Activation of the type 1 histamine (H1) or the type 2 protease-activated (PAR-2) G protein-coupled receptors interrupts E-cadherin adhesion and decreases the transepithelial resistance (TER) of epithelium. Several reports suggest that cadherin adhesive function depends on the association of cadherin with β-catenin and that this association is regulated by phosphorylation of tyrosines in β-catenin. We tested the hypothesis that loss of cadherin adhesion and compromise of TER on activation of the H1 or PAR-2 receptor is due to phosphorylation of tyrosines in β-catenin. L cells were stably transfected to express E-cadherin (L-E-cad cells) and H1 (L-H1-E-cad cells). L cells and Madin-Darby canine kidney (MDCK) cells constitutively express PAR-2. Stably transfected L-E-cad, L-H1-E-cad, and MDCK cells were also stably transfected with FLAG-tagged wild-type (WT) or mutant β-catenin, converting tyrosine 142, 489, or 654 to the nonphosphorylatable mimetic, phenylalanine (WT, Y142F, Y489F, or Y654F). Activation of H1 or PAR-2 interrupted adhesion to an immobilized E-cadherin-Fc fusion protein of L-H1-E-cad, L-E-cad, and MDCK cells expressing WT or Y142F β-catenin but did not interrupt adhesion of L-H1-E-cad, L-E-cad, and MDCK cells expressing the Y489F or Y654F mutant β-catenins. PAR-2 activation decreased the TER of monolayers of MDCK cells expressing WT or Y142F β-catenin 40–45%. However, PAR-2 activation did not decrease the TER of monolayers of MDCK cells expressing Y489F or Y654F β-catenin. The protein tyrosine phosphatase PTP1B binds to the cadherin cytoplasmic domain and dephosphorylates β-catenin. Inhibition of PTP1B interrupted adhesion to E-cadherin-Fc of MDCK cells expressing WT β-catenin but did not affect the adhesion of MDCK cells expressing Y489F or Y654F β-catenin. Similarly, inhibition of PTP1B compromised the TER of MDCK cells expressing WT β-catenin but did not affect the TER of MDCK cells expressing Y489F or Y654F β-catenin. We conclude that phosphorylation of tyrosines 489 and 654 in β-catenin is a necessary step in the process by which G protein-coupled H1 and PAR-2 receptors interrupt E-cadherin adhesion. We also conclude that activation of PAR-2 has no effect on the TER without first interrupting E-cadherin adhesion.
airway epithelia create an important barrier between the tissues of the lung and the environment. This barrier is dependent on molecules that bind the epithelial cells to each other and to molecules of the tight junction that act as resistors to paracellular molecular traffic. E-cadherin mediates calcium-dependent homotypic cell-cell adhesion in epithelia. We (24, 27) previously reported that activation of the type 1 histamine (H1) or the type 2 protease-activated receptor (PAR-2) decreased the transepithelial resistance (TER) of airway epithelium by interrupting E-cadherin adhesion. In these reports, activation of the H1 or PAR-2 receptor interrupted E-cadherin-dependent adhesion of E-cadherin expressing L cells and of primary airway epithelium to immobilized E-cadherin. Activation of H1 or PAR-2 also caused an immediate, ∼50% decrease in the TER of the primary airway epithelium. The resistance spontaneously restored itself to its initial level over the course of ∼5 min. Supporting an essential role of E-cadherin adhesion in the maintenance of the TER of the epithelium, antibody to E-cadherin abolished the TER. Our goals were to confirm the essential role of E-cadherin adhesion in preserving the TER by rescuing E-cadherin adhesion in the setting of G protein-coupled receptor activation and, in so doing, illuminate some of the mechanisms by which these receptors affect cadherin adhesion.
E-cadherin-mediated cell-cell adhesion appears to depend on the association of E-cadherin with β-catenin. β-Catenin binds to E-cadherin in the endoplasmic reticulum. In the absence of β-catenin binding, the cadherin cytoplasmic domain lacks structure, PEST sites are exposed, and cadherin is rapidly proteolyzed (9, 10). Peptides that compete with β-catenin for binding to cadherin also interrupt cadherin adhesion (1). These observations infer that persistent binding of β-catenin to E-cadherin would be essential for stable E-cadherin adhesion. Two specific tyrosines in the β-catenin-cadherin binding interface are well-conserved and important to β-catenin-cadherin binding. Src or EGF receptor phosphorylation of tyrosine 654 in β-catenin decreases the affinity of binding between β-catenin and E-cadherin ∼6-fold (17). Similarly, interruption of N-cadherin adhesion following activation of the Slit receptor, Robo, is caused by Abelson kinase phosphorylation of tyrosine 489 in β-catenin with a consequent reduction in the binding of β-catenin to N-cadherin (15, 16).
Based on this constellation of observations, we hypothesized that activation of the H1 and PAR-2 receptors stimulates tyrosine phosphorylation of β-catenin with consequent interruption of E-cadherin adhesion resulting in a transient decrease in the TER of epithelium. To test this hypothesis, we examined the effect of expressing mutant β-catenin, with tyrosine (Y) to phenylalanine (F) mutations at tyrosines 489 and 654, on E-cadherin adhesion and the TER of epithelium in the setting of activation of H1 and PAR-2. We found that either of the mutations rescued both E-cadherin adhesion and preserved the TER of epithelium in the setting of activation of H1 and PAR-2.
Fibronectin was from Collaborative Research (Bedford, MA). Tissue culture media and serum were from the Tissue Culture Core, University of Iowa. Madin-Darby canine kidney (MDCK) and L cells were from American Type Culture Collection (Rockville, MD). rr1, Function-blocking antibody to E-cadherin, was from the University of Iowa Hybridoma Core as previously described (24, 27). Monoclonal antibody to β-catenin was from BD Biosciences (Franklin Lakes, NJ), and polyclonal antibody from Santa Cruz Biotechnology (Santa Cruz, CA). HEK-293 cells expressing an E-cadherin-human-Fc fusion protein were a generous gift from W. James Nelson (Stanford University). Secondary antibody was sheep anti-mouse IgG conjugated with horseradish peroxidase from Amersham Biosciences (Piscataway, NJ). Fura 2 was from Molecular Probes (Eugene, OR). Reacti-Bind Protein G-coated strip plates were from Pierce (Rockford, IL).
L cells were grown in DMEM with 10% FBS, penicillin (100 μg/ml), and streptomycin (100 μg/ml). L cells transfected with the human histamine receptor H1 in the pcDNA3.1 vector and E-cadherin in the pHβAPr-1-neo vector were grown as above with the addition of zeocin (1 mg/ml) and G418 (1.4 mg/ml) for selection. When zeocin was present, penicillin and streptomycin were eliminated. To verify that clones expressed a functional histamine receptor at the surface, transfected cells were loaded with fura 2 and assayed for an increase in intracellular calcium in response to histamine as described previously (22, 27).
MDCK cells were grown in DMEM with 10% FBS, penicillin (100 μg/ml), and streptomycin (100 μg/ml). MDCK cells transfected with β-catenin in the LZRS-MS-zeo vector were grown as above with the addition of zeocin (1 mg/ml).
Plasmid preparation and transfection.
L cells expressing the histamine receptor in the vector pcDNA3.1/zeo were transfected with the E-cadherin-pHβAPr-1-neo plasmid using Lipofectamine Plus. Doubly transfected cells were selected with zeocin (1 mg/ml) and G418 (1.4 mg/ml). Clones were isolated based on expression of E-cadherin.
The cDNA for β-catenin [wild-type (WT), Y142F, Y489F, and Y654F] in the pGEX-KG vector were as described by Xu et al. (25). The LZRS-MS-gfp and LZRS-MS-zeo vectors were a generous gift from Albert Reynolds Lab (Vanderbilt University). Permission to use the Phoenix cell lines was granted by Gary Nolan from Stanford University. The β-catenin constructs were shuttled through the pFLAG-CMV-7 vector to pick up an NH2-terminal FLAG tag and then into the LZRS-MS-gfp and LZRS-MS-zeo vectors. They were cut out of the pGEX-KG vector using BamHI and then ligated into the pFLAG-CMV vector using the same restriction enzyme. To transfer the FLAG-tagged β-catenin into the LZRS-MS-gfp and the LZRS-MS-zeo vectors, an EcoRI site needed to be eliminated. To do this, we changed the EcoRI recognition sequence GAATTC to GGATTC on an internal EcoRI site using the QuickChange mutagenesis kit from Stratagene (La Jolla, CA) and the following primers: 5′-primer, 5′-GGTCACCTGTGCAGCTGGGATTCTCTCTAACC-3′; 3′-primer, 5′-GGTTAGAGAGAATCCCAGCTGCACAGGTGACC-3′. This did not change the amino acid sequence. The resulting DNA was sequenced to confirm the elimination of the internal EcoRI site, the maintenance of the mutations, and the presence of the FLAG tag in frame with the β-catenin. The FLAG-tagged β-catenin constructs were then cut out of the pFLAG-CMV vector using EcoRI and ligated into the LZRS-MS-gfp or LZRS-MS-zeo vectors using the same enzyme. The LZRS-MS-gfp or LZRS-MS-zeo vectors containing the β-catenin constructs were used to transfect Phoenix-Ampho cell (ATCC no. SD3443) using CaCl2. Twenty-four hours later, the virus was harvested and used to infect L cells stably transfected to express E-cadherin (L-E-cad cells) and MDCK cells. The L cells were also expressing E-cadherin in the pHβAPr-1-neo vector with or without pcDNA3.1 containing the H1 cDNA. L cells stably transfected to express H1 (L-H1-E-cad cells), L-E-cad, and MDCK cells were selected with G418 and zeocin or sorted for green fluorescent protein (GFP). The presence of FLAG-tagged β-catenin was confirmed by immunoprecipitating with FLAG antibody and blotting with antibody to β-catenin (Figs. 1 and 2).
Analysis of protein expression.
Protein was solubilized with cytoskeleton (CSK) buffer (50 mM NaCl, 10 mM PIPES, pH 6.8, 3 mM MgCl2, 1% Triton X-100, 300 mM sucrose, 5 mg/ml PMSF, 5 μg/ml aprotinin, 1 mg/ml leupeptin, 5 μg/ml pepstatin A, 4 mM sodium orthovanadate, 10 mM NaF, 10 mM NaP2O7), and cell proteins separated on 6% PAGE gels and transferred to polyvinylidene difluoride (PVDF) membranes in 25 mM Tris, 192 mM glycine, 20% methanol, and 0.04% SDS buffer for 1 h at 10 V in a semidry transfer. The membranes were blocked with SEA BLOCK Blocking Buffer (Pierce). Blocked membranes were incubated with primary antibody in 1:1 blocking buffer-to-PBS for 1 h at room temperature, washed 4× with 1:1 blocking buffer-to-PBS, incubated with the secondary antibody, IRDye 800 goat anti-mouse IgG (Rockland, Gilbertsville, PA), for 1 h at room temperature, washed 4× with 1:1 blocking buffer-to-PBS, and then washed 2× with PBS. Blots were scanned using a LI-COR Odyssey Infrared Imaging System (LI-COR, Lincoln, NE).
Immunoprecipitation and blotting.
For immunoprecipitation, L-H1-E-cad, L-E-cad, or MDCK cells expressing β-catenin (WT, Y142F, Y489F, and Y654F) were exposed to 10 μM histamine or PAR-2-AP activating peptide for 1 min. In experiments measuring phosphorylation in response to histamine, cells were preincubated for 3 min with 0.1 mM potassium bisperoxo(1,10-phenanthroline)oxovanadate(V) [bpV(phen)], and bpV(phen) was present during histamine treatment. After exposure, the cells were lysed as described by Hinck et al. (8) by adding 3 ml of CSK buffer (50 mM NaCl, 10 mM PIPES, pH 6.8, 3 mM MgCl2, 1% Triton X-100, 300 mM sucrose, 5 μg/ml PMSF, 5 μg/ml aprotinin, 1 μg/ml leupeptin, 5 μg/ml pepstatin A, 4 mM sodium orthovanadate, 10 mM NaF, and 10 mM NaP2O7) at 4°C and kept on ice for 20 min. The cell lysate was then scraped from the plates and centrifuged at 12,400 g for 10 min. The supernatant solution was collected for immunoprecipitation.
The supernatant material was precleared with 100 μl of sepharose (CL4B200; Sigma) and 50 μl of GammaBind (17–0886-01; Pharmacia) and preincubated with 60 μg of mouse IgG in a total volume of 3 ml of TBS containing protease inhibitors (10 mM Tris, pH 7.5, 150 mM NaCl, 2 mM MgCl2, 2 mM CaCl2, 1 mM PMSF, 2 μg/ml aprotinin, 15 mg/ml leupeptin, and 360 μM 1,10-phenanthroline) for ≥2 h at 4°C. The beads were pelleted, and the precleared solution was incubated with 100 μl of sepharose and 50 μl of GammaBind that had been preincubated with either 3 ml of E-cadherin antibody or 60 μg of FLAG or β-catenin antibody for ≥2 h at 4°C and rinsed 2× with TBS containing protease inhibitors. The samples were rotated overnight at 4°C in the antibody-sepharose-GammaBind mixture. The beads were pelleted at 2,000 g for 5 min. Beads were then washed in high-salt buffer (15 mM Tris, pH 7.5, 5 mM EDTA, 2.5 mM EGTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, and 1 M NaCl) followed by low-salt buffer (15 mM Tris, pH 7.5, and 5 mM EDTA) and boiled in SDS sample buffer before loading. The proteins were separated on 6% SDS-PAGE and transferred to PVDF membranes. Proteins and phosphotyrosine were identified by immunoblotting. Separate portions of the same immunoprecipitate (IP) were blotted for anti-phosphotyrosine or for β-catenin to control for loading for Fig. 3A. For Fig. 3B, the blots were stripped and reblotted with antibody to β-catenin to control for loading.
Adhesion of transfected L cells and MDCK cells to cadherin-Fc.
Cadherin-mediated adhesion to microtiter plate surfaces coated with E-cadherin-Fc was measured as previously described (24, 27). Briefly, protein G-coated strip plates were rinsed three times with 0.05% Tween 20 in PBS, and then 100 μl of the cadherin-Fc fusion protein (30 μg/ml) in PBS with 0.5 mM EGTA (PBS-EGTA) was added to the wells and allowed to bind.
Cells were labeled by replacing the media with HBSCGA (in mM: 135 NaCl, 1.2 CaCl2, 1.2 MgCl2, 10 HEPES, 10 glucose, and 0.05% BSA, pH 7.4 with KOH) containing calcein AM (5 ng/ml) and incubating for 45 min at 37°C. The cells were then rinsed with PBS, harvested following a 10-min (L cells) or 30-min (MDCK cells) incubation in 137 mM NaCl, 4.2 mM NaHCO3, 5.4 mM KCl, 5.6 mM glucose, 0.5 mM EDTA, pH 7.2, pelleted by mild centrifugation, and resuspended in HBSCGA at a final concentration of 100,000 cells/ml. Two hundred microliters of the cell suspension was added to each well and allowed to bind for 45 min at 37°C. Each eight-well strip was then treated with histamine or PAR-2-AP for 1 min or protein tyrosine phosphatase PTP1B inhibitor for 5 min and rinsed 3× with HBSCGA to remove nonadherent cells. The fluorescence remaining in each well was then measured (VICTOR2; EG&G Wallac, Gaithersburg, MD) and used as an estimate of the relative number of adherent cells following background subtraction (24, 27).
For measurement of transepithelial electrical properties, epithelia were mounted in Ussing chambers and studied as previously described (24). Epithelia were bathed in symmetrical solutions containing, in mM, 135 NaCl, 2.4 K2HPO4, 0.6 KH2PO4, 1.2 CaCl2, 1.2 MgCl2, 10 dextrose, 5 HEPES, at pH 7.2, 37°C and gassed with 100% O2. TER was measured before and after application of PAR-2-AP by applying a 5 mV spike every 5 s and measuring the change in current to calculate the resistance.
Changes were compared by ANOVA, and individual group comparisons were done using a Tukey honestly significant difference test for post hoc comparisons of means. Differences were considered significant at the P < 0.05 level.
Expression of β-catenin constructs.
To associate effects of histamine and PAR-2-AP on tyrosine phosphorylation of β-catenin with physiological responses to these agonists, FLAG-tagged β-catenin constructs with specific Y-to-F mutations were expressed in L cells and MDCK cells and tested for their ability to prevent the effects of these agonists on cadherin adhesion and the TER of an epithelium. Mutant constructs at the three tyrosines known to affect binding within the cadherin-β-catenin-α-catenin complex, Y142F, Y489F, and Y654F, were analyzed along with FLAG-tagged WT β-catenin. The FLAG-tagged β-catenin constructs WT, Y142F, Y489F, and Y654F were expressed in L cells and MDCK cells (Fig. 1, A and B; Ref. 24). The β-catenin constructs were associated with E-cadherin as they were present in IPs of E-cadherin, and the fraction of β-catenin in the E-cadherin IPs that was FLAG-tagged was not significantly different among the different constructs (P > 0.05; Fig. 2, A and B).
The hypothesis implies that activation of H1 or PAR-2 increases phosphorylation of β-catenin tyrosines. Indeed, activation of H1 in L-H1-E-cad cells, activation of PAR-2 in MDCK cells, and addition of the tyrosine phosphatase inhibitor, PTP1B, increased tyrosine phosphorylation of both native and transfected β-catenin (Fig. 3, A and B).
The effect of histamine and PAR-2-AP on adhesion of L-H1-E-cad, L-E-cad, and MDCK cells to immobilized E-cadherin.
Adhesion of L-E-cad and E-cadherin expressing epithelial cells to immobilized E-cadherin is dependent on E-cadherin as the adhesion is abolished by antibody to E-cadherin (24, 27). Histamine or PAR-2-AP (10 μM) caused ∼45% of L-H1-E-cad or L-E-cad cells expressing WT or Y142F β-catenin to lose adhesion to immobilized E-cadherin (Fig. 4A). In contrast, adhesion to immobilized E-cadherin of L-H1-E-cad or L-E-cad cells expressing Y489F or Y654F β-catenin was not reduced by histamine or PAR-2-AP.
Hence, in cells expressing WT β-catenin or Y142F β-catenin, histamine and PAR-2-AP stimulate signaling leading to loss of E-cadherin adhesion. In contrast, in cells expressing Y489F or Y654F β-catenin, activation of H1 or PAR-2 does not interrupt E-cadherin adhesion, implying that these agonists initiate signaling causing phosphorylation of β-catenin Y489 and Y654 that results in interruption of E-cadherin adhesion.
We also examined the effects of activation of PAR-2 on adhesion of MDCK epithelial cells to immobilized E-cadherin. The adhesion was abolished by function-blocking antibody to E-cadherin demonstrating that this adhesion is dependent on E-cadherin (data not shown). PAR-2-AP interrupted adhesion of MDCK cells expressing WT or Y142F β-catenin to immobilized E-cadherin (Fig. 4B). In contrast, PAR-2-AP did not interrupt cadherin-dependent adhesion to immobilized E- cadherin of MDCK cells expressing Y489F or Y654F β- catenin (Fig. 4B). These observations in MDCK cells confirm the observations in L cells and support the conclusion that activation of H1 and PAR-2 receptors interrupts E-cadherin adhesion via phosphorylation of tyrosines 489 and 654 in β-catenin.
The effect of PAR-2-AP on the TER of MDCK cell monolayers.
Since the β-catenin Y489F and Y654F constructs preserved E-cadherin-mediated adhesion, we asked whether preservation of cadherin-dependent adhesion would prevent PAR-2-AP from decreasing the TER of monolayers of MDCK cells. Indeed, PAR-2-AP (10 μM) immediately decreased the TER of monolayers of MDCK cells expressing WT or Y142F β-catenin by 40–50% (Fig. 5, A–C). As previously reported, the resistance was spontaneously restored within ∼5 min (24). In contrast, PAR-2-AP did not significantly decrease the TER of MDCK cell monolayers expressing either Y489F or Y654F β-catenin (Fig. 5, A–C). Hence, preventing phosphorylation of tyrosine 489 or 654 of β-catenin preserves cadherin-dependent adhesion and the TER of MDCK cell epithelium.
PTP1B, Y489F, Y654F, and the TER of MDCK epithelia.
We examined the effects of the β-catenin tyrosine mutations on MDCK cell E-cadherin-dependent adhesion and the TER of MDCK cell epithelia when an alternative mechanism was used to increase tyrosine phosphorylation of β-catenin. The protein tyrosine phosphatase PTP1B binds to the cytoplasmic domain of cadherins and limits tyrosine phosphorylation of β-catenin (2). PTP1B inhibitor, a derivative of benzbromarone (Calbiochem), is a highly specific (>2 logs) inhibitor of PTP1B (21). PTP1B inhibitor interrupted the adhesion to E-cadherin-Fc of MDCK cells expressing WT β-catenin, but it did not affect the adhesion to E-cadherin-Fc of MDCK cells expressing Y489F or Y654F β-catenin (Fig. 6). Similarly, PTP1B inhibitor decreased the TER of MDCK cell monolayers expressing WT β-catenin, but it did not significantly decrease the TER of MDCK cell monolayers expressing Y489F or Y654 F β-catenin (Fig. 7, A and B).
E-cadherin mediates calcium-dependent homotypic and homophilic cell-cell adhesion in epithelia. Chelation of extracellular calcium or function-blocking antibody to the E-cadherin extracellular domain abolishes the barrier created by epithelia (7, 19, 20, 24). Our prior reports (24, 27) suggested that activation of the H1 and PAR-2 receptors compromised the TER of airway epithelia by interrupting E-cadherin adhesion. The observations referenced in the Introduction indicating that tyrosine phosphorylation of β-catenin reduces binding of β-catenin to E-cadherin and destabilizes cadherin-mediated adhesion led us to hypothesize that activation of the H1 and PAR-2 G protein-coupled receptors interrupts E-cadherin adhesion by increasing tyrosine phosphorylation of β-catenin (6, 9–12, 24, 27). Confirmation of the hypothesis would rescue cadherin adhesion in the setting of activation of the receptors and demonstrate an essential role of cadherin adhesion in the response of the TER to activation of these G protein-coupled receptors. Confirmation would also further support the hypothesis that tyrosine phosphorylation of β-catenin destabilizes cadherin adhesion. Indeed, activation of H1 or PAR-2 decreased adhesion to immobilized E-cadherin of cells expressing WT or Y142F β-catenin, whereas adhesion to immobilized E-cadherin of cells expressing Y489F or Y654F β-catenin was not affected by activation of H1 or PAR-2. Similarly, activation of PAR-2 decreased the TER of epithelia expressing WT or Y142F β-catenin, but activation of PAR-2 did not alter the TER of epithelia expressing Y489F or Y654F β-catenin, confirming the essential role of E-cadherin adhesion in the response of the TER to activation of these G protein-coupled receptors.
Rescue was effected by either Y489F or Y654F. This suggests that both residues must be phosphorylated to compromise E-cadherin adhesion, whereas a recent study indicated phosphorylation of either of these residues compromised N- cadherin adhesion (15). Phosphorylation of β-catenin Y489 reduces binding affinity between β-catenin and N-cadherin ∼4-fold whereas phosphorylation of β-catenin Y654 reduces binding affinity between β-catenin and E-cadherin 6- to 10-fold (9, 10, 13, 15). It is possible that the fourfold reduction in affinity does not itself disrupt E-cadherin binding to β-catenin but facilitates the phosphorylation of Y654 that then in combination with phosphorylation of Y489 compromises binding of β-catenin and cadherin. This sequence would be in concert with Huber and Weis's observations about the binding of β-catenin to E-cadherin (10).
“Multiple, quasi-independent interactions provide the possibility of having a minimal 'core' binding region while allowing other interactions to be more dynamic. Separate binding regions can be regulated independently, enabling combinatorial regulation of the interaction and the integration of multiple input signals.”
We could not detect persistently significant differences in binding of WT or the Y-to-F β-catenin constructs to E-cadherin after activating H1 or PAR-2. Our observations are consistent with those of Ezaki et al. (5) who found that decreased tyrosine phosphorylation of β-catenin was associated with more stable E-cadherin adhesion in intestinal epithelium despite no detectable differences in the binding of β-catenin to E-cadherin.
Piedra et al. (14) reported that Fer or c-met tyrosine kinases could phosphorylate β-catenin tyrosine 142 with a resultant decrease in binding affinity between α-catenin and β-catenin. Historically, binding of α-catenin to β-catenin was believed to link the cadherin complex to the cortical actin cytoskeleton through the binding of actin to α-catenin either directly or through another actin binding protein. Drees et al. (4) and Yamada et al. (26) more recently reported that α-catenin cannot bind β-catenin and actin or other actin binding proteins at the same time. In our experiments, the Y142F β-catenin mutant did not protect E-cadherin adhesion or the TER of the epithelium following activation of H1 or PAR-2, indicating that phosphorylation of β-catenin Y142 is not involved in the effects of these G protein-coupled receptors on E-cadherin adhesion.
The cytoplasmic domain of E- and N-cadherin binds several regulatory molecules that contribute to determining the state of phosphorylation of β-catenin tyrosines. The tyrosine kinase Fer and the tyrosine phosphatase PTP1B are associated directly or indirectly with the cytoplasmic domain (2, 18, 25). Fer is responsible for tyrosine phosphorylation of PTP1B, a state necessary for PTP1B binding to cadherin (16, 25). PTP1B persistently dephosphorylates β-catenin tyrosines, enhancing binding of β-catenin to cadherin (2, 25). This organization of regulatory molecules is consistent with a recent report that acetaldehyde disrupted the E-cadherin-β-catenin complex in Caco-2 cells in a process dependent on tyrosine phosphorylation (18). In this report, tyrosine phosphorylation of E-cadherin was associated with interruption of the binding of PTP1B to E-cadherin with consequent increased tyrosine phosphorylation of β-catenin. β-Catenin tyrosines 489 and 654 were among those β-catenin tyrosines phosphorylated, although phosphorylation of four other tyrosines in β-catenin was also detected. Our observations that inhibition of PTP1B interrupts E-cadherin adhesion and the TER of MDCK cell epithelia expressing WT but not Y489F or Y654F β-catenin adds additional support for this role of PTP1B.
The TER of epithelial cell monolayers is determined primarily by the resistance to paracellular ion flux and by the state of activation of membrane ion channels (24). We (24, 27) previously reported that the effects of activation of H1 and PAR-2 on the TER of epithelium were not dependent on activation of ion channels. Although there are multiple precedents demonstrating that the TER of an epithelium is dependent on intact cadherin adhesion, it was possible that activation of the G protein-coupled H1 and PAR-2 receptors interrupted cadherin adhesion and also had cadherin-independent effects on the resistance (3, 7, 19, 20, 24). We found that the β-catenin mutations Y489F and Y654F preserved both cadherin adhesion and all of the TER of the epithelium. Hence, the effects of activating these G protein-coupled receptors on the epithelial barrier are solely a result of interrupting E-cadherin adhesion.
This study was supported by National Heart, Lung, and Blood Institute Grant HL-33540 (D. M. Shasby).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society