2100 x g for 30 min at 4 °C and assessed by immunoblot analysis using anti-vimentin antibody. The soluble/insoluble vimentin ratios were determined after scanning densitometry of the immunoblots. To assess the ratios of soluble to insoluble (vimentin-associated) CAS and the ratios of insoluble CAS to cytoskeletal vimentin, the membranes were stripped and reprobed with anti-CAS antibody (clone 24, BD Biosciences).
Cell Transfection and Assessment of Protein Expression— Cultured smooth muscle cells were transfected with plasmids encoding enhanced green fluorescent protein (EGFP)-tagged wild-type and mutant S56A vimentin using LipofectamineTM and PLUSTM transfection kits (Invitrogen). After 48 h, the cells were incubated in serum-free medium for 1 day. Cells were treated with 1x SDS sample buffer to extract and denature proteins, after which the proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes. To assess the expression of recombinant proteins, the membranes were detected with anti-EGFP antibody (clone GFP02, Lab Vision Corp.) or anti-vimentin antibody.
Cell Fluorescence Analysis—After transfection with plasmids containing EGFP-labeled proteins, cells on coverslips were fixed with paraformaldehyde and mounted on slides using Fluoromount-G (Fisher). Cells were also stained with 4',6-diamidino-2-phenylindole to visualize nuclei. Cell images were viewed under a confocal fluorescence microscope. Cells with reorganized vimentin filaments were determined and quantitatively analyzed by the method described previously (6).
Statistical Analysis—All statistical analysis was performed using Prism Version 4 software (GraphPad Software, San Diego, CA). Comparison among multiple groups was performed by one-way analysis of variance, followed by Tukey's multiple comparison post test. Differences between pairs of groups were analyzed by the Student-Newman-Keuls test or Dunn's test. Values of n refer to the number of experiments performed to obtain each value. p < 0.05 was considered to be significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
The activation of smooth muscle cells with 5-HT led to the enhancement of PAK phosphorylation at Thr-423. As shown in Fig. 1A, PAK1 phosphorylation at Thr-423 was increased 5 min after stimulation with 5-HT and was sustained for the 15-min duration. The phosphorylation levels in smooth muscle cells were increased by 1.5-fold 10-15 min after activation with 5-HT (n = 4) (Fig. 1B).
|
To assess whether Ser-56 is a major phosphorylation site on vimentin mediated by PAK, phosphorylation of mutant S56A vimentin was also evaluated. The phosphorylation levels of the S56A mutant were relatively lower compared with those of wild-type vimentin, and they were not significantly increased after the addition of PAK (n = 4; p < 0.05) (Fig. 2).
PAK Induces Disassembly of Wild-type (but Not S56A Mutant) Vimentin Filaments in Vitro—Vimentin phosphorylation is associated with vimentin disassembly in smooth muscle cells upon agonist activation (6). To assess whether vimentin phosphorylation at Ser-56 by PAK modulates its assembly/disassembly, we developed an in vitro filament assembly assay. Soluble wild-type vimentin and its S56A mutant were treated with or without activated PAK, after which 150 mM NaCl was added to initiate vimentin filament formation. These proteins were plated on slides and immunostained with anti-vimentin antibody.
In the absence of PAK, wild-type vimentin was filamentous, and a portion of the vimentin filaments was assembled into a supercoiled structure (Fig. 3a). The formation of supercoiled vimentin filaments might be due to lack of tension imposed on the filaments. Treatment with PAK led to the disassembly of wild-type vimentin intermediate filaments (Fig. 3b). Likewise, mutant S56A vimentin displayed a filamentous structure when it was not treated with PAK (Fig. 3c). However, incubation of the mutant with PAK did not trigger the disassembly of mutant S56A vimentin filaments (Fig. 3d). These results suggest that vimentin phosphorylation at Ser-56 by PAK leads directly to vimentin filament disassembly.
The Amount of CAS in Insoluble Vimentin Fractions Is Reduced in Smooth Muscle Cells Stimulated with 5-HT—The vimentin framework binds to certain signaling partners such as Ca2+/calmodulin-dependent protein kinase II and Rho kinase (7, 11). The adapter protein CAS and the small GTPase Cdc42 have been shown to regulate active force development in smooth muscle (12, 14, 24, 25). To determine whether CAS and Cdc42 associate with the vimentin network, soluble and insoluble (cytoskeletal) vimentin from unstimulated smooth muscle cells were separated by SDS-PAGE, and blots of the fractions were probed with antibodies against CAS, Cdc42, and vimentin. CAS (but not Cdc42) was found in the fractions of both insoluble and soluble vimentin (Fig. 4A). Quantification analysis showed that 25% of total CAS was associated with insoluble vimentin fractions (n = 8) (Fig. 4B).
|
The distribution of CAS in different fractions in response to agonist stimulation was also evaluated. Activation with 5-HT led to a decrease in the amount of CAS in insoluble fractions. The insoluble/soluble CAS ratios elicited by 5-HT stimulation were lower compared with those in unstimulated smooth muscle cells (n = 5; p < 0.05) (Fig. 4D).
|
We first assessed whether vimentin phosphorylation in response to 5-HT stimulation is depressed in cells expressing mutant S56A vimentin. pEGFP plasmid alone or plasmid encoding wild-type or mutant S56A vimentin was transfected into smooth muscle cells. Immunoblot analysis showed that EGFP-labeled recombinant vimentin was highly expressed; 
90% of the total vimentin was recombinant vimentin in the transfected cells (Fig. 5, A and B). Smooth muscle cells expressing wild-type or mutant S56A vimentin were stimulated with 10 µM 5-HT for 5-15 min or left unstimulated. Blots of protein extracts from these cells were probed using anti-phosphovimentin (Ser-56) antibody, stripped, and reprobed with anti-EGFP antibody.
Stimulation with 5-HT resulted in increases in vimentin phosphorylation at Ser-56 in cells transfected with plasmid encoding wild-type vimentin (Fig. 5C). The phosphorylation levels in cells expressing wild-type vimentin were increased by 1.7-fold after stimulation with 5-HT for 10-15 min (Fig. 5D), which is similar to our previous results in untransfected cells (6). However, the phosphorylation levels in response to 5-HT stimulation in cells expressing the S56A mutant were significantly lower compared with cells expressing wild-type vimentin (n = 4; p < 0.05) (Fig. 5, C and D).
|
Activation with 5-HT increased the amount of soluble vimentin in cells expressing wild-type vimentin; however, the 5-HT-mediated increase in the level of soluble vimentin was attenuated in cells expressing the S56A mutant (Fig. 6A). The ratios of soluble to cytoskeletal vimentin in response to 5-HT stimulation were lower in cells expressing mutant S56A vimentin than in cells expressing wild-type vimentin (n = 4; p < 0.05) (Fig. 6B).
Redistribution of CAS Elicited by 5-HT Is Depressed in Cells Expressing S56A Mutant Vimentin—To evaluate whether vimentin phosphorylation at Ser-56 is required for CAS redistribution, we assessed the ratios of vimentin-associated (insoluble) CAS to cytoskeletal vimentin in unstimulated and stimulated (10 µM 5-HT, 15 min) smooth muscle cells expressing wild-type or mutant S56A vimentin. Expression of the S56A mutant suppressed the decrease in the CAS/vimentin ratios in response to 5-HT stimulation. The ratios of insoluble CAS to cytoskeletal vimentin elicited by 5-HT were significantly decreased in cells expressing wild-type vimentin compared with unstimulated cells. In contrast, the ratios of insoluble CAS to cytoskeletal vimentin were not significantly different in unstimulated and 5-HT-stimulated cells expressing mutant S56A vimentin (n = 7; p < 0.05) (Fig. 7A).
Moreover, we assessed the effects of mutant S56A vimentin on the vimentin-associated/soluble CAS ratios. The insoluble/soluble CAS ratios were decreased in 5-HT-stimulated cells expressing wild-type vimentin compared with unstimulated cells. However, the ratios of insoluble to soluble CAS in response to 5-HT stimulation were not reduced in cells expressing the S56A mutant compared with unstimulated cells (n = 10; p < 0.05) (Fig. 7B).
|
|
Phosphorylation of PAK1 at Thr-423 in response to 5-HT stimulation was enhanced in cells expressing wild-type vimentin. Nevertheless, the agonist-induced PAK1 phosphorylation at this site was significantly suppressed in cells expressing the S56A mutant compared with cells expressing wild-type vimentin (n = 5; p < 0.05) (Fig. 8).
|
Vimentin phosphorylation at Ser-56 is crucial for its spatial reorganization in smooth muscle cells upon 5-HT stimulation. In unstimulated cells transfected with wild-type vimentin (Fig. 9A, panel a), vimentin exhibited curved filamentous structures. Exposure of cells transfected with wild-type vimentin to 5-HT induced vimentin filament straightness (Fig. 9A, panel b). The 5-HT-induced spatial reorientation of the EGFP-tagged vimentin fibers is similar to that of native vimentin filaments in smooth muscle cells (6). In unstimulated cells expressing mutant S56A vimentin, vimentin fibers also showed the curvature (Fig. 9A, panel c). However, stimulation with 5-HT did not induce the spatial rearrangement; the intermediate filaments remained curved in stimulated cells transfected with the S56A mutant (Fig. 9A, panel d). Stimulation with 5-HT resulted in an increase in the number of wild-type (but not mutant S56A) vimentin-expressing cells that displayed the reorganized intermediate filaments (Fig. 9B).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
In cultured smooth muscle cells, PAK silencing attenuates vimentin phosphorylation upon contractile activation (6). In this study, stimulation with 5-HT induced PAK1 autophosphorylation at Thr-423 in smooth muscle cells. In the inactive state, the N-terminal domain of PAK interacts with the C-terminal catalytic motif, forming an intramolecular inhibitory conformation and diminishing enzyme activity. Autophosphorylation at Thr-423 has been proposed to induce a conformational change for the activation of PAK (18). Thus, the results suggest that stimulation with 5-HT may increase PAK1 activity in smooth muscle cells.
|
There is evidence that the vimentin filament framework harbors Rho kinase and Ca2+/calmodulin-dependent protein kinase II, which may be dissociated from the network when vimentin partial disassembly or structural rearrangement occurs (7, 11). The adapter protein CAS may participate in the signaling process that mediates smooth muscle force development (12, 14). In this study, CAS associated with the vimentin cytoskeleton, and agonist activation triggered the dissociation of CAS from cytoskeletal vimentin. The amount of CAS dissociated from cytoskeletal vimentin was 50% of total CAS associated with vimentin filaments. Furthermore, the expression of mutant S56A vimentin in smooth muscle cells attenuated CAS redistribution during agonist stimulation. Because vimentin phosphorylation at Ser-56 and partial disassembly were also inhibited in cells expressing the S56A mutant, we propose that contractile stimulation of smooth muscle cells leads to vimentin phosphorylation at this position, inducing vimentin disassembly and CAS release from the vimentin filament network.
Changes in the physiological status of a substrate may conversely serve as a regulator of its upstream enzyme, forming a feedback mechanism. As mentioned above, PAK is able to mediate vimentin phosphorylation both in vitro and in vivo. In this study, PAK activation in response to 5-HT stimulation was inhibited in cells expressing mutant S56A vimentin, suggesting that vimentin phosphorylation at Ser-56 reciprocally regulates PAK activity. The positive feedback mechanism may render PAK in an active state during agonist stimulation. The mechanisms by which vimentin phosphorylation at this residue regulates PAK activity are unknown. Because CAS can be dissociated from the vimentin network during contractile activation, it is possible that the increase in soluble CAS mediated by vimentin phosphorylation may facilitate the activation of PAK. When activated, CAS may be able to form a protein complex containing CrkII, paxillin, the p95 paxillin kinase linker, and PAK, which may mediate the activation of PAK (21, 22).
In addition to the regulation of signaling molecule redistribution, vimentin phosphorylation may be associated with spatial restructuring of the vimentin network (6). In this study, 5-HT-stimulated vimentin phosphorylation at position 56 was reduced in cells expressing mutant S56A vimentin. Moreover, spatial reorganization of the vimentin cytoskeleton in response to stimulation with 5-HT was also attenuated in cells expressing the S56A mutant. These results lead us to suggest that phosphorylation at Ser-56 is required for vimentin structural reorganization during agonist activation of smooth muscle cells.
As mentioned above, PAK treatment of vimentin protein in vitro triggered almost complete filament disassembly, whereas vimentin filaments underwent partial disassembly and spatial rearrangement in the cells in response to agonist activation. The difference between the in vitro and in situ studies may stem from the lower extent of vimentin phosphorylation in the cells. Our previous investigation showed that 65% of soluble vimentin is phosphorylated at Ser-56 in smooth muscle cells stimulated with 5-HT, which is associated with the partial vimentin disassembly as estimated by a cell fractionation assay, and is regulated by PAK (6). Thus, we postulate that smooth muscle cells may possess a limited amount of PAK and/or that PAK has limited access to vimentin. Upon agonist stimulation, vimentin in the cells undergoes partial phosphorylation and partial disassembly, which may facilitate the spatial reorientation of vimentin filaments. In contrast, in the in vitro study, a sufficient amount of PAK is added to the reaction system, and/or PAK has full access to vimentin. PAK treatment is able to induce comprehensive filament disassembly.
|
Thus, we propose a unique vimentin cytoskeletal signaling mechanism in smooth muscle cells. Activation of PAK by 5-HT stimulation may catalyze vimentin phosphorylation at Ser-56, leading to vimentin disassembly and the dissociation of the adapter protein CAS from the vimentin network. The released CAS may facilitate the formation of a multiprotein complex including CAS and PAK, maintaining PAK in an active state. In addition, vimentin phosphorylation at this position may induce the spatial reorientation of vimentin filaments, which may be associated with contractile element reorganization in response to agonist stimulation (Fig. 10).
|
FOOTNOTES |
|---|
1 To whom correspondence should be addressed: Center for Cardiovascular Sciences, Albany Medical College, 47 New Scotland Ave., MC-8, Albany, NY 12208. Tel.: 518-262-6416; Fax: 518-262-8101; E-mail: tangd{at}mail.amc.edu.
2 The abbreviations used are: CAS, Crk-associated substrate; PAK, p21-activated kinase; 5-HT, 5-hydroxytryptamine; EGFP, enhanced green fluorescent protein.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  D. D. Tang and Y. Anfinogenova Physiologic Properties and Regulation of the Actin Cytoskeleton in Vascular Smooth Muscle Journal of Cardiovascular Pharmacology and Therapeutics, June 1, 2008; 13(2): 130 - 140. [Abstract] [PDF]
|
|
|
|
 |
|
 D. D. Tang Intermediate filaments in smooth muscle Am J Physiol Cell Physiol, April 1, 2008; 294(4): C869 - C878. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. T. Gerthoffer Migration of Airway Smooth Muscle Cells Proceedings of the ATS, January 1, 2008; 5(1): 97 - 105. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Anfinogenova, R. Wang, Q.-f. Li, A. M. Spinelli, and D. D. Tang Abl Silencing Inhibits CAS-Mediated Process and Constriction in Resistance Arteries Circ. Res., August 17, 2007; 101(4): 420 - 428. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Wang, A. Pandey, and G. W. Hart Dynamic Interplay between O-Linked N-Acetylglucosaminylation and Glycogen Synthase Kinase-3-dependent Phosphorylation Mol. Cell. Proteomics, August 1, 2007; 6(8): 1365 - 1379. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. T. Gerthoffer Mechanisms of Vascular Smooth Muscle Cell Migration Circ. Res., March 16, 2007; 100(5): 607 - 621. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||
