QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Vol. 19, Issue 5, 1862-1872, May 2008

This Article Abstract Full Text (PDF) Supplemental Materials All Versions of this Article: E07-09-0869v1 19/5/1862    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Severson, E. A. Articles by Parkos, C. A. PubMed PubMed Citation Articles by Severson, E. A. Articles by Parkos, C. A.

Cis-Dimerization Mediates Function of Junctional Adhesion Molecule A

Epithelial Pathobiology Research Unit, Department of Pathology, Emory University, Atlanta, GA 30322

Submitted September 6, 2007; Revised January 15, 2008; Accepted February 6, 2008
Monitoring Editor: M. Bishr Omary

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Junctional adhesion molecule-A (JAM-A) is a transmembrane component of tight junctions that has been proposed to play a role in regulating epithelial cell adhesion and migration, yet mechanistic structure–function studies are lacking. Although biochemical and structural studies indicate that JAM-A forms cis-homodimers, the functional significance of dimerization is unclear. Here, we report the effects of cis-dimerization–defective JAM-A mutants on epithelial cell migration and adhesion. Overexpression of dimerization-defective JAM-A mutants in 293T cells inhibited cell spreading and migration across permeable filters. Similar inhibition was observed with using dimerization-blocking antibodies. Analyses of cells expressing the JAM-A dimerization-defective mutant proteins revealed diminished β1 integrin protein but not mRNA levels. Further analyses of β1 protein localization and expression after disruption of JAM-A dimerization suggested that internalization of β1 integrin precedes degradation. A functional link between JAM-A and β1 integrin was confirmed by restoration of cell migration to control levels after overexpression of β1 integrin in JAM-A dimerization-defective cells. Last, we show that the functional effects of JAM dimerization require its carboxy-terminal postsynaptic density 95/disc-large/zonula occludins-1 binding motif. These results suggest that dimerization of JAM-A regulates cell migration and adhesion through indirect mechanisms involving posttranscriptional control of β1 integrin levels.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Junctional adhesion molecule-A (JAM-A) is a transmembrane component of tight junctions in epithelial and endothelial cells. In addition, JAM-A is expressed on the surface of blood cells, including leukocytes and platelets (Gupta et al., 2000). JAM-A has been implicated in a diverse array of functions, including intercellular junction assembly (Liang et al., 2000; Liu et al., 2000), cell adhesion (Mandell et al., 2005), leukocyte transmigration (Martin-Padura et al., 1998; Del Maschio et al., 1999; Ostermann et al., 2002; Corada et al., 2005; Khandoga et al., 2005; Ostermann et al., 2005), platelet activation (Kornecki et al., 1990; Naik et al., 1995; Sobocka et al., 2000; Babinska et al., 2002), and angiogenesis (Naik et al., 2003a,b). Additionally, JAM-A has been shown to be a receptor for reovirus (Barton et al., 2001; Forrest et al., 2003; Prota et al., 2003; Campbell et al., 2005). Structurally, JAM-A consists of an extracellular domain with two immunoglobulin (Ig)-like loops, a membrane-spanning segment, and a cytoplasmic tail containing a C-terminal postsynaptic density 95/disc-large/zonula occludins-1 (PDZ) binding motif. The cytoplasmic tail of JAM-A has been reported to associate, either directly or indirectly, with PDZ domain-containing proteins, such as zona occludens-1 (Bazzoni et al., 2000; Ebnet et al., 2000), AF-6/Afadin (Ebnet et al., 2000) and Par3/ASIP (Ebnet et al., 2001; Itoh et al., 2001), through characteristic hydrophobic residues (FLV) at the carboxy terminus. Evidence suggests that the cytoplasmic tail plays an important role in directing JAM-A localization to intercellular contacts (Ozaki et al., 2000), formation of tight junctions (Rehder et al., 2006), and transduction of intracellular signaling events (Bazzoni et al., 2005; Mandell et al., 2005; Naik and Naik, 2006). The extracellular domain of JAM-A can form homodimers through its N-terminal Ig loop (Prota et al., 2003). Furthermore, the human JAM-A crystal structure predicts dimers forming between molecules on the same cell (in cis) (Prota et al., 2003); however, the murine protein crystal structure predicts tetramer formation between the extracellular loops between cells (in trans) (Kostrewa et al., 2001). Despite these intriguing observations, mechanistic studies linking dimerization of JAM-A to these functions are lacking.

JAM-A is abundantly expressed in polarized epithelia, yet its role in epithelial cell migration has not been studied. In endothelial cells, controversy exists concerning the functional role of JAM-A in the regulation of cell migration. In a study by Bazzoni et al. (2005) the absence of JAM-A in endothelial cells enhanced spontaneous and random cell motility by reducing the stability of microtubules and impairing the formation of focal adhesions (Bazzoni et al., 2005). Transfection of full-length JAM-A, but not a C-terminal PDZ binding motif deleted JAM-A mutant restored random cell motility. Recently, JAM-A was shown to interact with integrin vβ3 and to enhance endothelial cell migration on vitronectin when overexpressed, as well as enhance phosphorylation of focal adhesion kinase and mitogen-activated protein kinase (MAPK) (Naik and Naik, 2006). None of the above-mentioned studies examined the role of JAM-A dimerization in mediating these effects on migration.

We recently reported that transient knockdown of JAM-A expression in epithelial cells resulted in decreased protein levels of β1 integrin that correlated with altered cell shape and decreased cell adhesion (Mandell et al., 2005). This study suggested that JAM-A may regulate cell adhesion by increasing integrin protein expression; however, the mechanisms for these JAM-A–mediated effects was not investigated and is the topic of this report.

Based upon these observations, we hypothesized that cis-dimerization of JAM-A plays a key role in regulation of cell migration. To test this hypothesis, we stably overexpressed wild-type and JAM-A with mutations in the putative dimerization domain in 293T cells, a human epithelial cell line that expresses low levels of JAM-A. Dimerization of the extracellular domain is mediated by the predicted formation of salt bridges in the membrane distal Ig loop D1. One dimerization-defective JAM-A mutant we studied has point mutations at two residues (E61A/K63A) predicted by the crystal structure to be required for dimerization (6163), and both mutations have been shown to disrupt dimerization in vitro (Mandell et al., 2004). Mutation of either residue has been shown to result in JAM-A formation of only monomers as assessed by gel filtration (Guglielmi et al., 2007). A second dimerization-defective construct that we tested consists of JAM-A with a deletion of the distal most immunoglobulin-like loop, which is necessary for dimerization (DL1). We observed that overexpression of both the 6163 and DL1 dimerization-defective mutants resulted in decreased 293T cell migration and spreading. We also determined that these cellular effects are mediated by decreased β1 integrin protein levels. Notably, for the dimerization-defective constructs to have an effect, we determined that there must be a carboxy-terminal PDZ binding motif on JAM-A, suggesting that dimerization-defective mutants mediate their effects through sequestration of scaffolding proteins. These observations indicate that JAM-A dimerization indirectly regulates cell migration through signaling events that ultimately increase β1 integrin protein levels resulting in increased cell adhesion, spreading, and migration.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Cell Culture
293T human embryonic kidney epithelial cells were grown in DMEM supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 100 IU of penicillin, 100 µg/ml streptomycin, 15 mM HEPES, and 1% nonessential amino acids (Cellgro, Mediatech, Herndon, VA). The cells were subcultured and harvested with 0.05% trypsin with EDTA in Hank's balanced salt solution (HBSS–) (Sigma-Aldrich, St. Louis, MO). SK-C0–15 cells, a transformed human colonic epithelial cell line (Lisanti et al., 1989; Mandell et al., 2005; Ivanov et al., 2006) were cultured as described previously in DMEM supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 15 mM HEPES, 1% nonessential amino acids, 40 µg/ml penicillin, and 100 µg/ml streptomycin, pH 7.4.

JAM-A Mutant Plasmid Production
Production of the plasmids encoding full-length JAM-A, the JAM-A truncation mutant lacking the N-terminal Ig-like loop (DL1), and a JAM-A point mutant containing substitutions at residues 61 and 63 (E61R/K63E, 6163) have been described previously (Mandell et al., 2004). Briefly, the coding region along with 1.5 kb of the 3' untranslated region was restriction enzyme digested and inserted into a pIRES-EGFP vector. Plasmids of all four constructs for transient transfections were amplified by polymerase chain reaction (PCR) (5'-ATATGGTACCAGCCACCATGGGGAC AAA-3'; 5'-ATATCTCGAGTCACACCAG GAATGACGAGGTCTG-3') and digested with KpnI and XhoI before ligation into pCDNA3.0. A construct lacking both the DL1 and last three amino acids (FLV) was made from the DL1 mutant by using PCR (5'-ATATGGTACCAGCCACCATGGGGACAAA-3'; 5'-ATATCTCGAGTCATGACGAGGT CTGTTTGAA-3'). PCR product was digested and ligated into pCDNA3.0 as described above.

Stable Cell Lines
293T cells were transfected with constructs and empty vector (pIRES2-EGFP) by using Lipofectamine 2000 (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. The transfected cells were enriched using flow-cytometric-based cell sorting by gating on enhanced green fluorescent protein (EGFP) fluorescence. Single clones were grown under Geneticin (G418; Invitrogen) selection, and expression of EGFP in colonies was verified. Cell lines were verified for expression of JAM-A constructs by Western blot and immunofluorescence. Early passages were frozen in FBS and 10% dimethyl sulfoxide. No cell lines were used for >10 passages from transfection in these studies, and expression of EGFP was monitored before each experiment to ensure that cells used remained stably transfected.

Transient Cell Transfections
For plasmid transfections, the 293T cells were transfected using Lipofectamine 2000 (Invitrogen) in Opti-MEM I (Invitrogen) according to the manufacturer's protocol. For β1 integrin overexpression, constructs containing full-lengthof β1 integrin cDNA in pCMV-XL6OriGene) or the empty pCMV-XL6 vector were used at a concentration of 1.0 µg of plasmid per ml, and assays were performed 48 h after transfection. SmartPool siRNA targeted to β1 integrin was obtained from Dharmacon RNA Technologies (Lafayette, CO). For transient overexpression of JAM-A or JAM-A mutants, pCDNA3.0 with the JAM-A protein coding sequence or empty vector (pCDNA3.0) were used in the same manner as the β1 integrin construct. siRNA transfections were performed in Opti-MEM I (Invitrogen) with HiPerFect (QIAGEN, Valencia, CA) according to the manufacturer's protocol using either a SmartPool for β1 integrin or siRNA for cyclophilin B (a control gene), and the final concentration of siRNA was 50 nM. Three JAM-A siRNA sequences were used at a total concentration of 50 nM: 5'-AGGGTCACATGCCAATAAA-3', 5'-CAGTCTATTTATTAACTTA-3', and 5'-TCCCTTCTAAGTAGACAGC-3'. Experimental time points for DNA and siRNA transfections were 48 and 72 h, respectively.

Antibodies
The murine monoclonal anti-JAM-A antibodies J10.4 and 1H2A9 were described previously (Liu et al., 2000; Mandell et al., 2004). All other antibodies were obtained commercially: murine anti-β1 integrin (BD Biosciences PharMingen, San Diego, CA), rat anti-β1 integrin (Mab13) (BD Biosciences PharMingen), rabbit anti-β4 integrin (Santa Cruz Biotechnology), rabbit anti-phospho-paxillin (Tyr118) (Cell Signaling Technology, Danvers, MA), rabbit anti-Rap1 (Upstate Biotechnology, Charlottesville, VA), and murine anti-tubulin (Sigma-Aldrich). Horseradish peroxidase-conjugated secondary antibodies were obtained from Jackson ImmunoResearch Laboratories (West Grove, PA), and fluorescently labeled secondary antibodies were obtained from Invitrogen.

Cell Migration Assays
Cells were washed in Hank's balanced salt solution without calcium (HBSS–), and then they were incubated in cell dissociation buffer (enzyme free, phosphate-buffered saline [PBS]-based) (Invitrogen) for 30 min to disrupt cell junctions. Cells were then washed in HBSS–, centrifuged, and resuspended in serum-free DMEM. Cells (1 x 10⁵) were added on the top side of 8-µm pore size Transwell (Corning Life Sciences, Acton, MA) inserts that had been coated with 10 µg/ml fibronectin overnight on the underside of the transwell (Figure 2A). Cells were allowed to migrate across inserts toward the fibronectin-coated side for 3 h at 37°C. Inserts were then washed, fixed with ethanol, and stained with phallodin. Confocal fluorescence microscopy was performed on the lower chamber side of inserts using a laser scanning confocal microscope (Axioplan2 microscope equipped with LSM510 Meta; Carl Zeiss, Thornwood, NY). Cells were counted from two randomly chosen fields of view from three separate inserts, and average counts with SEM were used to quantify the extent of migration. For the study on the effects of J10.4 and 1H2A9 antibodies on cell migration, cells were treated with 10 µg/ml J10.4 or 1H2A9 at 4°C for 1 h before the addition of cells into Transwell inserts.

Cell Spreading
Cells (5 x 10⁴) were added on coverslips coated with 10 µg/ml fibronectin in 24-well plates, and then they were incubated for 1 h at 37°C. Phase-contrast microscopy was used to capture images at 5x magnification for cell spreading assays. Two separate fields from each of three coverslips were averaged to assess the extent of cell spreading. Rounded cells with phase sharp edges and no protrusions were defined as rounded, and cells with protrusions and flattened morphology were defined as spreading. To study cell protrusion changes, 2.5 x 10⁴ cells were incubated on 10 µg/ml fibronectin coverslips at 37°C for 48 h, and then they were washed three times in HBSS+. Cells were then fixed in ethanol, blocked in 1% bovine serum albumin (BSA) in HBSS+, and stained with Alexa-488-phalloidin (1:1000) or Topro-3 (1:1000). Confocal fluorescence images were captured using a Zeiss laser-scanning microscope. The length of all cellular protrusions 30 high powered fields from three separate experiments were measured using the MetaMorph imaging program (Carl Zeiss) and reported as average length with SE of the mean.

Immunoblotting
Cells were homogenized in lysis buffer containing 20 mM Tris, 50 mM NaCl, 2 mM EDTA, 2 mM EGTA, 1% sodium deoxycholate, 1% Triton X-100, and 0.1% SDS, pH 7.4. Lysis buffer was supplemented with protease and phosphatase inhibitor cocktails containing 4-(2-aminoethyl)-benzenesulfonyl fluoride hydrochloride, pepstatin A, E-64, bestatin, leupeptin, aprotinin, microcystin LR, cantharidin, (–)-p-bromotetramisole, sodium vanadate, sodium molybdate, sodium tartrate, and imidazole (1:100 dilution; Sigma-Aldrich). Protein concentrations in lysates were quantified by bicinchoninic acid assay. Lysates were cleared by centrifugation and immediately boiled in SDS sample buffer. SDS-polyacrylamide gel electrophoresis (PAGE) and immunoblots were performed by standard methods. Immunoblots were probed for tubulin to ensure equal protein loading.

Immunofluorescence Microscopy
Cells were grown on transwell filters or glass coverslips, fixed in 3.7% paraformaldehyde for 20 min at 25°C, permeabilized with 0.2% Triton X-100 for 10 min at 25°C, and blocked in 1% BSA in HBSS+ for 1 h. Primary antibodies were diluted in blocking buffer and incubated with cells for 1 h at 25°C. The cells were washed in HBSS+, and then they were incubated in fluorescently labeled secondary antibodies or Alexa fluorophore-conjugated phallodin for 45 min at room temperature. Labeled cells were then washed and mounted in Prolong Antifade agent (Invitrogen). Confocal fluorescence images were captured using a laser-scanning microscope (Carl Zeiss).

Chemical Inhibitors
MG-132 (Calbiochem, San Diego, CA) was used at 10 µM final concentration, whereas MG-262 (BIOMOL Research Laboratories, Plymouth Meeting, PA) was used at 20 µM final concentration. Cyclohexamide (MP Biomedicals, Irvine, CA) was used at a final concentration of 50 µg/ml.

Real-Time Reverse Transcription-Polymerase Chain Reaction (RT-PCR)
Total RNA was extracted from cells using RNeasy Mini kit (QIAGEN) according to the manufacturer's protocol. RT-PCR was performed by using iScript One-Step RT-PCR kit with SYBR Green (Bio-Rad, Hercules, CA) on iCycle iQ real-time PCR detection system (Bio-Rad) according to the manufacturer's instructions. A pair of PCR primers (5'-ATCCCAGAGGCT CCAAAGAT-3' and 5'-CTAAATGGGCTGGTGCAG-3') was used to amplify β1 integrin. Primer pair (5'-CGGCTACCACATCCAAGGAA-3' and 5'-GCTGGAATTACCGCGGCT-3') targeting 18S RNA was used as internal control.

Rap1 Activity Assay
Active Rap1 was detected using a pull-down procedure (Franke et al., 1997). Cells were lysed at 4°C in lysis buffer (50 mM Tris-HCl, pH 7.4, 0.5M NaCl, 1% Nonidet P-40, 2.5 mM MgCl₂, and 10% glycerol) supplemented with protease inhibitor cocktails (1:100; Sigma-Aldrich). The lysates were clarified by centrifugation, and then 200 µg of protein from each lysate was incubated with 30 µl of agarose beads conjugated with Ral GDS-Rap-binding domain (Upstate Biotechnology) for 45 min at 4°C. The beads were washed four times in lysis buffer, resuspended in 2x reducing SDS sample buffer, and boiled for 15 min. The entire sample was then loaded into each well for separation by SDS-PAGE, and active Rap1 detected by immunoblot.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Expression of Dimerization-defective JAM-A Mutants Inhibits Migration of 293T Cells
To investigate the role of JAM-A cis-dimerization in cell migration, we used 293T human embryonic kidney epithelial cells that normally express low levels of JAM-A protein. We generated 293T stable cell lines expressing two different JAM-A mutants that either lack the distal extracellular Ig loop (DL1) or contain mutations in a characteristic dimerization motif within DL1 (6163). The 6163 mutant contains two point mutations at residues 61 and 63 that are involved in the formation of JAM-A homophilic dimer salt bridges (Figure 1A) (Prota et al., 2003; Mandell et al., 2004). Both mutants have been reported to be unable to form extracellular homophilic dimers as reported by gel filtration (Gugliemi et al., 2007). Furthermore, in Western blots of JAM-A from cells expressing the 6163 mutant that were treated with a cell impermeable cross-linker, we observed no JAM-A dimerization in contrast to results from cells expressing only wild-type protein (Supplemental Figure 1). We also generated two control cell lines, one line that overexpresses wild-type JAM-A and another line that contains the appropriate empty vector control (pIRES2-GFP). As shown in Figure 1, B and C, 293T cells normally express endogenous JAM-A, which localizes to cell–cell contacts. Expression levels of vector control, wild-type–overexpressing, and mutant JAM-A 293T cell lines were analyzed by Western blotting, and they are shown in Figure 1B1 after 10 s of exposure by enhanced chemiluminescence (ECL). Figure 1B2 highlights, by longer ECL exposure (2 min), endogenous JAM-A expression compared with that in cells stably transfected with JAM-A. Furthermore, immunofluorescence analyses indicated that in overexpressing cells, exogenous JAM-A localized to cell–cell contacts in a manner similar to endogenous, control JAM-A (Figure 1C).

View larger version (70K): [in this window] [in a new window]   Figure 1. Stable expression of JAM-A mutants in 293T cells. (A) Structure for endogenous JAM-A is shown and contains two Ig-like loops, a transmembrane domain, and a cytoplasmic tail with a carboxy-terminal PDZ binding domain. The star in the 6163 mutant highlights the region of mutations at amino acids 61 and 63 that forms a salt-bridge between two JAM-A molecules in cis. The DL1 mutant completely lacks the distal most Ig-like domain that mediates cis-dimerization. (B) Western blots demonstrating overexpression of wild-type and mutant proteins in 293T stable transfectants. Tubulin is shown as a protein loading control. A 10-s film exposure (B1) and a 2-min exposure (B2) are shown to demonstrate the presence of JAM-A overexpression in the stable cell lines and to highlight the presence of endogenous JAM-A, respectively. (C) Immunofluorescence labeling of 293T cells expressing mutant constructs for JAM-A protein demonstrating similar localization of endogenous JAM-A, exogenous JAM-A, and mutant JAM-A to cell–cell contacts.

View larger version (44K): [in this window] [in a new window]   Figure 2. JAM-A dimerization-defective mutant effects cell migration. (A) Schematic for cell migration assays. As detailed in the methods, 293T cells suspended in DMEM without serum were added to the upper chamber of transwell filters coated with 10 µg/ml fibronectin on the bottom of the filters. To assess the extent of cell migration, cells were stained with phallodin and confocal images were taken of the underside of the transwell filter. (B) Cell migration assays revealed that overexpression of DL1 and 6163, but not wild-type JAM-A resulted in decreased cell migration. Bar, 200 µm. (C) The number of cells that migrated per square millimeter over 3 h was determined from three separate filters after assessing two photomicrographs of each filter and counting the number of cells by using MetaMorph software. Average counts ± SEM are shown. As can be seen, overexpression of 6163 and DL1 mutants significantly decreased cell migration (*p < 0.05).

View larger version (62K): [in this window] [in a new window]   Figure 3. JAM-A mutants alter 293T cell spreading. (A) Cell spreading assays reveal that dimerization-defective mutants 6163 and DL1 decrease cell spreading. Serum-starved 293T cells were added to 10 µg/ml fibronectin-coated coverslips. After 1 h at 37°C, phase contrast images were taken (2 images each for 3 coverslips/sample). Black arrows point to spreading cells, and white arrows point to rounded cells. (B) Spreading versus rounded cells were counted manually for each image, and average ± SEM was plotted. As can be seen, expression of dimerization-defective JAM-A significantly decreases 293T cell spreading (*p < 0.05).

View larger version (85K): [in this window] [in a new window]   Figure 4. JAM-A dimerization defective-mutants decrease the length of 293T cell protrusions. (A) Confocal microscopy revealed a decrease in both the length and number of cellular protrusions (white arrows) in 293T cells overexpressing the 6163 and DL1 mutants. Bar, 50 µM. (B) All cell protrusions in eight images were measured using the program ImageJ. Bars are the average length in micrometers + SEM. As can be seen, cellular protrusions in JAM-A dimerization-defective cells are significantly shorter than those in cells expressing wild-type JAM-A (*p < 0.05).

View larger version (49K): [in this window] [in a new window]   Figure 5. The number of PY118-paxillin focal concentrations is decreased by overexpression of 6163 and DL1 mutants in 293T cells. (A) Staining with phallodin, anti-PY118 paxillin, and Topro revealed that overexpression of both the 6163 and DL1 mutants significantly decreased the number and density of focal contacts as determined by PY-118 paxillin staining. Bar, 20 µm. (B) PY118-paxillin containing rod-shaped structures were counted with MetaMorph software based on staining with anti-PY118 paxillin. Five slides were counted per sample, and average + SEM/mm2 was reported (*p < 0.05).

View larger version (77K): [in this window] [in a new window]   Figure 6. The β1 integrin protein expression, but not mRNA levels is decreased in the 293T JAM-A dimerization-defective cell lines. (A) Immunoblotting revealed that 293T cells expressing 6163 or DL1 had dramatically decreased amounts of β1 integrin protein levels, with no change in β4 integrin or tubulin, which was used as a loading control. (B) Real-time PCR analysis from control and transfected cell lines demonstrating no significant change in β1 Integrin mRNA for any of the overexpressing 293T cell lines.

Active Rap1 Is Decreased in JAM-A Dimerization-defective Cell Lines
Recently, we demonstrated a decrease in the active form of the small GTPase Rap1 after down-regulation of JAM-A expression in SK-CO15 cells (Mandell et al., 2005). Furthermore, we demonstrated that down-regulation of Rap1 by small interfering RNA (siRNA) resulted in decreased β1 integrin levels. Thus, we examined whether Rap1 levels were altered in the JAM-A dimerization-deficient cell lines. Using standard pull-down assays to isolate active, or guanosine triphosphate-bound Rap1, we analyzed the cell lines for activation status of Rap1. As shown in Supplemental Figure 2, the 6163 and DL1 mutant cell lines each had decreased active Rap1 levels compared with control cell lines. In concert with our previously reported findings, this observation suggests that active Rap1 may be a signaling link between JAM-A dimerization and β1 integrin protein levels.

Inhibition of JAM-A Dimerization with Antibody Reduces Cell Migration and β1 Integrin Protein Levels
Given that expression of dimerization-defective JAM-A mutants in 293T cells inhibited cell migration and decreased β1 integrin protein level, we tested antibodies known to inhibit dimerization for effects on cell migration and β1 integrins. We were particularly interested in the time course of effects of impaired dimerization of JAM-A, because the stable mutant cell lines represent long-term (chronic) effects. We have previously shown that the JAM-A monoclonal antibody (mAb) J10.4 inhibits the formation of JAM-A dimers, whereas the JAM-A mAb 1H2A9 binds to the D1 Ig loop, but it does not inhibit dimerization at similar concentrations (Mandell et al., 2004). 293T cells with endogenous levels of JAM-A were treated with 10 µg/ml J10.4 or 1H2A9 before the migration assays. As shown in Figure 7, A and B, J10.4, but not 1H2A9, significantly inhibited 293T cell migration toward fibronectin compared with a murine IgG control. Cells treated with murine IgG or 1H2A9 migrated at a rate of 590 cells per mm² per 3 h, whereas migration was decreased by treatment with J10.4 to a rate of 310 cells per mm² over 3 h, representing a 47% decrease. These results suggest that acute disruption of dimerization of JAM-A inhibits migration in 293T cells.

View larger version (37K): [in this window] [in a new window]   Figure 7. Antibodies blocking JAM-A dimerization inhibits 293T cell migration and alters β1 integrin localization. (A) JAM-A antibody J10.4 (inhibits JAM-A dimerization) inhibits 293T cell migration. Serum-starved 293T cells were treated with 10 µg/ml mouse IgG, 1H2A9 (does not inhibit JAM-A dimerization), or J10.4 before migration assays. Confocal analyses revealed that J10.4 significantly inhibited 293T cell migration. Bar, 200 µm. (B) Blocking of JAM-A dimerization (J10.4) significantly reduced 293T cell migration compared with isotype (mouse IgG) and antigen-binding (1H2A9) antibody controls (*p < 0.05). (C and D) 293T cells were Ca++ depleted for 5 min to expose intercellular junctions followed by treatment with J10.4 (JAM-A dimerization blocking antibody), 1H2A9 (nondimerization blocking JAM-A antibody), or mouse IgG at 10 µg/ml for 3 h in DMEM. Western blots of cell lysates were then probed for total levels of β1 integrin (C), demonstrating no change in levels of β1 integrin. However, immunofluorescence photomicrographs of such cells stained for β1 integrin by using a rat anti-β1 integrin antibody (D) reveal loss of lateral cell border staining after treatment with J10.4 but not 1H2A9 or mouse IgG. Note the appearance of β1 integrin labeled cytoplasmic vesicles (white arrow) in J10.4-treated cells, suggesting internalization. (E) Concurrent treatment with J10.4 and Cyclohexamide results in decreased B1 integrin.

Reduced Cell Migration in JAM-A Dimerization-defective Cell Lines Is Secondary to Decreased β1 Integrin Protein Levels
We reasoned that decreased levels of β1 integrin observed in the dimerization-defective JAM-A–expressing cell lines was linked to the observed JAM-A–mediated decreases in cell migration. We thus down-regulated β1 integrin expression by using siRNA in 293T cells to determine whether similar inhibitory effects would occur as observed with the JAM-A dimerization-defective cell lines. Compared with control siRNA specific for cyclophilin B, β1 integrin expression was virtually ablated in 293T cells as assessed by Western blot (Figure 8A). Importantly, reduction of β1 integrin expression had no effect on JAM-A protein levels (Figure 8A). When such cells were tested in cell migration assays, there was significant inhibition of cell migration compared with siRNA-treated controls in a manner not significantly different from that observed with JAM-A mutants. Cells with down-regulated β1 integrin protein levels migrated at a rate of 190 cells per mm² over the course of 3 h, whereas cells with control siRNA treatment migrated at a rate of 470 cells per mm² (Figure 8C).

View larger version (33K): [in this window] [in a new window]   Figure 8. β1 Integrin is the putative effector for JAM-A–mediated regulation of 293T cell migration. (A) Western blot demonstrating that transfection of 293T cells with siRNA specific for β1 integrin results in significantly decreased β1 integrin protein levels compared with transfection with siRNA targeting cyclophilin (CB). JAM-A protein levels were not changed. (B and C) Transfection of 293T cells with siRNA specific for β1 integrin revealed significantly reduced 293T cell migration compared with controls treated with siRNA specific for cyclophilin B. Bar, 200 µM. Bar graph is average number of cells per field + SEM (*p < 0.05). (D) Western blot demonstrating increased β1 integrin protein levels in JAM-A dimerization-defective cell lines transfected with a plasmid encoding β1 integrin. JAM-A protein levels were not changed. (E and F) Migration assays with mutant cell lines overexpressing β1 Integrin. As can be seen, overexpression of β1 integrin in the 6163 and DL1 mutant cell lines restores cell migration to that of control 293T cells. Transient transfection of a control plasmid had no effect on cell migration. Bar, 200 µm (*p < 0.05 for β1 integrin-transfected cells vs. control).

The PDZ Domain of JAM-A Is Necessary for the Dominant-Negative Effects of Dimerization-defective JAM-A Constructs
To clarify mechanisms of the dominant-negative effects on cell migration mediated by the dimerization-defective JAM-A mutants, we rationalized that these mutants are likely to sequester PDZ–containing JAM-A binding/signaling proteins away from native JAM-A dimers. To test this hypothesis, we created a modified DL1 mutant construct that lacks the C-terminal PDZ binding domain termed DL1-dFLV. Furthermore, we generated a JAM-A mutant lacking only the PDZ binding motif. Thus, if dimerization-defective JAM-A affects cell migration by sequestering PDZ domain-containing scaffolds, then DL1-dFLV should reverse the dominant-negative effect of DL1 on cell migration and the dFLV-only mutant should mimic the effects of dimerization-defective JAM-A by dimerizing with wild-type JAM-A and preventing endogenous JAM-A homodimerization.

The JAM-A mutants were transiently expressed in 293T cells (Figure 9A) with a transfection efficiency of 70–90%. Transfected cells were then used in cell migration assays on permeable filters as described above. In the passage of 293T cells used for this set of experiments, there was a higher rate of migration than observed in the passage of 293T cells clonally selected for the stable cell lines. In these transient transfections, we observed decreased β1 integrin levels with 6163 and DL1, indicating that degradation of β1 integrin due to interference with JAM-A dimerization occurs in a relatively short time frame of <48 h. In contrast, the DL1-dFLV double mutation had no effect on the levels of β1 integrin protein. Furthermore, as shown in Figures 9, B and C, deletion of the C-terminal PDZ binding motif in the DL1 mutant (DL1-dFLV) completely reversed the dominant-negative effects on migration observed with DL1, and it restored migration to control levels. As predicted, transfection with a JAM-A mutant lacking only the PDZ motif also decreased the levels of β1 integrin protein and cell migration. Overall, these data strongly suggest that dimerization-defective JAM-A mutants accelerated β1 integrin degradation and decreased cell migration by sequestering PDZ-containing scaffolding protein from native JAM-A dimers at the plasma membrane.

View larger version (45K): [in this window] [in a new window]   Figure 9. The PDZ binding domain of JAM-A is critical for dominant-negative effects on cell migration and β1 integrin expression by dimerization-defective mutants. (A) Expression of JAM-A mutant proteins and β1 integrin after transient transfection. As can be seen in the Western blot, full-length (wild type), and the 6163 mutant of JAM-A have an Mr of 37 kDa, whereas constructs lacking the membrane-distal loop (DL1 and DL1-dFLV) have an Mr of 25 kDa. Also shown are immunoblots for β1 integrin after transient transfection with JAM-A constructs, demonstrating decreased expression with 6163, DL1, and dFLV, respectively. However, transfection with the DL1-dFLV double mutant results in no change in β1 integrin protein expression. (B) Transient expression of DL1 or dFLV resulted in decreased cell migration, as measured by confocal analysis of Topro-3 nuclear staining, whereas expression of the DL1-dFLV construct had no effect. These findings suggest that the PDZ binding motif is required for regulation of β1 integrin protein levels and cell migration by the DL1 mutant. (C) Average cells migrated ± SEM are shown. As can be seen, overexpression of 6163 and DL1 mutants significantly decreased cell migration, whereas expression of the double mutant DL1-dFLV had no effect (*p < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Although there are no studies on the role of JAM-A migration in epithelial cells, there are a few reports examining JAM-A and migration in endothelial cells. Bazzoni et al. (2005) reported decreased two-dimensional cell migration in JAM-A–deficient endothelial cells, by using scratch wound assays, that was restored after transfection with full-length JAM-A but not protein lacking the PDZ binding motif. Furthermore, Naik et al. (2003a) reported that overexpression of JAM-A increased two-dimensional cell migration in endothelial cells through interactions with _vβ₃ integrin and activation of MAPK. However, neither of these studies addressed the role of JAM-A dimerization, nor did they provide a structural model for how JAM-A might regulate cell motility. In addition, there are no studies examining the role of JAM-A in models of cell migration in three dimensions.

We previously reported that transient down-regulation of JAM-A expression in intestinal epithelial cells by siRNA resulted in altered epithelial cell morphology, decreased cell–matrix adhesion, and decreased levels of β1 integrin and Rap1 (Mandell et al., 2005); however, mechanistic insight(s) was lacking. This study was directed at better understanding the mechanism linking structural determinants of JAM-A to β1 integrin levels and cell migration. We determined that overexpression of JAM-A dimerization-defective mutants in 293T cells resulted in decreased cell migration across matrix-coated permeable filters, decreased spreading, and reduced length of cellular protrusions. The role of dimerization in cell migration was further confirmed through experiments demonstrating inhibition of cell migration after treatment with specific JAM-A dimer-disrupting antibodies. Furthermore, treatment with a dimerization-inhibiting antibody and cyclohexamide lead to degradation of β1 integrin more quickly than treatment with cyclohexamide and IgG. Decreased cell migration in our assays correlated with decreased β1 integrin levels, alterations in β1 integrin protein localization, decreased levels of the active form of the small GTPase Rap1, and diminished numbers of focal concentrations of phosphorylated paxillin. An effector role for β1 integrin in the observed JAM-A–mediated effects was supported by experiments with siRNA specific to β1 integrin, demonstrating decreased cell migration in control cell lines after down-regulation of β1 integrin protein levels. Furthermore, we observed increased cell migration comparable with that observed in control cell line levels after overexpression of β1 integrins in cells expressing JAM-A dimerization-defective mutations.

Although the mechanism of decreased β1 integrin in the JAM-A dimerization mutants remains unclear, our results provide important new insights. Our data obtained using two different approaches to disrupt JAM-A dimers (JAM-A mutants and antibodies) are consistent with a scenario in which disruption of JAM-A dimerization causes internalization and degradation of β1. This is consistent with the observation of no decrease in β1 integrin mRNA levels in cells expressing dimerization-defective JAM-A. We tested whether increased proteosomal degradation might account for diminished β1 integrins in JAM-A mutant cell lines. Treatment of the dimerization-defective JAM-A cell lines with the proteosome inhibitor MG262 failed to increase β1 integrins despite increasing levels of ubiquitinated proteins (Supplemental Figure 3). This finding suggests that β1 integrin degradation in our cell lines may not be mediated by the proteosome. Further studies are necessary to determine the mechanism for the degradation of β1 integrin in the presence of the JAM-A dimerization-defective mutants.

The mechanism(s) behind regulation of β1 integrin expression by JAM-A remain to be determined. Possibilities include direct interactions of β1 integrins with JAM-associated scaffolding proteins; activation of signaling molecules that affect β1 integrin turnover, such as the small GTPase Rap1; or sequestration of negative regulators of β1 integrin stability by scaffolding complexes. In other studies, JAM-A has been reported to physically interact with vβ3 (Naik and Naik, 2006) and β2 integrin (Fraemohs et al., 2004) and to regulate migration of endothelial cells (Fraemohs et al., 2004; Naik and Naik, 2006); however, we have been unable to detect a direct association between JAM-A and β1 integrin in coimmunoprecipitation experiments. These observations suggest that JAM-A mediates decreased β1 integrin protein levels through an indirect mechanism(s).

A signaling link between JAM-A dimerization and β1 integrin protein internalization/degradation is suggested by the correlation between decreased β1 integrin and levels of the active form of the GTPase Rap1. We previously demonstrated a decrease in active Rap1 after down-regulation of JAM-A expression in SK-CO15 cells (Mandell et al., 2005). Furthermore, we demonstrated that down-regulation of Rap1 by siRNA resulted in decreased β1 integrin levels. Additionally, other studies have linked Rap1 activity and increased integrin protein levels and/or integrin activation (Reedquist et al., 2000; Katagiri et al., 2003; Shimonaka et al., 2003). In concert with our findings in the dimerization-defective cell lines, it is thus likely that Rap1 is a signaling element between JAM-A and β1 integrin. We speculate that JAM-A dimer-dependent activation of Rap1 may be required to prevent internalization and degradation of β1 integrin.

Intriguingly, we observed that the dominant-negative effects of DL1 were abrogated after an additional mutation removed the PDZ binding domain. Because PDZ domains are responsible for interactions with scaffolding proteins, these results suggest that the dimerization-defective mutations may affect cell migration and β1 integrin levels through sequestration of scaffolding proteins. This hypothesis is consistent with our data demonstrating that 293T cells transfected with JAM-A containing a mutation in the PDZ-binding domain (dFLV) led to decreased levels of β1 integrin protein and decreased rates of cell migration. These data suggest that the effects on β1 integrin and migration are mediated by dimerization of wild-type JAM-A with dFLV-JAM-A. Given that transfection of cells with dFLV resulted in much higher levels of expression of the mutant than endogenous JAM-A, it is likely that a majority of the endogenous JAM-A would dimerize with the dFLV mutant. Under such conditions, functional dimers of JAM-A would not be expected to form, resulting in effects similar to those observed with dimerization-defective mutants.

From these findings, we present a hypothetical model of JAM-A function (Figure 10). In the model, cis-dimerization of JAM-A brings into proximity two molecules of JAM-A. Each JAM-A molecule has a C-terminal PDZ binding motif that can interact directly or indirectly with scaffolding proteins such as ZO-1 or Afadin (Bazzoni et al., 2000; Ebnet et al., 2000). This scaffolding complex might interact with β1 integrin via yet unidentified partners leading to stabilization of β1 integrin at the plasma membrane. In other studies, JAM-A has been reported to physically interact with vβ3 (Naik and Naik, 2006) and β2 integrin (Fraemohs et al., 2004) and regulate migration of endothelial cells (Fraemohs et al., 2004; Naik and Naik, 2006); however, we have been unable to detect a direct association between JAM-A and β1 integrin in coimmunoprecipitation experiments. Therefore, it is most likely that scaffolding complexes associated with JAM-A dimers bind and regulate activity of some signaling and endocytic proteins such as Rap1, which can mediate internalization/trafficking of β1 integrin. We speculate that the dimerization-defective JAM-A–expressing mutants disrupt these scaffolding complexes at endogenous JAM-A dimers by sequestering their certain components. This may activate/release yet unknown signaling cascade resulting in accelerated internalization and subsequent degradation of β1 integrin.

View larger version (36K): [in this window] [in a new window]   Figure 10. Model for effect of JAM-A dimerization on β1 integrin levels. In this model, JAM-A dimers are required for close apposition of cytoplasmic tails bound to scaffolding complexes containing signaling elements. Under this scenario, disruption of cis-dimerization by either mutagenesis or a blocking antibody would result in loss of close apposition of these cytoplasmic tail complexes, inhibition of signaling, and subsequent internalization/degradation of β1 integrin through an as of yet undetermined mechanism. Diminished cell surface expression of β1 Integrin would then result in decreased adhesion to matrix and affect cell migration.

Importantly, our model is also compatible with JAM-A dimers interacting with one another in trans, as has been observed in the mouse, but not human crystal structures (Kostrewa et al., 2001; Prota et al., 2003). Transinteracting JAM-A dimers would allow for the close apposition of multiple sets of PDZ binding domains and the formation of large scaffolding/signaling complexes. Cis-dimerization may thus only be a prerequisite for the changes described in this report to occur.

The physiological significance of formation of JAM-A cis-dimers is highlighted in the various functions described for JAM-A per se, which include regulation of cell migration, barrier function (Liu et al., 2000), angiogenesis (Naik et al., 2003a), cell adhesion (Mandell et al., 2005), and determination of cell polarity (Itoh et al., 2001). We recently demonstrated that loss of JAM-A leads to changes in basal intestinal permeability and increased sensitivity to dextran sulfate sodium-induced colitis in vivo (Laukoetter et al., 2007). In this study, loss of JAM-A was shown to result in an altered claudin expression profile. It is tempting to hypothesize that such changes in claudins may be due to altered JAM-A dimer-mediated signaling.

It is possible to speculate on pathophysiological conditions that would result in dissociation of JAM-A cis-dimers. It is likely that the formation of such complexes results from low-affinity interactions that would be very sensitive to changes in levels of JAM-A in the plasma membrane. Therefore, stimuli that decrease abundance of plasma membrane JAM-A by inhibiting protein expression or enhancing internalization would diminish cis-JAM-A dimers. Interestingly, we and others have shown that junctional proteins, including JAM-A are internalized and decrease after a variety of stimuli including exposure to inflammatory cytokines and oxidant stress (Bruewer et al., 2005; Utech et al., 2005). We have also observed similar internalization and diminished levels of junctional proteins and JAM-A in the mucosa from individuals with inflammatory bowel disease (Kucharzik et al., 2001). It is thus tempting to speculate that loss of JAM-A dimers at the cell surface through inflammatory stimuli contributes to the altered permeability and pathophysiology of chronic intestinal inflammatory states. Clearly, more work is needed to fully understand the physiological relevance of JAM-A dimerization.

In summary, these results suggest that dimerization of JAM-A is required for regulating several aspects of cell migration through signaling events. Disruption of JAM-A dimerization presumably prevents the formation of scaffolding protein complexes that prevent and/or lead to signaling events that result in loss of β1 integrin and decreased cell migration. Further studies are needed to better understand the mechanisms of JAM-mediated regulation of integrin expression and cell migration as well as identification of scaffolding complexes involved in JAM-A–mediated signaling events.

	ACKNOWLEDGMENTS

 
We thank Susan Voss for tissue culture expertise and assistance. This study was supported by National Institutes of Health grants R01-DK72564, R01-DK61379, and R01-DK 79392 (to C.A.P.); DK-53202, DK-55679, and DK-59888 (to A.N.), DK-64399 (National Institutes of Health Digestive Disease Research Center tissue culture and morphology grant), and the Crohn's and Colitis Foundation of America (career development award to A.I.I.).

	Footnotes

 
This was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E07-09-0869) on February 13, 2008.

Address correspondence to: Charles A. Parkos (cparkos{at}emory.edu)

Abbreviations used: 6163, dimerization-defective JAM-A mutant E61A/K63A; DL1, dimerization-defective JAM-A mutant with deletion of the distal most immunoglobulin-like loop; FA, focal adhesion; JAM-A, junctional adhesion molecule A; QRT-PCR, quantitative real-time reverse transcription-polymerase chain reaction analysis.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Babinska, A., Kedees, M. H., Athar, H., Ahmed, T., Batuman, O., Ehrlich, Y. H., Hussain, M. M., and Kornecki, E. (2002). F11–receptor (F11R/JAM) mediates platelet adhesion to endothelial cells: role in inflammatory thrombosis. Thromb. Haemost 88, 843–850.[Medline]

Barton, E. S., Forrest, J. C., Connolly, J. L., Chappell, J. D., Liu, Y., Schnell, F. J., Nusrat, A., Parkos, C. A., and Dermody, T. S. (2001). Junction adhesion molecule is a receptor for reovirus. Cell 104, 441–451.[CrossRef][Medline]

Bazzoni, G., Martinez-Estrada, O. M., Orsenigo, F., Cordenonsi, M., Citi, S., and Dejana, E. (2000). Interaction of junctional adhesion molecule with the tight junction components ZO-1, cingulin, and occludin. J. Biol. Chem 275, 20520–20526.[Abstract/Free Full Text]

Bazzoni, G., Tonetti, P., Manzi, L., Cera, M. R., Balconi, G., and Dejana, E. (2005). Expression of junctional adhesion molecule-A prevents spontaneous and random motility. J. Cell Sci 118, 623–632.[Abstract/Free Full Text]

Bruewer, M., Utech, M., Ivanov, A. I., Hopkins, A. M., Parkos, C. A., and Nusrat, A. (2005). Interferon-gamma induces internalization of epithelial tight junction proteins via a macropinocytosis-like process. FASEB J 19, 923–933.[Abstract/Free Full Text]

Campbell, J. A., Schelling, P., Wetzel, J. D., Johnson, E. M., Forrest, J. C., Wilson, G. A., Aurrand-Lions, M., Imhof, B. A., Stehle, T., and Dermody, T. S. (2005). Junctional adhesion molecule a serves as a receptor for prototype and field-isolate strains of mammalian reovirus. J. Virol 79, 7967–7978.[Abstract/Free Full Text]

Corada, M. et al. (2005). Junctional adhesion molecule-A-deficient polymorphonuclear cells show reduced diapedesis in peritonitis and heart ischemia-reperfusion injury. Proc. Natl. Acad. Sci. USA 102, 10634–10639.[Abstract/Free Full Text]

Del Maschio, A. et al. (1999). Leukocyte recruitment in the cerebrospinal fluid of mice with experimental meningitis is inhibited by an antibody to junctional adhesion molecule (JAM). J. Exp. Med 190, 1351–1356.[Abstract/Free Full Text]

Ebnet, K., Schulz, C. U., Meyer Zu Brickwedde, M. K., Pendl, G. G., and Vestweber, D. (2000). Junctional adhesion molecule interacts with the PDZ domain-containing proteins AF-6 and ZO-1. J. Biol. Chem 275, 27979–27988.[Abstract/Free Full Text]

Ebnet, K., Suzuki, A., Horikoshi, Y., Hirose, T., Meyer Zu Brickwedde, M. K., Ohno, S., and Vestweber, D. (2001). The cell polarity protein ASIP/PAR-3 directly associates with junctional adhesion molecule (JAM). EMBO J 20, 3738–3748.[CrossRef][Medline]

Forrest, J. C., Campbell, J. A., Schelling, P., Stehle, T., and Dermody, T. S. (2003). Structure-function analysis of reovirus binding to junctional adhesion molecule 1. Implications for the mechanism of reovirus attachment. J. Biol. Chem 278, 48434–48444.[Abstract/Free Full Text]

Fraemohs, L., Koenen, R. R., Ostermann, G., Heinemann, B., and Weber, C. (2004). The functional interaction of the beta 2 integrin lymphocyte function-associated antigen-1 with junctional adhesion molecule-A is mediated by the I domain. J. Immunol 173, 6259–6264.[Abstract/Free Full Text]

Franke, B., Akkerman, J. W., and Bos, J. L. (1997). Rapid Ca2+-mediated activation of Rap1 in human platelets. EMBO J 16, 252–259.[CrossRef][Medline]

Guglielmi, K. M., Kirchner, E., Holm, G. H., Stehle, T., and Dermody, T. S. (2007). Reovirus binding determinants in junctional adhesion molecule-a. J. Biol. Chem 282, 17930–17940.[Abstract/Free Full Text]

Gupta, S. K., Pillarisetti, K., and Ohlstein, E. H. (2000). Platelet agonist F11 receptor is a member of the immunoglobulin superfamily and identical with junctional adhesion molecule (JAM): regulation of expression in human endothelial cells and macrophages. IUBMB Life 50, 51–56.[Medline]

Hamazaki, Y., Itoh, M., Sasaki, H., Furuse, M., and Tsukita, S. (2002). Multi-PDZ domain protein 1 (MUPP1) is concentrated at tight junctions through its possible interaction with claudin-1 and junctional adhesion molecule. J. Biol. Chem 277, 455–461.[Abstract/Free Full Text]

Itoh, M., Sasaki, H., Furuse, M., Ozaki, H., Kita, T., and Tsukita, S. (2001). Junctional adhesion molecule (JAM) binds to PAR-3, a possible mechanism for the recruitment of PAR-3 to tight junctions. J. Cell Biol 154, 491–497.[Abstract/Free Full Text]

Ivanov, A. I., McCall, I. C., Babbin, B., Samarin, S. N., Nusrat, A., and Parkos, C. A. (2006). Microtubules regulate disassembly of epithelial apical junctions. BMC Cell Biol 7, 12.[CrossRef][Medline]

Katagiri, K., Maeda, A., Shimonaka, M., and Kinashi, T. (2003). RAPL, a Rap1-binding molecule that mediates Rap1-induced adhesion through spatial regulation of LFA-1. Nat. Immunol 4, 741–748.[CrossRef][Medline]

Khandoga, A., Kessler, J. S., Meissner, H., Hanschen, M., Corada, M., Motoike, T., Enders, G., Dejana, E., and Krombach, F. (2005). Junctional adhesion molecule-A deficiency increases hepatic ischemia-reperfusion injury despite reduction of neutrophil transendothelial migration. Blood 106, 725–733.[Abstract/Free Full Text]

Kornecki, E., Walkowiak, B., Naik, U. P., and Ehrlich, Y. H. (1990). Activation of human platelets by a stimulatory monoclonal antibody. J. Biol. Chem 265, 10042–10048.[Abstract/Free Full Text]

Kostrewa, D. et al. (2001). X-ray structure of junctional adhesion molecule: structural basis for homophilic adhesion via a novel dimerization motif. EMBO J 20, 4391–4398.[CrossRef][Medline]

Kucharzik, T., Walsh, S. V., Chen, J., Parkos, C. A., and Nusrat, A. (2001). Neutrophil transmigration in inflammatory bowel disease is associated with differential expression of epithelial intercellular junction proteins. Am. J. Pathol 159, 2001–2009.[Abstract/Free Full Text]

Laukoetter, M. G. et al. (2007). JAM-A regulates permeability and inflammation in the intestine in vivo. J. Exp. Med 204, 3067–3076.[Abstract/Free Full Text]

Liang, T. W. et al. (2000). Characterization of huJAM: evidence for involvement in cell-cell contact and tight junction regulation. Am. J. Physiol. Cell Physiol 279, C1733–C1743.[Abstract/Free Full Text]

Lisanti, M. P., Caras, I. W., Davitz, M. A., and Rodriguez-Boulan, E. (1989). A glycophospholipid membrane anchor acts as an apical targeting signal in polarized epithelial cells. J. Cell Biol 109, 2145–2156.[Abstract/Free Full Text]

Liu, Y., Nusrat, A., Schnell, F. J., Reaves, T. A., Walsh, S., Pochet, M., and Parkos, C. A. (2000). Human junction adhesion molecule regulates tight junction resealing in epithelia. J. Cell Sci 113, 2363–2374.[Abstract]

Mandell, K. J., Babbin, B. A., Nusrat, A., and Parkos, C. A. (2005). Junctional adhesion molecule 1 regulates epithelial cell morphology through effects on beta1 integrins and Rap1 activity. J. Biol. Chem 280, 11665–11674.[Abstract/Free Full Text]

Mandell, K. J., McCall, I. C., and Parkos, C. A. (2004). Involvement of the junctional adhesion molecule-1 (JAM1) homodimer interface in regulation of epithelial barrier function. J. Biol. Chem 279, 16254–16262.[Abstract/Free Full Text]

Martin-Padura, I. et al. (1998). Junctional adhesion molecule, a novel member of the immunoglobulin superfamily that distributes at intercellular junctions and modulates monocyte transmigration. J. Cell Biol 142, 117–127.[Abstract/Free Full Text]

Naik, M. U., Mousa, S. A., Parkos, C. A., and Naik, U. P. (2003a). Signaling through JAM-1 and alphavbeta3 is required for the angiogenic action of bFGF: dissociation of the JAM-1 and alphavbeta3 complex. Blood 102, 2108–2114.[Abstract/Free Full Text]

Naik, M. U., and Naik, U. P. (2006). Junctional adhesion molecule-A-induced endothelial cell migration on vitronectin is integrin alpha v beta 3 specific. J. Cell Sci 119, 490–499.[Abstract/Free Full Text]

Naik, M. U., Vuppalanchi, D., and Naik, U. P. (2003b). Essential role of junctional adhesion molecule-1 in basic fibroblast growth factor-induced endothelial cell migration. Arterioscler. Thromb. Vasc. Biol 23, 2165–2171.[Abstract/Free Full Text]

Naik, U. P., Ehrlich, Y. H., and Kornecki, E. (1995). Mechanisms of platelet activation by a stimulatory antibody: cross-linking of a novel platelet receptor for monoclonal antibody F11 with the Fc gamma RII receptor. Biochem. J 310, 155–162.[Medline]

Ostermann, G., Fraemohs, L., Baltus, T., Schober, A., Lietz, M., Zernecke, A., Liehn, E. A., and Weber, C. (2005). Involvement of JAM-A in mononuclear cell recruitment on inflamed or atherosclerotic endothelium: inhibition by soluble JAM-A. Arterioscler. Thromb. Vasc. Biol 25, 729–735.[Abstract/Free Full Text]

Ostermann, G., Weber, K. S., Zernecke, A., Schroder, A., and Weber, C. (2002). JAM-1 is a ligand of the beta(2) integrin LFA-1 involved in transendothelial migration of leukocytes. Nat. Immunol 3, 151–158.[CrossRef][Medline]

Ozaki, H., Ishii, K., Arai, H., Horiuchi, H., Kawamoto, T., Suzuki, H., and Kita, T. (2000). Junctional adhesion molecule (JAM) is phosphorylated by protein kinase C upon platelet activation. Biochem. Biophys. Res. Commun 276, 873–878.