Calcium-dependent homotypic cell-cell adhesion, mediated by molecules such as E-cadherin, guides the establishment of classical epithelial cell polarity and contributes to the control of migration, growth, and differentiation. These actions involve additional proteins, including a-and ,B-catenin (or plakoglobin) and p120, as well as linkage to the cortical actin cytoskeleton. The molecular basis for these interactions and their hierarchy of interaction remain controversial. We demonstrate a direct interaction between Factin and a(E)-catenin, an activity not shared by either the cytoplasmic domain of E-cadherin or 1-catenin. Sedimentation assays and direct visualization by transmission electron microscopy reveal that ai(E)-catenin binds and bundles Factin in vitro with micromolar affinity at a catenin/G-actin monomer ratio of -1:7 (mol/mol). Recombinant human ,B-catenin can simultaneously bind to the a-catenin/actin complex but does not bind actin directly. Recombinant fragments encompassing the amino-terminal 228 residues of al(E)-catenin or the carboxyl-terminal 447 residues individually bind actin in cosedimentation assays with reduced affinity compared with the full-length protein, and neither fragment bundles actin. Except for similarities to vinculin, neither region contains sequences homologous to established actin-binding proteins. Collectively these data indicate that a1(E)-catenin is a novel actin-binding and -bundling protein and support a model in which ac(E)-catenin is responsible for organizing and tethering actin filaments at the zones of E-cadherin-mediated cell-cell contact.

Because of its requirement for signaling by multiple cytokines, Janus kinase 3 (JAK3) is an excellent target for clinical immunosuppression. We report the development of a specific, orally active inhibitor of JAK3, CP-690,550, that significantly prolonged survival in a murine model of heart transplantation and in cynomolgus monkeys receiving kidney transplants. CP-690,550 treatment was not associated with hypertension, hyperlipidemia, or lymphoproliferative disease. On the basis of these preclinical results, we believe JAK3 blockade by CP-690,550 has potential for therapeutically desirable immunosuppression in human organ transplantation and in other clinical settings.

The cadherin-based transmembrane cell-cell adhesive complex is thought to be composed of a cadherin molecule, a ␤-catenin, and an ␣-catenin, which connects the complex to the cytoskeleton. The precise stoichiometry of this complex remains uncertain. We have used a series of recombinant molecules and biophysical techniques to assess the multimeric state of human ␣-and ␤-catenin in vitro and then visualized them by electron microscopy after rotary shadowing. Calculated solution molecular masses are 213 kDa for ␣-catenin, 73 kDa for ␤-catenin, and 186 kDa for both. This suggests that ␣-catenin exists as a homodimer in solution, ␤-catenin is a monomer, and when both are present, they form ␣/␤-catenin heterodimers. Co-precipitation and surface plasmon resonance assays localize the site of ␣-catenin dimerization to the NH 2 -terminal 228 amino acids. This region encompasses a high-affinity (K d ‫؍‬ 100 nM) binding site for ␤-catenin that lies between residues 54 and 157. We anticipate that the oligomeric state of ␣-catenin and the relative stoichiometry of the components in the membrane adhesion complex will be dynamic and regulated by ␤-catenin, cell adhesion, and probably other factors as well.

Cadherin mediated cell-cell adhesion requires cytoplasmic connections to the cytoskeleton mediated by a-catenin. Original descriptions of the catenins, as well as our own in vitro studies, have suggested that this connection was mediated by the interaction of a-catenin to actin. Loss of adhesion in the human colon carcinoma cell line "Clone A" is the result of an internal deletion mutation of 158 residues near the N-terminus of the protein resulting in an 80 kD mutated protein. Transfection of these cells with the full length protein restores the normal adhesive phenotype. We have characterized this mutant protein in efforts to understand the normal function of a-catenin and, in particular, the region deleted in the Clone A mutant. Coprecipitation experiments using whole cell lysates indicate that the mutant form of a-catenin binds p-catenin and plakoglobin, and can form a structural complex with E-cadherin via these interactions. Actin ca-sedimentation assays show that the recombinant mutant binds and bundle,s F-actin and binds both actin and p-catenin simultaneously, as seen with wild type a-catenin. These results suggest that the stabilization of the E-cadherin-catenin complex may be mediated by factors beyond its direct interaction with actin. We conclude that a region near the N-terminus of a-catenin mediates additional interactions between the adhesive complex and the cytoskeleton that are critical for functional adhesion.