Showing papers in "Journal of Cell Biology in 1993"

Occludin: a novel integral membrane protein localizing at tight junctions.

Abstract: Recently, we found that ZO-1, a tight junction-associated protein, was concentrated in the so called isolated adherens junction fraction from the liver (Itoh, M., A. Nagafuchi, S. Yonemura, T. Kitani-Yasuda, Sa. Tsukita, and Sh. Tsukita. 1993. J. Cell Biol. 121:491-502). Using this fraction derived from chick liver as an antigen, we obtained three monoclonal antibodies specific for a approximately 65-kD protein in rats. This antigen was not extractable from plasma membranes without detergent, suggesting that it is an integral membrane protein. Immunofluorescence and immunoelectron microscopy with these mAbs showed that this approximately 65-kD membrane protein was exclusively localized at tight junctions of both epithelial and endothelial cells: at the electron microscopic level, the labels were detected directly over the points of membrane contact in tight junctions. To further clarify the nature and structure of this membrane protein, we cloned and sequenced its cDNA. We found that the cDNA encoded a 504-amino acid polypeptide with 55.9 kDa. A search of the data base identified no proteins with significant homology to this membrane protein. A most striking feature of its primary structure was revealed by a hydrophilicity plot: four putative membrane-spanning segments were included in the NH2-terminal half. This hydrophilicity plot was very similar to that of connexin, an integral membrane protein in gap junctions. These findings revealed that an integral membrane protein localizing at tight junctions is now identified, which we designated as "occludin."

Transforming growth factor-beta 1 induces alpha-smooth muscle actin expression in granulation tissue myofibroblasts and in quiescent and growing cultured fibroblasts.

Abstract: Granulation tissue fibroblasts (myofibroblasts) develop several ultrastructural and biochemical features of smooth muscle (SM) cells, including the presence of microfilament bundles and the expression of alpha-SM actin, the actin isoform typical of vascular SM cells. Myofibroblasts have been proposed to play a role in wound contraction and in retractile phenomena observed during fibrotic diseases. We show here that the subcutaneous administration of transforming growth factor-beta 1 (TGF beta 1) to rats results in the formation of a granulation tissue in which alpha-SM actin expressing myofibroblasts are particularly abundant. Other cytokines and growth factors, such as platelet-derived growth factor and tumor necrosis factor-alpha, despite their profibrotic activity, do not induce alpha-SM actin in myofibroblasts. In situ hybridization with an alpha-SM actin probe shows a high level of alpha-SM actin mRNA expression in myofibroblasts of TGF beta 1-induced granulation tissue. Moreover, TGF beta 1 induces alpha-SM actin protein and mRNA expression in growing and quiescent cultured fibroblasts and preincubation of culture medium containing whole blood serum with neutralizing antibodies to TGF beta 1 results in a decrease of alpha-SM actin expression by fibroblasts in replicative and non-replicative conditions. These results suggest that TGF beta 1 plays an important role in myofibroblast differentiation during wound healing and fibrocontractive diseases by regulating the expression of alpha-SM actin in these cells.

Flexural rigidity of microtubules and actin filaments measured from thermal fluctuations in shape.

Abstract: Microtubules are long, proteinaceous filaments that perform structural functions in eukaryotic cells by defining cellular shape and serving as tracks for intracellular motor proteins. We report the first accurate measurements of the flexural rigidity of microtubules. By analyzing the thermally driven fluctuations in their shape, we estimated the mean flexural rigidity of taxol-stabilized microtubules to be 2.2 x 10(-23) Nm2 (with 6.4% uncertainty) for seven unlabeled microtubules and 2.1 x 10(-23) Nm2 (with 4.7% uncertainty) for eight rhodamine-labeled microtubules. These values are similar to earlier, less precise estimates of microtubule bending stiffness obtained by modeling flagellar motion. A similar analysis on seven rhodamine-phalloidin-labeled actin filaments gave a flexural rigidity of 7.3 x 10(-26) Nm2 (with 6% uncertainty), consistent with previously reported results. The flexural rigidity of these microtubules corresponds to a persistence length of 5,200 microns showing that a microtubule is rigid over cellular dimensions. By contrast, the persistence length of an actin filament is only approximately 17.7 microns, perhaps explaining why actin filaments within cells are usually cross-linked into bundles. The greater flexural rigidity of a microtubule compared to an actin filament mainly derives from the former's larger cross-section. If tubulin were homogeneous and isotropic, then the microtubule's Young's modulus would be approximately 1.2 GPa, similar to Plexiglas and rigid plastics. Microtubules are expected to be almost inextensible: the compliance of cells is due primarily to filament bending or sliding between filaments rather than the stretching of the filaments themselves.

Signal transduction from the extracellular matrix.

Abstract: Adhesive interactions between cells and the insoluble meshwork of extracellular matrix proteins play a vital role in embryonic morphogenesis (33, 36, 94, 109, 135, 145), and in the regulation of gene expression in cells of the adult organism (1, 6, 105, 124). While the overall phenomenology ofextracellular matrix (ECM) 1 effects on cell differentiation is well known, the biochemical and molecular bases for these effects have remained elusive. It is clear that many of the interactions between cells and the ECM are mediated by the integrin family of cell surface receptors (2, 3, 13, 72). However, the precise mechanism(s) whereby signals from ECM proteins are transduced via integfins to the intraceUular machinery that controls cell growth, behavior, and differentiation, remains poorly defined. There are many compelling examples of control of cell differentiation and gene expression through adhesive interactions with extracellular matrix. In fibroblasts, cell attachment has been reported to rapidly increase expression of c-los and pro al(I) collagen messages (26, 27). Adhesion to fibronectin fragments, or cross-linking of the integfin oe5/~l fibronectin receptor with antibody, induced the expression of metalloprotease genes in fibroblastic cells; interestingly, intact fibronectin did not provoke this response nor did fibronectin fragments in solution (137). In a somewhat similar vein, stimulation of the C~v//~3 integrin in melanoma cells induced the expression of type IV collagenase and increased the invasive ability of these cells (115). The capacity of breast epithelial cells to express milk proteins in response to hormonal stimuli is quite dependent on the presence of an appropriate ECM (124). Studies in this system have led to the preliminary identification of matrix-dependent elements in the promoter region of the ~ casein gene (111). In the immune system, activation of T-lymphocytes through the T-cell antigen receptor is markedly enhanced by integrin-mediated adhesion to fibronectin or laminin (85, 97, 119). This process is part of a complex dialogue involving adhesive receptors occurring between mature T-cells and antigen presenting cells, as well as during lymphocyte differentiation (40, 132, 133). There is extensive signaling \"cross talk\" between

A role for the dystrophin-glycoprotein complex as a transmembrane linker between laminin and actin

Abstract: The dystrophin-glycoprotein complex was tested for interaction with several components of the extracellular matrix as well as actin. The 156-kD dystrophin-associated glycoprotein (156-kD dystroglycan) specifically bound laminin in a calcium-dependent manner and was inhibited by NaCl (IC50 = 250 mM) but was not affected by 1,000-fold (wt/wt) excesses of lactose, IKVAV, or YIGSR peptides. Laminin binding was inhibited by heparin (IC50 = 100 micrograms/ml), suggesting that one of the heparin-binding domains of laminin is involved in binding dystroglycan while negatively charged oligosaccharide moieties on dystroglycan were found to be necessary for its laminin-binding activity. No interaction between any component of the dystrophin-glycoprotein complex and fibronectin, collagen I, collagen IV, entactin, or heparan sulfate proteoglycan was detected by 125I-protein overlay and/or extracellular matrix protein-Sepharose precipitation. In addition, laminin-Sepharose quantitatively precipitated purified dystrophin-glycoprotein complex, demonstrating that the laminin-binding site is accessible when dystroglycan is associated with the complex. Dystroglycan of nonmuscle tissues also bound laminin. However, the other proteins of the striated muscle dystrophin-glycoprotein complex appear to be absent, antigenically dissimilar or less tightly associated with dystroglycan in nonmuscle tissues. Finally, we show that the dystrophin-glycoprotein complex cosediments with F-actin but does not bind calcium or calmodulin. Our results support a role for the striated muscle dystrophin-glycoprotein complex in linking the actin-based cytoskeleton with the extracellular matrix. Furthermore, our results suggest that dystrophin and dystroglycan may play substantially different functional roles in nonmuscle tissues.

Mis-assembly of clathrin lattices on endosomes reveals a regulatory switch for coated pit formation.

Abstract: The clathrin-coated pit lattice is held onto the plasma membrane by an integral membrane protein that binds the clathrin AP-2 subunit with high affinity. In vitro studies have suggested that this protein controls the assembly of the pit because membrane bound AP-2 is required for lattice assembly. If so, the AP-2 binding site must be a resident protein of the coated pit and recycle with other receptors that enter cells through this pathway. Proper recycling, however, would require the switching off of AP-2 binding to allow the binding site to travel through the endocytic pathway unencumbered. Evidence for this hypothesis has been revealed by the cationic amphiphilic class of drugs (CAD), which have previously been found to inhibit receptor recycling. Incubation of human fibroblasts in the presence of these drugs caused clathrin lattices to assemble on endosomal membranes and at the same time prevented coated pit assembly at the cell surface. These effects suggest that CADs reverse an on/off switch that controls AP-2 binding to membranes. We conclude that cells have a mechanism for switching on and off AP-2 binding during the endocytic cycle.

Signal transducing molecules and glycosyl-phosphatidylinositol-linked proteins form a caveolin-rich insoluble complex in MDCK cells

Abstract: GPI-linked protein molecules become Triton-insoluble during polarized sorting to the apical cell surface of epithelial cells. These insoluble complexes, enriched in cholesterol, glycolipids, and GPI-linked proteins, have been isolated by flotation on sucrose density gradients and are thought to contain the putative GPI-sorting machinery. As the cellular origin and molecular protein components of this complex remain unknown, we have begun to characterize these low-density insoluble complexes isolated from MDCK cells. We find that these complexes, which represent 0.4-0.8% of the plasma membrane, ultrastructurally resemble caveolae and are over 150-fold enriched in a model GPI-anchored protein and caveolin, a caveolar marker protein. However, they exclude many other plasma membrane associated molecules and organelle-specific marker enzymes, suggesting that they represent microdomains of the plasma membrane. In addition to caveolin, these insoluble complexes contain a subset of hydrophobic plasma membrane proteins and cytoplasmically-oriented signaling molecules, including: (a) GTP-binding proteins--both small and heterotrimeric; (b) annex II--an apical calcium-regulated phospholipid binding protein with a demonstrated role in exocytic fusion events; (c) c-Yes--an apically localized member of the Src family of non-receptor type protein-tyrosine kinases; and (d) an unidentified serine-kinase activity. As we demonstrate that caveolin is both a transmembrane molecule and a major phospho-acceptor component of these complexes, we propose that caveolin could function as a transmembrane adaptor molecule that couples luminal GPI-linked proteins with cytoplasmically oriented signaling molecules during GPI-membrane trafficking or GPI-mediated signal transduction events. In addition, our results have implications for understanding v-Src transformation and the actions of cholera and pertussis toxins on hetero-trimeric G proteins.

Loss of epithelial differentiation and gain of invasiveness correlates with tyrosine phosphorylation of the E-cadherin/beta-catenin complex in cells transformed with a temperature-sensitive v-SRC gene.

Abstract: Loss of histotypic organization of epithelial cells is a common feature in normal development as well as in the invasion of carcinomas. Here we show that the v-src oncogene is a potent effector of epithelial differentiation and invasiveness. MDCK epithelial cells transformed with a temperature-sensitive mutant of v-src exhibit a strictly epithelial phenotype at the nonpermissive temperature for pp60v-src activity (40.5 degrees C) but rapidly loose cell-to-cell contacts and acquire a fibroblast-like morphology after culture at the permissive temperature (35 degrees C). Furthermore, the invasiveness of the cells into collagen gels or into chick heart fragments was increased at the permissive temperature. The profound effects of v-src on intercellular adhesion were not linked to changes in the levels of expression of the epithelial cell adhesion molecule E-cadherin. Rather, we observed an increase in tyrosine phosphorylation of E-cadherin and, in particular, of the associated protein beta-catenin. These results suggest a mechanism by which v-src counteracts junctional assembly and thereby promotes invasiveness and dedifferentiation of epithelial cells through phosphorylation of the E-cadherin/catenin complex.

The cell proliferation-associated antigen of antibody Ki-67: a very large, ubiquitous nuclear protein with numerous repeated elements, representing a new kind of cell cycle-maintaining proteins.

Abstract: The antigen defined by mAb Ki-67 is a human nuclear protein the expression of which is strictly associated with cell proliferation and which is widely used in routine pathology as a "proliferation marker" to measure the growth fraction of cells in human tumors. Ki-67 detects a double band with apparent molecular weights of 395 and 345 kD in immunoblots of proteins from proliferating cells. We cloned and sequenced the full length cDNA, identified two differentially spliced isoforms of mRNA with open reading frames of 9,768 and 8,688 bp encoding for this cell proliferation-associated protein with calculated molecular weights of 358,761 D and 319,508 D, respectively. New mAbs against a bacterially expressed part and a synthetic polypeptide deduced from the isolated cDNA react with the native Ki-67 antigen, thus providing a circle of evidence that we have cloned the authentic Ki-67 antigen cDNA. The central part of the Ki-67 antigen cDNA contains a large 6,845-bp exon with 16 tandemly repeated 366-bp elements, the "Ki-67 repeats", each including a highly conserved new motif of 66 bp, the "Ki-67 motif", which encodes for the epitope detected by Ki-67. Computer analysis of the nucleic acid and the deduced amino acid sequence of the Ki-67 antigen confirmed that the cDNA encodes for a nuclear and short-lived protein without any significant homology to known sequences. Ki-67 antigen-specific antisense oligonucleotides inhibit the proliferation of IM-9 cell line cells, indicating that the Ki-67 antigen may be an absolute requirement for maintaining cell proliferation. We conclude that the Ki-67 antigen defines a new category of cell cycle-associated nuclear nonhistone proteins.

Identification and differential subcellular localization of the neuronal class C and class D L-type calcium channel alpha 1 subunits.

Abstract: To identify and localize the protein products of genes encoding distinct L-type calcium channels in central neurons, anti-peptide antibodies specific for the class C and class D alpha 1 subunits were produced Anti-CNC1 directed against class C immunoprecipitated 75% of the L-type channels solubilized from rat cerebral cortex and hippocampus Anti-CND1 directed against class D immunoprecipitated only 20% of the L-type calcium channels Immunoblotting revealed two size forms of the class C L-type alpha 1 subunit, LC1 and LC2, and two size forms of the class D L-type alpha 1 subunit, LD1 and LD2 The larger isoforms had apparent molecular masses of approximately 200-210 kD while the smaller isoforms were 180-190 kD, as estimated from electrophoresis in gels polymerized from 5% acrylamide Immunocytochemical studies using CNC1 and CND1 antibodies revealed that the alpha 1 subunits of both L-type calcium channel subtypes are localized mainly in neuronal cell bodies and proximal dendrites Relatively dense labeling was observed at the base of major dendrites in many neurons Staining in more distal dendritic regions was faint or undetectable with CND1, while a more significant level of staining of distal dendrites was observed with CNC1, particularly in the dentate gyrus and the CA2 and CA3 areas of the hippocampus Class C calcium channels were concentrated in clusters, while class D calcium channels were generally distributed in the cell surface membrane of cell bodies and proximal dendrites Our results demonstrate multiple size forms and differential localization of two subtypes of L-type calcium channels in the cell bodies and proximal dendrites of central neurons The differential localization and multiple size forms may allow these two channel subtypes to participate in distinct aspects of electrical signal integration and intracellular calcium signaling in neuronal cell bodies The preferential localization of these calcium channels in cell bodies and proximal dendrites implies their involvement in regulation of calcium-dependent functions occurring in those cellular compartments such as protein phosphorylation, enzyme activity, and gene expression

Scatter factor/hepatocyte growth factor and its receptor, the c-met tyrosine kinase, can mediate a signal exchange between mesenchyme and epithelia during mouse development.

Abstract: Scatter factor/hepatocyte growth factor (SF/HGF) has potent motogenic, mitogenic, and morphogenetic activities on epithelial cells in vitro. The cell surface receptor for this factor was recently identified: it is the product of the c-met protooncogene, a receptor-type tyrosine kinase. We report here the novel and distinct expression patterns of SF/HGF and its receptor during mouse development, which was determined by a combination of in situ hybridization and RNase protection experiments. Predominantly, we detect transcripts of c-met in epithelial cells of various developing organs, whereas the ligand is expressed in distinct mesenchymal cells in close vicinity. In addition, transient SF/HGF and c-met expression is found at certain sites of muscle formation; transient expression of the c-met gene is also detected in developing motoneurons. SF/HGF and the c-met receptor might thus play multiple developmental roles, most notably, mediate a signal given by mesenchyme and received by epithelial. Mesenchymal signals are known to govern differentiation and morphogenesis of many epithelia, but the molecular nature of the signals has remained poorly understood. Therefore, the known biological activities of SF/HGF in vitro and the embryonal expression pattern reported here indicate that this mesenchymal factor can transmit morphogenetic signals in epithelial development and suggest a molecular mechanism for mesenchymal epithelial interactions.

Mutations in human dynamin block an intermediate stage in coated vesicle formation

Abstract: The role of human dynamin in receptor-mediated endocytosis was investigated by transient expression of GTP-binding domain mutants in mammalian cells. Using assays which detect intermediates in coated vesicle formation, the dynamin mutants were found to block endocytosis at a stage after the initiation of coat assembly and preceding the sequestration of ligands into deeply invaginated coated pits. Membrane transport from the ER to the Golgi complex was unaffected indicating that dynamin mutants specifically block early events in endocytosis. These results demonstrate that mutations in the GTP-binding domain of dynamin block Tfn-endocytosis in mammalian cells and suggest that a functional dynamin GTPase is required for receptor-mediated endocytosis via clathrin-coated pits.

Differential expression of mRNAs for neurotrophins and their receptors after axotomy of the sciatic nerve

Abstract: The neurotrophin family includes NGF, brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3), and neurotrophin-4 (NT-4). Previous studies have demonstrated that expression of NGF and its low-affinity receptor is induced in nonneuronal cells of the distal segment of the transected sciatic nerve suggesting a role for NGF during axonal regeneration (Johnson, E. M., M. Taniuchi, and P. S. DeStefano. 1988. Trends Neurosci. 11:299-304). To assess the role of the other neurotrophins and the members of the family of Trk signaling neurotrophin receptors, we have here quantified the levels of mRNAs for BDNF, NT-3, and NT-4 as well as mRNAs for trkA, trkB, and trkC at different times after transection of the sciatic nerve in adult rats. A marked increase of BDNF and NT-4 mRNAs in the distal segment of the sciatic nerve was seen 2 wk after the lesion. The increase in BDNF mRNA was mediated by a selective activation of the BDNF exon IV promoter and adrenalectomy attenuated this increase by 50%. NT-3 mRNA, on the other hand, decreased shortly after the transection but returned to control levels 2 wk later. In Schwann cells ensheathing the sciatic nerve, only trkB mRNA encoding truncated TrkB receptors was detected with reduced levels in the distal part of the lesioned nerve. Similar results were seen using a probe that detects all forms of trkC mRNA. In the denervated gastrocnemius muscle, the level of BDNF mRNA increased, NT-3 mRNA did not change, while NT-4 mRNA decreased. In the spinal cord, only small changes were seen in the levels of neutrophin and trk mRNAs. These results show that expression of mRNAs for neurotrophins and their Trk receptors is differentially regulated after a peripheral nerve injury. Based on these results a model is presented for how the different neurotrophins could cooperate to promote regeneration of injured peripheral nerves.

Growth factors induce nuclear translocation of MAP kinases (p42mapk and p44mapk) but not of their activator MAP kinase kinase (p45mapkk) in fibroblasts

Abstract: Mitogen-activated protein kinases (p42mapk and p44mapk) are serine/threonine kinases that are activated rapidly in cells stimulated with various extracellular signals. This activation is mediated via MAP kinase kinase (p45mapkk), a dual specificity kinase which phosphorylates two key regulatory threonine and tyrosine residues of MAP kinases. We reported previously that the persistent phase of MAP kinase activation is essential for mitogenically stimulated cells to pass the "restriction point" of the cell cycle. Here, using specific polyclonal antibodies and transfection of epitope-tagged recombinant MAP kinases we demonstrate that these signaling protein kinases undergo distinct spatio-temporal localization in growth factor-stimulated cells. In G0-arrested hamster fibroblasts the activator p45mapkk and MAP kinases (p42mapk, p44mapk) are mainly cytoplasmic. Subsequent to mitogenic stimulation by serum or alpha-thrombin both MAP kinase isoforms translocate into the nucleus. This translocation is rapid (seen in 15 min), persistent (at least during the entire G1 period up to 6 h), reversible (by removal of the mitogenic stimulus) and apparently 'coupled' to the mitogenic potential; it does not occur in response to nonmitogenic agents such as alpha-thrombin-receptor synthetic peptides and phorbol esters that fail to activate MAP kinases persistently. When p42mapk and p44mapk are expressed stably at high levels, they are found in the nucleus of resting cells; this nuclear localization is also apparent with kinase-deficient mutants (p44mapk T192A or Y194F). In marked contrast the p45mapkk activator remains cytoplasmic even during prolonged growth factor stimulation and even after high expression levels achieved by transfection. We propose that the rapid and persistent nuclear transfer of p42mapk and p44mapk during the entire G0-G1 period is crucial for the function of these kinases in mediating the growth response.

Maximal migration of human smooth muscle cells on fibronectin and type IV collagen occurs at an intermediate attachment strength.

Abstract: Although a biphasic dependence of cell migration speed on cell-substratum adhesiveness has been predicted theoretically, experimental data directly demonstrating a relationship between these two phenomena have been lacking. To determine whether an optimal strength of cell-substratum adhesive interactions exists for cell migration, we measured quantitatively both the initial attachment strength and migration speed of human smooth muscle cells (HSMCs) on a range of surface concentrations of fibronectin (Fn) and type IV collagen (CnIV). Initial attachment strength was measured in order to characterize short time-scale cell-substratum interactions, which may be representative of dynamic interactions involved in cell migration. The critical fluid shear stress for cell detachment, determined in a radial-flow detachment assay, increased linearly with the surface concentrations of adsorbed Fn and CnIV. The detachment stress required for cells on Fn, 3.6 +/- 0.2 x 10(-3) mu dynes/absorbed molecule, was much greater than that on CnIV, 5.0 +/- 1.4 x 10(-5) mu dynes/absorbed molecule. Time-lapse videomicroscopy of individual cell movement paths showed that the migration behavior of HSMCs on these substrates varied with the absorbed concentration of each matrix protein, exhibiting biphasic dependence. Cell speed reached a maximum at intermediate concentrations of both proteins, with optimal concentrations for migration at 1 x 10(3) molecules/micron2 and 1 x 10(4) molecules/micron2 on Fn and CnIV, respectively. These optimal protein concentrations represent optimal initial attachment strengths corresponding to detachment shear stresses of 3.8 mu dyne/micron2 on Fn and 1.5 mu dyne/micron2 on CnIV. Thus, while the optimal absorbed protein concentrations for migration on Fn and CnIV differed by an order of magnitude, the optimal initial attachment strengths for migration on these two proteins were very similar. Further, the same minimum strength of initial attachment, corresponding to a detachment shear stress of approximately 1 mu dyne/micron2, was required for movement on either protein. These results suggest that initial cell-substratum attachment strength is a central variable governing cell migration speed, able to correlate observations of motility on substrata differing in adhesiveness. They also demonstrate that migration speed depends in biphasic manner on attachment strength, with maximal migration at an intermediate level of cell-substratum adhesiveness.

Peptides derived from two separate domains of the matrix protein thrombospondin-1 have anti-angiogenic activity

Abstract: Thrombospondin-1 (TSP1) is a large modular matrix protein containing three identical disulfide-linked 180-kD chains that inhibits neovascularization in vivo (Good et al., 1990). To determine which of the structural motifs present in the 180-kD TSP1 polypeptide mediate the anti-angiogenic activity, a series of protease-generated fragments were tested using several in vitro and in vivo assays that reflect angiogenic activity. The majority of the anti-angiogenic activity of TSP1 resides in the central 70-kD stalk region which alone could block neovascularization induced by bFGF in the rat cornea in vivo and inhibit both migration in a modified Boyden chamber and [3H]thymidine incorporation stimulated by bFGF in cultured capillary endothelial cells. Although TSP1 has been shown to bind active TGF beta 1, this cytokine could not account for the inhibitory effects of the stalk region of TSP1 on cultured endothelial cells. Peptides and truncated molecules were used to further localize inhibitory activity to two domains of the central stalk, the procollagen homology region and the properdin-like type 1 repeats. Trimeric recombinant TSP1 containing NH2-terminal sequences truncated after the procollagen-like module inhibited endothelial cell migration in vitro and corneal neovascularization in vivo whereas trimeric molecules truncated before this domain were inactive as was the NH2-terminal heparin-binding domain that is present in both recombinant molecules. A series of peptides from the procollagen-like region, the smallest of which consisted of residues 303-309 of TSP1, inhibited angiogenesis in vivo in the rat cornea and the migration of endothelial cells in vitro. A 19-residue peptide containing these sequences blocked vessel formation in the granulation tissue invading a polyvinyl sponge implanted into the mouse. Nineteen residue peptides derived from two of the three type 1 repeats present in the intact TSP1 molecule blocked neovascularization in vivo in the rat cornea and inhibited the migration of cultured endothelial cells with ED50's of 0.6-7 microM. One of these peptides, containing residues 481-499 of TSP1, also inhibited vessel formation in granulation tissue invading sponges in vivo. These results suggest that the large TSP1 molecule employs at least two different structural domains and perhaps two different mechanisms to accomplish a single physiological function, the inhibition of neovascularization. The definition of short peptides from each of these domains that are able to block the angiogenic process may be of use in designing targeted inhibitors of the pathological neovascularization that underlies many diseases.

Fluorescent labeling of nascent RNA reveals transcription by RNA polymerase II in domains scattered throughout the nucleus

Abstract: Several nuclear activities and components are concentrated in discrete nuclear compartments. To understand the functional significance of nuclear compartmentalization, knowledge on the spatial distribution of transcriptionally active chromatin is essential. We have examined the distribution of sites of transcription by RNA polymerase II (RPII) by labeling nascent RNA with 5-bromouridine 5'-triphosphate, in vitro and in vivo. Nascent RPII transcripts were found in over 100 defined areas, scattered throughout the nucleoplasm. No preferential localization was observed in either the nuclear interior or the periphery. Each transcription site may represent the activity of a single gene or, considering the number of active pre-mRNA genes in a cell, of a cluster of active genes. The relation between the distribution of nascent RPII transcripts and that of the essential splicing factor SC-35 was investigated in double labeling experiments. Antibodies against SC-35 recognize a number of well-defined, intensely labeled nuclear domains, in addition to labeling of more diffuse areas between these domains (Spector, D. L., X. -D. Fu, and T. Maniatis. 1991. EMBO (Eur. Mol. Biol. Organ.) J. 10:3467-3481). We observe no correlation between intensely labeled SC-35 domains and sites of pre-mRNA synthesis. However, many sites of RPII synthesis colocalize with weakly stained areas. This implies that contranscriptional splicing takes place in these weakly stained areas. These areas may also be sites where splicing is completed posttranscriptionally. Intensely labeled SC-35 domains may function as sites for assembly, storage, or regeneration of splicing components, or as compartments for degradation of introns.

The 220-kD protein colocalizing with cadherins in non-epithelial cells is identical to ZO-1, a tight junction-associated protein in epithelial cells: cDNA cloning and immunoelectron microscopy.

Abstract: We previously identified a 220-kD constitutive protein of the plasma membrane undercoat which colocalizes at the immunofluorescence microscopic level with cadherins and occurs not only in epithelial M., S. Yonemura, A. Nagafuchi, Sa. Tsukita, and Sh. Tsukita. 1991. J. Cell Biol. 115:1449-1462). To clarify the nature and possible functions of this protein, we cloned its full-length cDNA and sequenced it. Unexpectedly, we found mouse 220-kD protein to be highly homologous to rat protein ZO-1, only a part of which had been already sequenced. This relationship was confirmed by immunoblotting with anti-ZO-1 antibody. As protein ZO-1 was originally identified as a component exclusively underlying tight junctions in epithelial cells, where cadherins are not believed to be localized, we analyzed the distribution of cadherins and the 220-kD protein by ultrathin cryosection immunoelectron microscopy. We found that in non-epithelial cells lacking tight junctions cadherins and the 220-kD protein colocalize, whereas in epithelial cells (e.g., intestinal epithelial cells) bearing well-developed tight junctions cadherins and the 220-kD protein are clearly segregated into adherens and tight junctions, respectively. Interestingly, in epithelial cells such as hepatocytes, which tight junctions are not so well developed, the 220-kD protein is detected not only in the tight junction zone but also at adherens junctions. Furthermore, we show in mouse L cells transfected with cDNAs encoding N-, P-, E-cadherins that cadherins interact directly or indirectly with the 220-kD protein. Possible functions of the 220-kD protein (ZO-1) are discussed with special reference to the molecular mechanism for adherens and tight junction formation.

The I domain is a major recognition site on the leukocyte integrin Mac-1 (CD11b/CD18) for four distinct adhesion ligands.

Abstract: Despite the identification and characterization of several distinct ligands for the leukocyte integrin (CD11/CD18) family of adhesion receptors, little is known about the structural regions on these molecules that mediate ligand recognition. In this report, we use alpha subunit chimeras of Mac-1 (CD11b/CD18) and p150,95 (CD11c/CD18), and an extended panel of newly generated and previously characterized mAbs specific to the alpha chain of Mac-1 to map the binding sites for four distinct ligands for Mac-1: iC3b, fibrinogen, ICAM-1, and the as-yet uncharacterized counter-receptor responsible for neutrophil homotypic adhesion. Epitopes of mAbs that blocked ligand binding were mapped with the chimeras and used to localize the ligand recognition sites because the data obtained from functional assays with the Mac-1/p150,95 chimeras were not easily interpreted. Results show that the I domain on the alpha chain of Mac-1 is an important recognition site for all four ligands, and that the NH2-terminal and perhaps divalent cation binding regions but not the COOH-terminal segment may contribute. The recognition sites in the I domain appear overlapping but not identical as individual Mac-1-ligand interactions are distinguished by the discrete patterns of inhibitory mAbs. Additionally, we find that the alpha subunit NH2-terminal region and divalent cation binding region, despite being separated by over 200 amino acids of the I domain, appear structurally apposed because three mAbs require the presence of both of these regions for antigenic reactivity, and chimeras that contain the NH2 terminus of p150,95 require the divalent cation binding region of p150,95 to associate firmly with the beta subunit.

Extracellular matrix contains insulin-like growth factor binding protein-5 : potentiation of the effects of IGF-I

Abstract: Insulin-like growth factor binding proteins (IGFBPs) have been shown to serve as carrier proteins for the insulin-like growth factors (IGFs) and to modulate their biologic effects. Since extracellular matrix (ECM) has been shown to be a reservoir for IGF-I and IGF-II, we examined the ECM of cultured human fetal fibroblasts and found that IGFBP-5 was incorporated intact into ECM, while mostly inert proteolytic fragments were found in the medium. In contrast, two other forms of IGFBP that are secreted by these cells were either present in ECM in minimal amounts (IGFBP-3) or not detected (IGFBP-4). Likewise, when purified IGFBPs were incubated with ECM, IGFBP-5 bound preferentially. IGFBP-5 was found to bind to types III and IV collagen, laminin, and fibronectin. Increasing salt concentrations inhibited the binding of IGFBP-5 to ECM and accelerated the release of IGFBP-5 from ECM, suggesting an ionic basis for this interaction. ECM-associated IGFBP-5 had a sevenfold decrease in affinity for IGF-I compared to IGFBP-5 in solution. Furthermore, when IGFBP-5 was present in cell culture substrata, it potentiated the growth stimulatory effects of IGF-I on fibroblasts. When IGFBP-5 was present only in the medium, it was degraded to a 22-kD fragment and had no effect on IGF-I-stimulated growth. We conclude that IGFBP-5 is present in fibroblast ECM, where it is protected from degradation and can potentiate the biologic actions of IGF-I. These findings provide a molecular explanation for the association of the IGF's with the extracellular matrix, and suggest that the binding of the IGF's to matrix, via IGFBP-5, may be important in mediating the cellular growth response to these growth factors.

Temporal analysis of events associated with programmed cell death (apoptosis) of sympathetic neurons deprived of nerve growth factor.

Abstract: The time course of molecular events that accompany degeneration and death after nerve growth factor (NGF) deprivation and neuroprotection by NGF and other agents was examined in cultures of NGF-dependent neonatal rat sympathetic neurons and compared to death by apoptosis. Within 12 h after onset of NGF deprivation, glucose uptake, protein synthesis, and RNA synthesis fell precipitously followed by a moderate decrease of mitochondrial function. The molecular mechanisms underlying the NGF deprivation-induced decrease of protein synthesis and neuronal death were compared and found to be different, demonstrating that this decrease of protein synthesis is insufficient to cause death subsequently. After these early changes and during the onset of neuronal atrophy, inhibition of protein synthesis ceased to halt neuronal degeneration while readdition of NGF or a cAMP analogue remained neuroprotective for 6 h. This suggests a model in which a putative killer protein reaches lethal levels several hours before the neurons cease to respond to readdition of NGF with survival and become committed to die. Preceding loss of viability by 5 h and concurrent with commitment to die, the neuronal DNA fragmented into oligonucleosomes. The temporal and pharmacological characteristics of DNA fragmentation is consistent with DNA fragmentation being part of the mechanism that commits the neuron to die. The antimitotic and neurotoxin cytosine arabinoside induced DNA fragmentation in the presence of NGF, supporting previous evidence that it mimicked NGF deprivation-induced death closely. Thus trophic factor deprivation-induced death occurs by apoptosis and is an example of programmed cell death.

A subpopulation of Mac-1 (CD11b/CD18) molecules mediates neutrophil adhesion to ICAM-1 and fibrinogen.

Abstract: We report that a subpopulation (10%) of the Mac-1 (CD1 1b/CD18) molecules on activated neutrophils mediates adhesion to ICAM-1 and fibrinogen. We describe a novel mAb (CBRM1/5) that binds to an activation-specific neoepitope on a subset of Mac-1 molecules on neutrophils and monocytes after stimulation with chemoattractants or phorobol esters but does not recognize Mac-1 on resting myeloid cells. CBRM1/5 immunoprecipitates a subpopulation of Mac-1 molecules from detergent lysates of neutrophils, binds to immunoaffinity-purified Mac-1, and localizes to the I domain on the alpha chain of Mac-1. Because CBRM1/5 recognizes a fraction of Mac-1 on activated neutrophils, but still blocks Mac-1-dependent adhesion to fibrinogen and ICAM-1, we suggest that only a small subset of Mac-1 molecules is competent to mediate adhesion.

Inhibition of nuclear protein import by nonhydrolyzable analogues of GTP and identification of the small GTPase Ran/TC4 as an essential transport factor.

Abstract: We have investigated a possible involvement of GTPases in nuclear protein import using an in vitro transport system involving digitonin-permeabilized cells supplemented with exogenous cytosol. Transport in this system was measured with a novel ELISA-based assay that allows rapid quantitative analysis. GTP gamma S and other nonhydrolyzable analogues of GTP were found to rapidly inhibit the rate of in vitro nuclear import. Transport inhibition by GTP gamma S was dependent on the concentrations of permeabilized cells and cytosol, and was strongly enhanced by a cytosolic factor(s). The predominant cytosolic component responsible for this inhibition was found in a 20-30-kD fraction in molecular sieving chromatography. Furthermore, a component(s) of this 20-30-kD fraction was itself required for efficient nuclear import. Biochemical complementation with bacterially expressed protein demonstrated that this essential GTP gamma S-sensitive transport factor was Ran/TC4, a previously described GTPase of the Ras superfamily found in both nucleus and cytoplasm. Ran/TC4 and its guanine nucleotide release protein RCC1 have previously been implicated in DNA replication, cell cycle checkpoint control, and RNA synthesis, processing and export. Our results suggest that Ran/TC4 serves to integrate nuclear protein import with these other nuclear activities.

CHIP28 water channels are localized in constitutively water-permeable segments of the nephron.

Abstract: The sites of water transport along the nephron are well characterized, but the molecular basis of renal water transport remains poorly understood. CHIP28 is a 28-kD integral protein which was proposed to mediate transmembrane water movement in red cells and kidney (Preston, G. M., T. P. Carroll, W. B. Guggino, and P. Agre. 1992. Science [Wash. DC]. 256:385-387). To determine whether CHIP28 could account for renal epithelial water transport, we used specific polyclonal antibodies to quantitate and localize CHIP28 at cellular and subcellular levels in rat kidney using light and electron microscopy. CHIP28 comprised 3.8% of isolated proximal tubule brush border protein. Except for the first few cells of the S1 segment, CHIP28 was immunolocalized throughout the convoluted and straight proximal tubules where it was observed in the microvilli of the apical brush border and in basolateral membranes. Very little CHIP28 was detected in endocytic vesicles or other intracellular structures in proximal tubules. Uninterrupted, heavy immunostaining of CHIP28 was also observed over both apical and basolateral membranes of descending thin limbs, including both short and long loops of Henle. These nephron sites have constitutively high osmotic water permeabilities. CHIP28 was not detected in ascending thin limbs, thick ascending limbs, or distal tubules, which are highly impermeable to water. Moreover, CHIP28 was not detected in collecting duct epithelia, where water permeability is regulated by antidiuretic hormone. These determinations of abundance and structural organization provide evidence that the CHIP28 water channel is the predominant pathway for constitutive transepithelial water transport in the proximal tubule and descending limb of Henle's loop.

Cortactin, an 80/85-kilodalton pp60src substrate, is a filamentous actin-binding protein enriched in the cell cortex.

Abstract: Two related cellular proteins, p80 and p85 (cortactin), become phosphorylated on tyrosine in pp60src-transformed cells and in cells stimulated with certain growth factors. The amino-terminal half of cortactin is comprised of multiple copies of an internal, tandem 37-amino acid repeat. The carboxyl-terminal half contains a distal SH3 domain. We report that cortactin is an F-actin-binding protein. The binding to F-actin is specific and saturable. The amino-terminal repeat region appears to be both necessary and sufficient to mediate actin binding, whereas the SH3 domain had no apparent effect on the actin-binding activity. Cortactin, present in several different cell types, is enriched in cortical structures such as membrane ruffles and lamellipodia. The properties of cortactin indicate that it may be important for microfilament-membrane interactions as well as transducing signals from the cell surface to the cytoskeleton. We suggest the name cortactin, reflecting the cortical subcellular localization and its actin-binding activity.

Sorting of membrane components from endosomes and subsequent recycling to the cell surface occurs by a bulk flow process

Abstract: A central question in the endocytic process concerns the mechanism for sorting of recycling components (such as transferrin or low density lipoprotein receptors) from lysosomally directed components; membrane-associated molecules including receptors are generally directed towards the recycling pathway while the luminal content of sorting endosomes, consisting of the acid-released ligands, are lysosomally targeted. However, it is not known whether recycling membrane receptors follow bulk membrane flow or if these proteins are actively sorted from lysosomally directed material because of specific protein sequences and/or structural features. Using quantitative fluorescence microscopy we have determined the endocytic route and kinetics of traffic of the bulk carrier, membrane lipids, to address this issue directly. We show that N-[N-(7-nitro-2,1,3-benzoxadiazol-4-yl)-epsilon-aminohexanoyl]- sphingosylphosphorylcholine (C6-NBD-SM) in endocytosed as bulk membrane, and it transits the endocytic system kinetically and morphologically identically to fluorescently labeled transferrin in a CHO cell line. With indistinguishable kinetics, the two labeled markers sort from lysosomally destined molecules in peripherally located sorting endosomes, accumulate in a peri-centriolar recycling compartment, and finally exit the cell. Other fluorescently labeled lipids, C6-NBD-phosphatidylcholine and galactosylceramide also traverse the same pathway. The constitutive nature of sorting of bulk membrane towards the recycling pathway and the lysosomal direction of fluid phase implies a geometric basis of sorting.

Morphogenesis in the yeast cell cycle: regulation by Cdc28 and cyclins.

Abstract: Analysis of cell cycle regulation in the budding yeast Saccharomyces cerevisiae has shown that a central regulatory protein kinase, Cdc28, undergoes changes in activity through the cell cycle by associating with distinct groups of cyclins that accumulate at different times. The various cyclin/Cdc28 complexes control different aspects of cell cycle progression, including the commitment step known as START and mitosis. We found that altering the activity of Cdc28 had profound effects on morphogenesis during the yeast cell cycle. Our results suggest that activation of Cdc28 by G1 cyclins (Cln1, Cln2, or Cln3) in unbudded G1 cells triggers polarization of the cortical actin cytoskeleton to a specialized pre-bud site at one end of the cell, while activation of Cdc28 by mitotic cyclins (Clb1 or Clb2) in budded G2 cells causes depolarization of the cortical actin cytoskeleton and secretory apparatus. Inactivation of Cdc28 following cyclin destruction in mitosis triggers redistribution of cortical actin structures to the neck region for cytokinesis. In the case of pre-bud site assembly following START, we found that the actin rearrangement could be triggered by Cln/Cdc28 activation in the absence of de novo protein synthesis, suggesting that the kinase may directly phosphorylate substrates (such as actin-binding proteins) that regulate actin distribution in cells.

Cell cycle-dependent specific positioning and clustering of centromeres and telomeres in fission yeast.

Abstract: Fluorescence in situ hybridization (FISH) shows that fission yeast centromeres and telomeres make up specific spatial arrangements in the nucleus. Their positioning and clustering are cell cycle regulated. In G2, centromeres cluster adjacent to the spindle pole body (SPB), while in mitosis, their association with each other and with the SPB is disrupted. Similarly, telomeres cluster at the nuclear periphery in G2 and their associations are disrupted in mitosis. Mitotic centromeres interact with the spindle. They remain undivided until the spindle reaches a critical length, then separate and move towards the poles. This demonstrated, for the first time, that anaphase A occurs in fission yeast. The mode of anaphase A and B is similar to that of higher eukaryotes. In nda3 and cut7 mutants defective in tubulin of a kinesin-related motor, cells are blocked in early stages of mitosis due to the absence of the spindle, and centromeres dissociate but remain close to the SPB, whereas in a metaphase-arrested nuc2 mutant, they reside at the middle of the spindle. FISH is therefore a powerful tool for analyzing mitotic chromosome movement and disjunction using various mutants. Surprisingly, in top2 defective in DNA topoisomerase II, while most chromatid DNAs remain undivided, sister centromeres are separated. Significance of this finding is discussed. In contrast, most chromatid DNAs are separated but telomeric DNAs are not in cut1 mutant. In cut1, the dependence of SPB duplication on the completion of mitosis is abolished. In crm1 mutant cells defective in higher-order chromosome organization, the interphase arrangements of centromeres and telomeres are disrupted.

Transport and localization of exogenous myelin basic protein mRNA microinjected into oligodendrocytes.

Abstract: We have studied transport and localization of MBP mRNA in oligodendrocytes in culture by microinjecting labeled mRNA into living cells and analyzing the intracellular distribution of the injected RNA by confocal microscopy. Injected mRNA initially appears dispersed in the perikaryon. Within minutes, the RNA forms granules which, in the case of MBP mRNA, are transported down the processes to the periphery of the cell where the distribution again becomes dispersed. In situ hybridization shows that endogenous MBP mRNA in oligodendrocytes also appears as granules in the perikaryon and processes and dispersed in the peripheral membranes. The granules are not released by extraction with non-ionic detergent, indicating that they are associated with the cytoskeletal matrix. Three dimensional visualization indicates that MBP mRNA granules are often aligned in tracks along microtubules traversing the cytoplasm and processes. Several distinct patterns of granule movement are observed. Granules in the processes undergo sustained directional movement with a velocity of approximately 0.2 micron/s. Granules at branch points undergo oscillatory motion with a mean displacement of 0.1 micron/s. Granules in the periphery of the cell circulate randomly with a mean displacement of approximately 1 micron/s. The results are discussed in terms of a multi-step pathway for transport and localization of MBP mRNA in oligodendrocytes. This work represents the first characterization of intracellular movement of mRNA in living cells, and the first description of the role of RNA granules in transport and localization of mRNA in cells.

Salmonella typhimurium attachment to human intestinal epithelial monolayers: transcellular signalling to subepithelial neutrophils.

Abstract: In human intestinal disease induced by Salmonella typhimurium, transepithelial migration of neutrophils (PMN) rapidly follows attachment of the bacteria to the epithelial apical membrane. In this report, we model those interactions in vitro, using polarized monolayers of the human intestinal epithelial cell, T84, isolated human PMN, and S. typhimurium. We show that Salmonella attachment to T84 cell apical membranes did not alter monolayer integrity as assessed by transepithelial resistance and measurements of ion transport. However, when human neutrophils were subsequently placed on the basolateral surface of monolayers apically colonized by Salmonella, physiologically directed transepithelial PMN migration ensued. In contrast, attachment of a non-pathogenic Escherichia coli strain to the apical membrane of epithelial cells at comparable densities failed to stimulate a directed PMN transepithelial migration. Use of the n-formyl-peptide receptor antagonist N-t-BOC-1-methionyl-1-leucyl-1- phenylalanine (tBOC-MLP) indicated that the Salmonella-induced PMN transepithelial migration response was not attributable to the classical pathway by which bacteria induce directed migration of PMN. Moreover, the PMN transmigration response required Salmonella adhesion to the epithelial apical membrane and subsequent reciprocal protein synthesis in both bacteria and epithelial cells. Among the events stimulated by this interaction was the epithelial synthesis and polarized release of the potent PMN chemotactic peptide interleukin-8 (IL-8). However, IL-8 neutralization, transfer, and induction experiments indicated that this cytokine was not responsible for the elicited PMN transmigration. These data indicate that a novel transcellular pathway exists in which subepithelial PMN respond to lumenal pathogens across a functionally intact epithelium. Based on the known unique characteristics of the intestinal mucosa, we speculate that IL-8 may act in concert with an as yet unidentified transcellular chemotactic factor(s) (TCF) which directs PMN migration across the intestinal epithelium.