Cadherins are a family of cell-cell adhesion molecules which play a central role in controlling morphogenetic movements during development. Cadherin function is regulated by its association with the actin containing cytoskeleton, an association mediated by a complex of cytoplasmic proteins, the catenins: alpha, beta, and gamma. Phosphorylated tyrosine residues on beta-catenin are correlated with loss of cadherin function. Consistent with this, we find that only nontyrosine phosphorylated beta-catenin is associated with N-cadherin in E10 chick retina tissue. Moreover, we demonstrate that a PTP1B-like tyrosine phosphatase associates with N-cadherin and may function as a regulatory switch controlling cadherin function by dephosphorylating beta-catenin, thereby maintaining cells in an adhesion-competent state. The PTP1B-like phosphatase is itself tyrosine phosphorylated. Moreover, both direct binding experiments performed with phosphorylated and dephosphorylated molecules, and treatment of cells with tyrosine kinase inhibitors indicate that the interaction of the PTP1B-like phosphatase with N-cadherin depends on its tyrosine phosphorylation. Concomitant with the tyrosine kinase inhibitor-induced loss of the PTP1B-like phosphatase from its association with N-cadherin, phosphorylated tyrosine residues are retained on beta-catenin, the association of N-cadherin with the actin containing cytoskeleton is lost and N-cadherin-mediated cell adhesion is prevented. Tyrosine phosphatase inhibitors also result in the accumulation of phosphorylated tyrosine residues on beta-catenin, loss of the association of N-cadherin with the actin-containing cytoskeleton, and prevent N-cadherin mediated adhesion, presumably by directly blocking the function of the PTP1B-like phosphatase. We previously showed that the binding of two ligands to the cell surface N-acetylgalactosaminylphosphotransferase (GalNAcPTase), the monoclonal antibody 1B11 and a proteoglycan with a 250-kD core protein, results in the accumulation of phosphorylated tyrosine residues on beta-catenin, uncoupling of N-cadherin from its association with the actin containing cytoskeleton, and loss of N-cadherin function. We now report that binding of these ligands to the GalNAcPTase results in the absence of the PTP1B-like phosphatase from its association with N-cadherin as well as the loss of the tyrosine kinase and tyrosine phosphatase activities that otherwise co-precipitate with N-cadherin. Control antibodies and proteoglycans have no such effect. This effect is similar to that observed with tyrosine kinase inhibitors, suggesting that the GalNAcPTase/proteoglycan interaction inhibits a tyrosine kinase, thereby preventing the phosphorylation of the PTP1B-like phosphatase, and its association with N-cadherin. Taken together these data indicate that a PTP1B-like tyrosine phosphatase can regulate N-cadherin function through its ability to dephosphorylate beta-catenin and that the association of the phosphatase with N-cadherin is regulated via the interaction of the GalNAcPTase with its proteoglycan ligand. In this manner the GalNAcPTase-proteoglycan interaction may play a major role in morphogenetic cell and tissue interactions during development.

The subcellular distribution of the 1,4-dihydropyridine receptor was determined in rabbit skeletal muscle in situ by immunofluorescence and immunoelectron microscopy. Longitudinal and transverse cryosections (5-8 microns) of rabbit gracilis muscle were labeled with monoclonal antibodies specific against either the alpha 1-subunit (170,000-D polypeptide) or the beta-subunit (52,000-D polypeptide) of the 1,4-dihydropyridine receptor by immunofluorescence labeling. In longitudinal sections, specific labeling was present only near the interface between the A- and I-band regions of the sarcomeres. In transverse sections, specific labeling showed a hexagonal staining pattern within each myofiber however, the relative staining intensity of the type II (fast) fibers was judged to be three- to fourfold higher than that of the type I (slow) fibers. Specific immunofluorescence labeling of the sarcolemma was not observed in either longitudinal or transverse sections. These results are consistent with the idea that the alpha 1-subunit and the beta-subunit of the purified 1,4-dihydropyridine receptor are densely distributed in the transverse tubular membrane. Immunoelectron microscopical localization with a monoclonal antibody to the alpha 1-subunit of the 1,4-dihydropyridine receptor showed that the 1,4-dihydropyridine receptor is densely distributed in the transverse tubular membrane. Approximately half of these were distributed in close proximity to the junctional region between the transverse tubules and the terminal cisternae. Specific labeling was also present in discrete foci in the subsarcolemmal region of the myofibers. The size and the nonrandom distribution of these foci in the subsarcolemmal region support the possibility that they correspond to invaginations from the sarcolemma called caveolae. In conclusion, our results demonstrate that the 1,4-dihydropyridine receptor in skeletal muscle is localized to the transverse tubular membrane and discrete foci in the subsarcolemmal region, possibly caveolae but absent from the lateral portion of the sarcolemma.

Microtubule assembly in diploid human skin fibroblasts was studied by [3H]colchicine binding to disaggregated microtubule subunits and to total cell tubulin. Microtubule content per milligram of cell protein was critically dependent upon cell density. As cultures neared confluence, microtubules increased and total cell tubulin decreased; the percent of tubulin assembled into microtubules increased from 5.3 in spare cultures to 58.3 in confluent cultures. Microtubules disappeared with a half-time of 2 min in response to 0 degree C incubation and reformed upon rewarming. Brief treatment of intact cells with concanavalin A or cytochalasin A depolymerized microtubules to 55 or 56% of control levels. The effect of concanavalin A was prevented by alpha-methylmannoside. Fibroblast microtubule assembly was not significantly altered by cyclic nucleotides, ascorbate, glucose, insulin, medium calcium concentration, or calcium ionophore A23187.

Hog anterior pituitary secretory granules sediment at 3,000 g. When rat or rabbit skeletal muscle actin filaments are present with the granules, the sedimentation decreases markedly. Depolymerized actin or viscous solutions of Ficoll and collagen have no effect on granule sedimentation. With this assay, actin filaments bind secretory granules (consisting of the proteinaceous core plus limiting membrane), secretory granule membranes, mitochondria, artificial lecithin liposomes, and styrene-butadiene microspheres, but have little or no interaction with membrane-free secretory granule cores and albumin microspheres. A secretory granule-actin complex sedimentable between 3,000 g and 25,000 g can be isolated. Metal ions, nucleotides, salts, dithiothreitol, or pretreatment of the granules with trypsin do not destroy the binding, which appears to be a lipophilic interaction.