Recognition of E-cadherin by integrin alpha(E)beta(7): requirement for cadherin dimerization and implications for cadherin and integrin function.

We have investigated the importance of dimerization of E-cadherin in the heterophilic adhesive interaction between E-cadherin and integrin alpha(E)beta(7). Dimerization of cadherin molecules in parallel alignment is known to be essential for homophilic adhesion and has been attributed to Ca(2+)-dependent interactions in the domain 1-2 junction or to cross-intercalation of Trp2 from one molecule to the other. We have disrupted either or both of these proposed mechanisms by point mutations in E-cadherin-Fc and have tested the modified proteins for alpha(E)beta(7)-mediated cell adhesion. Prevention of Trp2 intercalation had no adverse effect on integrin-mediated adhesion, whereas disruption of Ca(2+) binding permitted adhesion but with reduced efficiency. Both modifications in combination abolished recognition by alpha(E)beta(7). In EGTA, alpha(E)beta(7) adhered to wild type E-cadherin but not to the Trp2 deletion mutant. Independent evidence that the mutations prevented either or both mechanisms for dimerization is presented. The data show that dimerization is required for recognition by alpha(E)beta(7) and that it can take place by either of two mechanisms. Implications for the roles of the alpha(E) and beta(7) integrin subunits in ligand binding and for Trp2 and Ca(2+) in the assembly of cadherin complexes are discussed.

Identification of a binding site for integrin alphaEbeta7 in the N-terminal domain of E-cadherin.

The integrin alphaEbeta7, which is predominantly expressed on mucosal T lymphocytes, has recently been shown to recognize the cell adhesion molecule, E-cadherin, on epithelial cells. We have carried out mutations on E-cadherin, involving domain deletions as well as substitutions of specific amino acids, in order to identify the sites recognized by the integrin. Binding of alphaEbeta7 required the presence of the first two N-terminal domains of E-cadherin. Deletion of extracellular domains 3 and 4 or truncation of the cytoplasmic domain of E-cadherin had no consequence on integrin binding. Substitution of a glutamic acid in the BC loop of the Ig structure of the fist, N-terminal, domain of E-cadherin abrogated binding of alphaEbeta7. This mutation did not appear to affect the conformation of the domain nor the pattern of expression of E-cadherin on the cell surface. Synthetic peptides encompassing the first domain of E-cadherin had very little inhibitory effect on the interaction with alphaEbeta7. Our results highlight structural dissimilarities between recognition of E-cadherin by alphaEbeta7 and recognition of other members of the IgSF by integrins and show that the heterophilic (integrin binding) and homophilic sites in the N-terminal domain of E-cadherin are distinct.

Recognition of E-cadherin on epithelial cells by the mucosal T cell integrin alpha M290 beta 7 (alpha E beta 7).

The integrin alpha M290 beta 7 on the surface of a T cell hybridoma, MTC-1, mediated adhesion of these cells to the mouse epithelial cell line CMT93. This interaction was critically dependent on the presence of divalent cations; Mn2+ strongly promoted adhesion, Ca2+ was ineffective and Mg2+ gave intermediate results. Antibodies to molecules on the surface of CMT93 cells were tested for inhibition of adhesion. One monoclonal antibody (mAb) against E-cadherin, ECCD-2, was found to have significant inhibitory activity. Other mAb to E-cadherin and antibodies to other molecules had no effect. To show that inhibition by ECCD-2 was specific for adhesion mediated by alpha M290 beta 7, MTC-1 cells were induced to adhere to CMT93 via the LFA-1/ICAM-1 pathway. For this purpose, the epithelial cells were treated with interferon-gamma and tumor necrosis factor-alpha to induce ICAM-1 expression and, in addition, alpha M290 beta 7 on MTC-1 cells was down-regulated by culturing the cells in the absence of transforming growth factor beta. Under these circumstances adhesion of MTC-1 cells to CMT93 was inhibited by an antibody to LFA-1 but not by ECCD-2. Transfection of mouse L cells with cDNA for mouse E-cadherin enabled MTC-1 cells to adhere to them through the alpha M290 beta 7 integrin; this interaction was inhibited both by ECCD-2 and by blocking antibody against the integrin. These data strongly suggest that E-cadherin is a principal ligand for alpha M290 beta 7.

Adhesion of human epidermal keratinocytes to laminin.

We have examined the mechanism by which human epidermal keratinocytes adhere to the A/B1/B2 (alpha 1 beta 1 gamma 1) form of laminin. Adhesion could be completely inhibited with an antibody to the beta 1 integrin subunit or a combination of antibodies recognising the alpha 2 beta 1, alpha 3 beta 1 and alpha 6 beta 4 integrins. Keratinocytes adhered in the presence of magnesium and manganese ions, but calcium ions did not support adhesion and inhibited adhesion when combined with magnesium and manganese. The effects of anti-integrin antibodies (including a stimulatory antibody to the beta 1 subunit) were not influenced by specific cations, with the exception that inhibition by an antibody to alpha 2 beta 1 was abrogated by the presence of manganese ions. The E3 and E8 proteolytic fragments of laminin did not support keratinocyte adhesion and heat inactivation of the E8 site in intact laminin did not reduce adhesion. Three laminin fragments that did support adhesion were P1, E4 and E1X-Nd, P1 activity being attributable at least in part to the RGD site; antibody blocking experiments suggested that adhesion to these fragments was primarily via alpha 3 beta 1. The synthetic peptide GD-6, derived from the carboxy terminus of the laminin A chain (included within E3) did support adhesion, but the significance of this observation is unclear, since a scrambled control peptide could also support adhesion. In conclusion, keratinocyte adhesion to A/B1/B2 laminin involves three integrins and multiple binding sites that are different from those defined previously.

Interaction of membranes of the Golgi complex with microtubules in vitro.

We have developed an in vitro assay for characterizing the binding of elements of the Golgi complex to microtubules. The binding assay comprises three distinct components, Golgi elements purified from Vero cells by subcellular fractionation, taxol-polymerized tubulin from bovine brain coupled to magnetic beads and cytosol from HeLa cells. Binding of Golgi elements to microtubules is quantitated by measuring the activity of the Golgi marker enzyme, galactosyltransferase, associated with the microtubule-coated beads retrieved with a magnet. In the presence of cytosol, 35 to 45% of the total input of galactosyltransferase activity (Golgi elements) bind to microtubules; only 3% of the Golgi elements bind to microtubules, however, in the absence of cytosolic factors. This binding is saturable at a cytosol concentration of approximately 5 mg/ml or at a high input of Golgi elements. Cytosol-stimulated binding of Golgi elements to microtubules is decreased to less than 15% when cytosol is pretreated with 2 mM N-ethylmaleimide (NEM) and it is abolished when cytosolic proteins are inactivated by heat or when microtubules have been coated with heat-stable microtubule-associated proteins (MAPs). Trypsinization of the membranes of the Golgi elements abolishes their ability to bind to microtubules. Furthermore, inactivation of cytoplasmic dynein by UV/vanadate treatment does not affect the binding. This suggests that the interaction of Golgi elements with microtubules depends on NEM-sensitive cytosolic factors and membrane-associated receptors, but not on the microtubule-based motor protein cytoplasmic dynein.

Movement of interphase Golgi apparatus in fused mammalian cells and its relationship to cytoskeletal elements and rearrangement of nuclei.

Virus-induced Vero cell fusion was used to analyze the rearrangement of Golgi apparatus during the development of syncytia. Individual Golgi apparatus, associated initially with the separate microtubule-organizing centers in the perinuclear area of fused cells, congregated in the center of the syncytia and formed an extended Golgi complex within 3 to 5 h. The relocation of the Golgi apparatus, but not of nuclei, depended on the presence of an intact microtubule network, since both the microtubule depolymerizing drug nocodazole and the microtubule-stabilizing drug taxol interfered with the formation of an extended Golgi complex. Depolymerization of microfilaments with cytochalasin D and the complete collapse of intermediate filaments induced by microinjected monoclonal antibodies against vimentin had no effect on these processes. Cooling cells to 20 degrees C inhibited both congregation of Golgi apparatus and relocation of nuclei. Visualization of the movement of Golgi apparatus labeled in living cells with fluorescent metabolites of C6-NBD-ceramide showed that relocation of the Golgi apparatus was a process in which congregation and coalescence of the intact organelles was seen, rather than dispersal and reassembly of smaller Golgi elements in the center of the polykaryons. Thus, movement of intact Golgi apparatus in fused interphase cells depends on an undisturbed microtubule network and occurs independently of the relocation of nuclei.