| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 281, Issue 29, 20542-20554, July 21, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of Biological Sciences, University of Pittsburgh, Pittsburgh, Pennsylvania 15260
Received for publication, November 21, 2005 , and in revised form, March 22, 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Shroom (Shrm)2 is an actin-binding protein that localizes to the apical junctional complex (AJC) of epithelial cells and works through myosin II to regulate apical constriction, a process by which columnar cells adopt a wedge-shaped morphology due to a decrease in their apical area (4, 5). This function is essential for neural tube closure in both mouse and Xenopus embryos, as deficiency in Shrm results in exencephaly and spina bifida (5, 6). Shrm appears both necessary and sufficient for the recruitment or maintenance of an actomyosin network in the AJC of neural epithelial cells (4).
In mice and humans, the Shrm family of proteins consists of Shrm, Apxl, and Kiaa1202 (6-8). This family is defined by the arrangement of at least two of three conserved sequence motifs. Shrm, Kiaa1202, and Apxl all contain an N-terminal PDZ domain and a C-terminal motif termed ASD2 (Apx/Shrm domain 2). Apxl and Shrm also contain a centrally located motif termed ASD1 (6). Based on the analysis of Shrm, ASD1 is an actin binding motif that regulates subcellular localization and ASD2 regulates apical constriction. The activity of Shrm is dependent on ASD1 and ASD2 working in cis. To date, the ASD1 and ASD2 elements have only been found in Shrm-related proteins and do not appear in combination with other conserved domains as is observed for many other defined protein modules (such as PDZ, SH2, SH3, and PH domains).
Here we show that Shrm, Apxl, and Kiaa1202 have both shared and unique cellular and biochemical properties. Whereas wild-type Apxl and Kiaa1202 are unable to induce apical constriction, they have the innate capacity to do so, as targeting their ASD2 motifs to the AJC results in constriction. The intrinsic functional differences between Shrm, Apxl, and Kiaa1202 are regulated by variations in subcellular localization that are conferred by association with distinct compartments of the actin cytoskeleton. We further demonstrate that Apxl colocalizes with non-muscle myosin II and may function to regulate endothelial cell morphology during processes involving dynamic actin organization. We also present evidence that the Shrm pathway may be conserved in invertebrate organisms. Thus, we predict that these proteins utilize a common mechanism to regulate diverse cellular processes, and that their activities are temporally and spatially controlled by differential actin binding. Additionally, we suggest that this pathway has been conserved throughout metazoan evolution.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
clones encompassing the entire open reading frame of mouse Apxl were isolated from a mouse brain cDNA library (Stratagene) and assembled into a complete cDNA encoding Apxl (GenBankTM DQ358971
[GenBank]
). Portions of Apxl were amplified by PCR and cloned into the pCS3mt expression vector in-frame with the Myc tags. For bacterial expression of Apxl proteins, portions of the cDNA encoding amino acids 513-880 or 878-1287 were amplified by PCR and cloned into pGex3X. The cDNAs for CG8603 (GenBank NM137108) and human KIAA1202 (GenBank NM020717) were obtained from the BDGP (Bloomington) and the Riken Corporation (Japan), respectively. To generate vectors expressing ShrmS
ASD2 and the various Shrm chimeric proteins, pCS2-ShrmS (6) was digested with EcoRI to remove the portion of the cDNA encoding amino acids 1572-1986 (encompassing ASD2) either religated (for ShrmSASD2) or ligated to DNA fragments encoding amino acids 1068-1480, 1054-1502, and 237-667 of mApxl, hKIAA1202, or CG8603, respectively, that had been amplified by PCR and digested with EcoRI.
Antibody Production and Protein Expression—Antibodies to Shrm (UPT120 and UPT132) have been described previously (4, 5). Apxl-specific sera (UPT115) was generated in rabbits using an antigen consisting of amino acids 878-1287 fused to GST. Specific antibodies were affinity purified using the same antigen coupled to Sepharose. A region of Apxl encoding the localization domain (amino acids 513-880) was expressed in BL21 Escherichia coli and purified using glutathione-Sepharose. Briefly, expression of GST-Apxl was induced for 3 h by the addition of isopropyl 1-thio-
-D-galactopyranoside (0.5 mM final concentration) to log phase cells growing at 30 °C. Cells were harvested by centrifugation, resuspended in NETN (100 mM NaCl, 1 mM EDTA, 20 mM Tris, pH 8.0, 0.05% Nonidet P-40) containing a protease inhibitor mixture, and lysed by sonication. Cleared bacterial lysate was incubated with glutathione-Sepharose (Amersham Biosciences) for 2 h at 4°C. Beads were collected by centrifugation and washed extensively with NETN followed by TBS. GST-Apxl was eluted with free glutathione. Concentration and protein integrity were determined by SDS-PAGE followed by Coomassie Blue staining. His-tagged Shrm was generated by cloning the portion of the mouse Shrm cDNA encoding amino acids 754-1108 into pQE32. Protein expression was induced by the addition of isopropyl 1-thio--D-galactopyranoside to log phase BL21 E. coli to a final concentration of 0.5 mM. Bacteria were grown overnight at 22 °C, pelleted, and lysed by sonication. His-tagged protein was purified under native conditions using nickel-nitrilotriacetic acid-Sepharose as recommended by the manufacturer (Qiagen).
Western Blotting and Actin Binding—Purified His-Shrm or GST-Apxl were added to polymerized F-actin (0.08 µM) and incubated for 1 h at room temperature. For co-sedimentation assays, F-actin was pelleted at 100,000 x g in an Airfuge. The pellets were resuspended in a volume equal to the supernatant fraction, and equal volumes of the pellet and supernatant fractions were resolved by SDS-PAGE and visualized by Coomassie Blue staining. For EM analysis of actin bundles, Shrm-induced actin bundles were spotted onto grids, negatively stained, and observed by transmission electron microscopy. Alternatively, Shrm-actin complexes were pelleted at 14,000 x g in a Microfuge. Pellet fractions were resuspended in sample buffer to a final volume equal to that of the supernatant fractions. Equal volumes of pellet and supernatant fractions were resolved by SDS-PAGE and stained with Coomassie Blue.
Endogenous Apxl was detected by Western blotting using affinity purified UPT115 diluted 1:500 in TBST + 4% milk. Lysate from embryonic day (e) 10.5 mouse embryos was generated by homogenizing embryos in RIPA buffer (+protease inhibitors) with a Dounce homogenizer at 4 °C. Apxl was precipitated from 1 ml (1 mg/ml) of lysate with UPT115 and protein A-Sepharose (Amersham Biosciences) for 1 h at 4°C with constant rocking. Immune complexes were washed with RIPA buffer and TBS and resuspended in Laemmli SDS-PAGE sample buffer. Total lysate or immune complexes were resolved by SDS-PAGE and transferred to nitrocellulose. Proteins were detected by UPT115 diluted 1:500 in TBST. Primary antibody was detected using horseradish peroxidase-conjugated goat anti-rabbit IgG (Amersham Biosciences) diluted 1:2,500 in TBST and ECL (Amersham Biosciences). Ectopically expressed Apxl proteins, Kiaa1202, and CG8603 were detected by Western blotting using the monoclonal antibody 9E10 (anti-Myc) diluted 1:200 in TBST + 4% milk. Shrm chimeric proteins were detected using the Shrm-specific polyclonal antibody UPT120 raised to amino acids 754-1108 of mouse Shrm.
Cell Culture—T23 MDCK and C166 endothelial cells (9) were maintained in Eagle's minimal essential medium supplemented with 10% fetal bovine serum, penicillin/streptomycin, and L-glutamine at 37 °C and 5% CO2. Rat1 fibroblasts and CtBP90 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, penicillin/streptomycin, and L-glutamine at 37 °C and 5% CO2. Cells growing on coverslips or Transwell filters were transfected with 1-2 µg of DNA using Lipofectamine 2000 as per the manufacturer's recommendations (Invitrogen). For drug treatments, cells were treated with either Me2SO (vehicle) or ± blebbistatin (100 µg/ml, CalBiochem) for 90 min. For adhesion assays, cells were removed from plates with trypsin, washed in soybean trypsin inhibitor (1 mg/ml, Sigma), resuspended in serum-free Eagle's minimal essential medium, and plated onto fibronectincoated coverslides (3 µg/cm2, Sigma). In some cases, cells were first transfected with Apxl expression vectors and grown for 18 h prior to plating onto fibronectin.
Immunocytochemistry—Cells growing on coverslips or Transwell filters were washed three times with phosphate-buffered saline (PBS), fixed in either methanol (-20 °C) for 5 min or 4% paraformaldehyde (in PBS) for 15 min at room temperature. Paraformaldehyde-fixed cells were permeablized with 0.2% Triton (in PBS) for 5 min at room temperature. Cells were stained for 1 h at room temperature with the following primary antibodies diluted in PBT (PBS, 0.1% Tween 20): rat anit-ZO-1 (1:400, Chemicon), mouse anti-E-cadherin, mouse anti--catenin, rat anti-PECAM (1:400, BD Pharmingen), rabbit anti-di-phosphorylated myosin regulatory light chain (1:200, Cell Signaling), and mouse anti-non-muscle myosin II (1:400, clone CMII23, Developmental Studies Hybridoma Bank). Cells were washed three times for 5 min each at room temperature. Cells were then stained with secondary antibodies diluted in PBT for 1 h at room temperature. Cells were washed as above and mounted on microscope slides with Vectashield (Vector Labs). Secondary antibodies were Alexa 488, Alexa 568, or Alexa 633 conjugated to goat anti-rabbit, mouse, or rat IgG (Molecular Probes). TRITC-phalloidin was used to detect F-actin (Sigma).
For staining sections, embryos or adult tissues were isolated and fixed in 4% paraformaldehyde for 1-3 h (depending on the size of the sample) at 4 °C. Samples were washed extensively in cold PBS, equilibrated in 30% sucrose overnight at 4 °C, embedded in OCT, and stored at -80 °C. Samples were processed at -20 °C using a Leica cryostat and ribbons of 8-µm sections were placed on glass slides. Sections were equilibrated in PBS, blocked with PBT, 2% goat serum for 30 min, and then incubated overnight at 4 °C in primary antibody diluted in blocking buffer. Sections were washed at room temperature in PBT for 15 min with constant rocking and then incubated for 2 h at room temperature in secondary antibody or TRITC-phalloidin diluted in blocking buffer. Slides were washed and mounted using Vectashield. Images were acquired using a Bio-Rad Radiance 2000 Laser Scanning System mounted on a Nikon E800 microscope (x40 or 60 oil objectives). Collected images were processed using Adobe Photoshop and NIH Image. For cellular measurements, NIH Image was used to measure the area of a cell outlined by E-cadherin staining at the basal surface (a distance of 1.5 µm from the filter) and the AJC (as defined by the distribution of ZO-1).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
200 kDa in both total lysate and immunoprecipitations using lysate derived from e10.5 mouse embryos (Fig. 1A). To determine what cell types express Apxl and where it is distributed within those cells during vertebrate development, frozen sections of e10.5 mouse embryos were stained to detect Apxl. Apxl is predominantly expressed in cells that exhibit polarized growth, including the vasculature and multiple epithelial populations (Fig. 1B). To verify the distribution of Apxl in the vasculature, sections were co-stained to detect Apxl and PECAM, a marker of vascular endothelial cells (Fig. 1C). Apxl and PECAM are expressed in the same cells and exhibit significant co-distribution, suggesting that Apxl may associate with PECAM-based adhesion structures. In most populations of epithelial cells, including gut and neural epithelium (Fig. 1, D and E, respectively), Apxl is localized to the putative AJC or tight junctions. Because the founding member of this protein family, Apx, was identified in Xenopus kidney cells (10), we assessed the expression of Apxl in adult kidney by staining frozen sections. Apxl is expressed in kidney tubules and exhibits a unique distribution in these cells relative to other epithelial cells, such that it is associated with the apical plasma membrane (Fig. 1F). In humans, APXL has been linked to the human disease ocular albinism due to its genomic location on the X-chromosome (7). In mice, Apxl is expressed in the cell junctions of the retinal pigmented epithelium (RPE) in e15.5 mouse embryos (Fig. 1, G-I). Western blot analysis indicates that many adult tissues express Apxl, however, based on immunohistochemistry, a majority of this Western blot signal appears to be contributed by the associated vasculature (not shown).
To more carefully investigate the distribution of Apxl we looked at its subcellular localization in mouse embryo fibroblasts that exhibit epithelial characteristic due to mutations in Ctbp1 and Ctbp2 (CtBP90 cells (11)). In confluent cells, Apxl is localized to the plasma membrane (Fig. 1J). Apxl is localized to the plasma membrane in sparsely plated cells as well, suggesting that its localization is independent of cell-cell adhesion (Fig. 1K). Ectopically expressed Apxl exhibits a subcellular distribution that is similar that of endogenous Apxl. In fibroblasts, Apxl is seen at the plasma membrane with cortical actin (Fig. 1L). In MDCK cells, a population of Apxl is localized to the AJC as defined by ZO-1 distribution (Fig. 1M). The distribution of Apxl in tight junctions in MDCK cells appears somewhat inefficient when compared with its recruitment to these structures in vivo (Fig. 1, D and E). This could be a function of overexpression or result from a mechanism that regulates Apxl localization in vivo that is not present in MDCK cells grown in vitro. The distribution we observe is somewhat inconsistent with that described previously, where Apxl appears to be largely perinuclear with some being localized to the cell membrane and the apical domain (12). This discrepancy might be due to the fact that the antibodies used in that study were directed to a related protein from Xenopus, Apx, which is not the mouse ortholog of Apxl.
Mapping the Functional Sequences of Apxl—We next wanted to determine what region of Apxl mediates its localization in both fibroblast and epithelial cells. To accomplish this, we generated vectors for expressing various portions of Apxl in cells (Fig. 2A). Western blot analysis indicates that the expressed proteins are of the predicted size (Fig. 2B). To assay localization, these proteins were transiently expressed in either Rat1 fibroblasts (Fig. 2) or polarized MDCK epithelial cells (Fig. 3) and their distribution examined by immunofluorescence. In Rat1 cells, proteins containing either the N-terminal PDZ-(1-513) or C-terminal ASD2-(880-1481) motif localize to the cytoplasm, suggesting that these sequences play little role in mediating localization to cortical actin (Fig. 2, D and H, respectively). In contrast, the centrally located ASD1 sequence motif appears to mediate targeting to cortical actin, as all of the proteins that contain it (1-880, 513-880, and 513-1481) localize to cortical actin (Fig. 2, E-G). These results are similar to what has been observed for Shrm, where ASD1 is necessary and sufficient for targeting Shrm to actin stress fibers (6).
|
|
Shrm and Apxl Exhibit Different Actin Binding Properties—It has been shown that Shrm is capable of directly binding F-actin and that this interaction is mediated by a region of Shrm containing ASD1 (6). Because Apxl localizes with actin and ASD1 appears to mediate this distribution, we tested if Apxl localization is regulated by a direct interaction with F-actin. The membrane distribution of Apxl is dependent on an intact actin cytoskeleton, as treatment of cells with cytochalasin D perturbs Apxl localization (Fig. 4, A and B). Following drug treatment, Apxl and actin remain co-localized in discrete puncta, suggesting that Apxl and actin are in a complex (Fig. 4B, inset). To determine whether Apxl directly binds F-actin, the region of Apxl that contains ASD1 and localizes to cortical actin in Rat1 fibroblasts (513-880) was expressed and purified from E. coli and assayed by F-actin co-sedimentation. In these experiments, GST-Apxl-(513-880) is found in the pellet fraction with F-actin (Fig. 4C). These data provide further evidence that ASD1 from both Shrm and Apxl play a similar role in mediating subcellular localization via direct binding to F-actin. Co-sedimentation of Apxl is not due to nonspecific trapping, as an unrelated Apxl protein, Apxl-(878-1287) does not co-sediment with F-actin to any appreciable level.
Unlike Apxl, Shrm localizes with actin stress fibers in vivo and causes them to form bundles when expressed in Rat1 fibroblasts (Fig. 4D). Either full-length Shrm or a portion of Shrm containing ASD1 (amino acids 754-1108) display this activity (Fig. 4E). To assess the direct impact of Shrm on actin organization, we purified the actin-binding domain of Shrm, mixed it with F-actin, and visualized the Shrm-actin complexes by electron microscopy. The addition of Shrm causes actin to form tightly packed bundles, consistent with what is seen in vivo (Fig. 4F). The addition of Shrm to F-actin also allows for sedimentation of F-actin at low speed, consistent with the notion that Shrm bundles actin filaments (not shown). The actin-binding domain of Apxl exhibits no bundling activity in these assays (not shown). Thus, whereas both Shrm and Apxl utilize a similar sequence motif to bind F-actin in vitro, the proteins interact with distinct populations of F-actin in vivo and have differing abilities to organize actin into higher order structures. We predict that the differential nature of actin binding is a critical aspect of regulating the in vivo function of Shrm, Apxl, and other related family members (see below).
|
The identification of CG8603 prompted us to analyze the sequence databases for Ciona intestinalis, Strongylocentrotus purpuratus (sea urchin), and other invertebrates. Results from this analysis suggest that the ASD2 motif may have appeared very early in the animal lineage, as a distantly related partial open reading frame can be identified in Hydra magnipapillata (CX054681 [GenBank] ). This analysis also identified putative orthologs in both Ciona (CiGC24a06, Ref. 14) and sea urchin (XM778480). Unlike CG8603, the predicted Shrm-like proteins from both Ciona and sea urchin possess both the N-terminal PDZ and C-terminal ASD2 motifs (not shown and Fig. 5A). Analysis of the yeast and Arabidopsis genomes indicates there are no Shrm-like proteins in these organisms, suggesting that this protein family is found only in animals. However, not all animals possess Shrm genes, as there are no discernable orthologs for Caenorhabditis elegans. Phylogenetic analysis of the ASD2 sequences indicates that in vertebrates, a duplication of the ancestral Shrm gene gave rise to two homologs, one that became Kiaa1202 and one that duplicated again to give rise to Shrm and Apxl (not shown). This is consistent with the observation that Apxl and Shrm contain the ASD1 motif, whereas Kiaa1202 does not. Ciona and sea urchin are predicted to have only one Shrm-related gene, suggesting that duplication of the ancestral Shrm gene occurred after the divergence of the Urochordates. The predicted Shrm proteins from Ciona and sea urchin do not possess ASD1 sequences, suggesting the homolog that gave rise to Shrm and Apxl acquired this sequence element after the initial duplication.
Alignment of ASD2 sequences from Apx (Xenopus), mShrm, mApxl, KIAA1202, and the invertebrate family members shows that this domain possesses a well conserved series of leucine residues that exhibit spacing consistent with that of a leucine zipper motif (Fig. 5A). This suggests that this domain may serve as a binding interface for homo- or heterotypic interactions or for other cellular proteins and that this domain may reflect an evolutionarily conserved mechanism for regulating cellular architecture. To date, there are no deposited structures in the Protein Data Bank that are significantly similar to the ASD2 motif.
|
ASD2 (amino acid 286-1572) localizes to the AJC but does not cause constriction (Fig. 6B) and a protein containing just ASD2-(1372-1986) is cytoplasmic and also does not cause constriction (Fig. 6C). In contrast to Shrm, expression of Apxl, KIAA1202, or CG8603 in MDCK cells does not result in apical constriction (Fig. 6, D-F). The inability of these proteins to trigger constriction could stem from the fact that they lack the intrinsic ability to cause constriction or that they localize to a different subcellular compartment and thus are not positioned to interface with the pathway. To test these possibilities, we expressed chimeric proteins consisting of amino acids 286-1572 of Shrm (ShrmSASD2) and the ASD2 motifs from Apxl, KIAA1202, or CG8603 (Fig. 5B). Western blot analysis of lysates from transiently transfected MDCK cells indicates that the chimeric proteins are expressed at roughly equivalent levels and are of the correct size (Fig. 5C).
To assess the ability of these chimeric proteins to cause constriction, we transiently expressed these proteins in MDCK cells growing on Transwell filters and stained the cells to detected the exogenous protein and the tight junction protein ZO-1. When expressed in MDCK cells, all of the Shrm chimeric proteins co-localize with ZO-1 and cause apical constriction (Fig. 6, G-I). As a control, a chimeric protein consisting of amino acids 1-1028 of Apxl and 1572-1986 of Shrm was generated and expressed in MDCK cells (Fig. 5, B and C). The Apxl-Shrm chimeric protein is targeted to the AJC as efficiently as either Apxl (Fig. 6, K versus D) or ApxlASD2 (1-1028, Fig. 6J), but is unable to elicit apical constriction (Fig. 6K). These results show that the conserved ASD2 motifs from all of the Shrm-related proteins tested are capable of activating the constriction pathway in MDCK cells and that this activity is regulated at the level of protein distribution.
To quantify the results depicted in Fig. 6, transfected cells were stained to detect E-cadherin, ZO-1, and Shrm (not shown) and the areas of transected cells were measured at the basal and apical surfaces. The results are summarized in Table 1. Untransfected cells have average basal and apical areas of 124 ± 12 and 121 ± 4 µm2, respectively. Cells transfected with ShrmS, Shrm-Apxl, Shrm-Kiaa1202, and Shrm-CG8603 display apical areas of 51 ± 12, 56 ± 6, 66 ± 8, and 54 ± 9 µm2, respectively. The basal areas of these cells are unchanged relative to each other and control cells. Cells expressing either Apxl or Apxl-Shrm show no changes in basal or apical areas relative to control cells. Consistent with what has been published for Shrm (4), apical constriction caused by the various Shrm chimeric proteins is reversed by treatment with the myosin II inhibitor blebbistatin (15), suggesting that these proteins are working through myosin II to regulate cell morphology (Fig. 6L and Table 1).
|
ASD2 localize to stress fibers and bundle them (Fig. 7, A and B), whereas ASD2 is cytoplasmic and has no effect on actin stress fibers (Fig. 7C). Apxl and KIAA1202 both localize to populations of F-actin that are distinct from Shrm, with Apxl targeted to cortical actin and Kiaa1202 localized with a punctate population of actin (Fig, 7, D-F). This distribution is similar to that of the endogenous Kiaa1202 3 and what has been previously reported for ectopically expressed hKIAA1202 (8). Interestingly, CG8603 also localizes to the cortical actin cytoskeleton (Fig. 7F). All of the Shrm chimeras localize to actin stress fibers and bundle them (Fig. 7, G-I). Apxl-(1-1028) and the Apxl-Shrm fusion protein localize to cortical actin (Fig. 7, J and K). These results correlate with the ability of these proteins to elicit constriction, such that only those proteins that bind and bundle stress fibers are capable of causing constriction in epithelial cells.
|
|
-catenin exhibit significant overlap in nascent adherens junctions (Fig. 8A). After 12 h, Apxl and -catenin co-localize in mature cell-cell junctions (Fig. 8B). We next wanted to determine the distribution of endogenous Apxl in sparsely plated, non-polarized endothelial cells (Fig. 8C). In these cells, Apxl is localized to the periphery of the cell in a manner that is largely distinct from focal adhesions. In well spread cells, Apxl is localized in a membrane-proximal position relative to the prominent actin ring at the periphery of the cell. To further investigate the function of Apxl in endothelial cells, we assayed the localization of Apxl, actin, and myosin II during cell spreading (Fig. 8, D-J). As cells start to spread, Apxl and actin co-localize in a prominent ring at the basal surface of the cell (Fig. 8, D-F). This is clearly seen in X-Z projections generated from a series of confocal optical sections (arrowhead). As spreading proceeds, Apxl and actin remain localized to the cell periphery, but Apxl adopts a membrane-proximal position relative to the F-actin ring (Fig. 8, E and F, arrowheads). Because Shrm-like proteins have the capacity to regulate cell shape in a myosin II-dependent manner, we examined the distribution of myosin II and the di-phosphorylated regulatory light chain relative to Apxl and actin in spreading cells. Early during spreading, Apxl, myosin II, and di-phosphorylated regulatory light chain all appear to co-distribute to the basal actin ring (Fig. 8, G-J). These results suggest that Apxl may be involved in the changes in endothelial cell morphology observed during cell spreading.
To determine whether Apxl plays a role in regulating the observed endothelial cell morphology and architecture, cells transiently expressing Apxl or the N-terminal half of Apxl-(1-880) were plated onto fibronectin and assayed by three criteria: cell shape, cell area, and the presence of a peripheral actin ring. Based on results using the actin binding motif of Shrm, we predict that Apxl-(1-880) likely acts as a dominant negative protein (5). Cells expressing the N-terminal half of Apxl show clear deficiency in the ability to organize the peripheral actin ring and adopt a normal morphology, whereas exogenous fulllength Apxl has no impact on morphology (Fig. 8, K-M). The phenotype caused by dominant negative Apxl is lost over time, as by 6 h most transfected cells exhibit a normal morphology (Fig. 8M).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Based on our results, we propose the following model that accounts for the different in vivo functions of Shrm family proteins (Fig. 8N). We predict that all Shrm-like proteins are capable of regulating the activity of myosins via the well conserved ASD2 motif located in their C termini. This ability to regulate myosin is then modulated spatially and temporally by unique actin-binding properties that are specified by N-terminal and centrally positioned sequence elements, namely the PDZ and ASD1 motifs, respectively. Thus, the disparate localization of these proteins can account for their different in vivo activities and allow them to participate in distinct cellular processes.
The ability of Shrm to bind F-actin seems critical for both its localization and its ability to stimulate apical constriction. One question that remains is if the actin bundling activity of Shrm is required for apical constriction. Shrm-induced actin bundling alone is insufficient to cause apical constriction in either cell culture or Xenopus embryos (5) and ASD2 can activate constriction in the absence of actin binding if targeted to the apical plasma membrane (4). However, because the apically targeted ASD2 likely acts on a different population of F-actin, the need to bundle actin may be eliminated. Apxl is unable to cause constriction despite the fact that the protein is targeted to the AJC in MDCK cells and that its ASD2 motif is capable of activating the constriction pathway if fused to the actin binding region of Shrm. These observations lend credence to the notion that the ability of Shrm to cause constriction may be dependent on its ability to alter the organization of F-actin within the AJC. This suggests a model in which Shrm binds and organizes actin in a way that allows the ASD2 region to facilitate the assembly of a contractile actomyosin network in the AJC.
Our results indicate that Apxl is involved in the morphogenesis, maintenance, and/or function of vascular endothelial cells. We are particularly interested in understanding how Apxl might function in vasculogenesis and angiogenesis, the developmental processes by which mesoderm-derived mesenchymal cells differentiate into endothelial cells and become organized into the tubules that make up the vascular system. These processes likely require extensive cytoskeletal remodeling and precisely choreographed alterations in cell morphology and adhesion. The localization of Apxl to nascent and mature cell-cell junctions and the ability of a dominant negative Apxl protein to alter endothelial cell morphology points to a possible role for Apxl in these developmental processes. In addition to embryonic function, Apxl may function in the adult vasculature, as myosin II activity is important for the ability of endothelial cells to regulate permeability (16).
|
KIAA1202 is found on the X-chromosome and is implicated in X-linked mental retardation in humans (8). The cellular basis for this phenotype is unknown, but the association of Kiaa1202 with a distinct population of F-actin in cells and its inherent ability to regulate cell shape suggests that it likely functions to regulate aspects of intracellular architecture or cell morphology. Our analysis of Kiaa1202 suggests that it is capable of regulating the formation and organization of the actin population to which it is localized.3
The discovery of Shrm-like proteins in multiple invertebrate organisms sheds light on the evolution of this protein family and opens new avenues to dissect the molecular function of these proteins. Our results suggest that the ASD2 sequence may represent a relatively old protein motif that is found only in animals and whose activity is evolutionarily conserved. During their embryonic development, invertebrates undergo numerous complex morphological changes that are driven by changes in cell shape and cellular organization. It will be interesting to determine when Shrm-like proteins are expressed, what cell types express them, and where these proteins are localized during invertebrate embryogenesis. This line of experimentation will allow us to determine whether Shrm regulates similar processes in all organisms or if in vertebrates the activity of the ASD2 motif has been subverted to regulate vertebrate-specific processes such as neural tube closure and elaboration of a closed vasculature. Based on our results, we predict that Shrm family proteins will be utilized in multiple morphogenic and developmental processes across animal phyla to regulate cells shape or intracellular architecture in an actin and myosin-dependent manner.
|
|
FOOTNOTES |
|---|
* This work was supported by NIGMS, National Institutes of Health, Grant RO1GM067525 (to J. D. H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 
The on-line version of this article (available at http://www.jbc.org) contains supplemental Fig. S1.
1 To whom correspondence should be addressed: 4249 Fifth Ave., Pittsburgh, PA 15260. Tel.: 412-624-6987; Fax: 412-624-4759; E-mail: jeffh{at}pitt.edu.
2 The abbreviations used are: Shrm, Shroom; AJC, apical junctional complex; SH, Src homology; PH, pleckstrin homology; GST, glutathione S-transferase; TBS, Tris-buffered saline; MDCK, Madin-Darby canine kidney; RPE, retinal pigmented epithelium; PBS, phosphate-buffered saline; TRITC, tetramethylrhodamine isothiocyanate; ASD1, Apx/Shrm domain 1; e, embryonic day; PECAM, platelet endothelial cell adhesion molecule.
3 M. Yoder and J. D. Hildebrand, unpublished data.
4 J. Wallingford, University of Texas, Austin, personal communication.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Sevilla-Perez, M. Konigshoff, G. Kwapiszewska, O. V. Amarie, W. Seeger, N. Weissmann, R. T. Schermuly, R. E. Morty, and O. Eickelberg Shroom expression is attenuated in pulmonary arterial hypertension Eur. Respir. J., October 1, 2008; 32(4): 871 - 880. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Nishimura and M. Takeichi Shroom3-mediated recruitment of Rho kinases to the apical cell junctions regulates epithelial and neuroepithelial planar remodeling Development, April 15, 2008; 135(8): 1493 - 1502. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Etournay, I. Zwaenepoel, I. Perfettini, P. Legrain, C. Petit, and A. El-Amraoui Shroom2, a myosin-VIIa- and actin-binding protein, directly interacts with ZO-1 at tight junctions J. Cell Sci., August 15, 2007; 120(16): 2838 - 2850. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
|
Deceased.