DISRUPTION OF CELL—CELL ADHESION IN THE PRESENCE OF SODIUM BUTYRATE ACTIVATES EXPRESSION OF THE 92kDa TYPE IV COLLAGENASE IN MDCK CELLS

We have investigated the effects of altered cell shape on the regulation of the 92kDa type IV collagenase. In MDCK cells, anti-E-cadherin antibodies alter cell shape by disrupting normal cell—cell contacts, while sodium butyrate causes a marked flattening and spreading of cells. The disruption of cell—cell contacts led to a faint expression of the 92kDa collagenase. This effect was enhanced by sodium butyrate, which by itself did not induce collagenase expression. In contrast, stromelysin expression was not regulated in these conditions. Although mRNA expression was enhanced, the secreted collagenase activity was not altered in these conditions in either cell line. Examination of cytoskeletal and extracellular matrix proteins and cell—cell and cell—matrix adhesion proteins by immunofluorescence and Western blot revealed a disruption of the actin network, tight junctions, and fibronectin deposition by anti E-cadherin antibodies, and alterations in actin, cytokeratin 8, cytokeratin 14, laminin and β1 integrin induced by sodium butyrate. Thus, the induction of collagenase expression in epithelial cells by disrupted cell—cell adhesion and sodium butyrate is associated with changes in cell shape and structure.     

Phorbol ester modulation of integrin-mediated cell adhesion: a postreceptor event

Chinese hamster ovary (CHO) suspension culture cells adhere readily to substrata coated with extracellular matrix proteins such as fibronectin, vitronectin, or laminin. In the case of fibronectin, it is known that adhesion is mediated by an integrin-type, cell surface fibronectin receptor (FnR). We demonstrate here that treatment of CHO cells with submicromolar concentrations of phorbol ester produces a remarkable increase in the ability of these cells to adhere to fibronectin. Both the rate of adhesion and the efficiency of adhesion are enhanced about four- to fivefold. Further, phorbol ester treatment renders the fibronectin-mediated adhesion process less sensitive to inhibitors, including GRGDSP peptide and PB1, a monoclonal anti-FnR antibody. By contrast, nonspecific adhesion processes, for example cell attachment to substrata coated with polylysine or concanavalin A, are not affected by phorbol ester treatment. Thus integrin-mediated adhesion is modulated by phorbol esters, but nonspecific adhesion is not. Neither the number of cell surface FnRs nor the receptor affinity, as measured by 125I-fibronectin and 125I-anti-FnR antibody binding, is altered by phorbol ester treatment. Thus, the effect of phorbol ester on cell adhesion seems to occur at a step subsequent to initial ligand-receptor binding events. Since phorbol ester is a potent activator of protein kinase C, we examined phosphorylation patterns in control and phorbol-treated cells. In immunoprecipitates of lysates from suspension culture cells, there was no evidence of phorbol ester-stimulated phosphorylation of FnR or of talin, a protein thought to interact with FnR. These results suggest that phorbol ester effects on fibronectin-dependent adhesion are not due to phosphorylation of the FnR itself but rather may be due to postreceptor events, possibly the phosphorylation of cytoskeletal proteins involved in integrin-mediated adhesion.     

Isolation and characterization of Chinese hamster ovary cell variants deficient in the expression of fibronectin receptor

Chinese hamster ovary cell populations were enriched for cells displaying low surface expression of the 140-kD integrin fibronectin receptor (FnR) by means of fluorescence-activated cell sorting using monoclonal anti-FnR antibodies. Selected cells were cloned by limiting dilution, and the resulting clones were screened for low cell surface FnR expression by ELISA. Two multiply sorted populations gave rise to variant clones possessing approximately 20 or 2% FnR expression, respectively, compared with wild-type cells. Growth rates of the "20%" and "2%" clones on serum-coated plastic dishes were similar to that of wild-type cells. Variant cells expressing 20% FnR could attach and spread on substrata coated with purified fibronectin, although somewhat more slowly than wild-type cells, while cells expressing 2% FnR could not attach or spread. Cells from all variant clones attached normally to vitronectin substrata, but some of the 2% clones displayed altered morphology on this type of substratum. Motility assays in blind well chambers showed a correlation of movement with level of expression of FnR. The number of cells migrating in response to fibronectin was greatly reduced compared with wild-type cells for the 20% FnR variant clones, while variant clones with 2% FnR showed virtually no migratory activity. Surface labeling with 125I and immunoaffinity purification of FnR showed reduced levels of intact FnR on the plasma membranes of variants with 20% FnR, while none was detected in variants expressing 2% FnR. Nevertheless, beta subunits were detected on the surfaces of all variant clones. Immunoblots of cell lysates from wild-type cells and from both types of variant clones showed substantial amounts of FnR beta chain as well as enhanced amounts of a pre-beta moiety in the variants. alpha chain was markedly reduced in the 20% variants and essentially absent in the 2% variants, indicating that failure to assemble intact FnR in these variants was due to deficiencies of alpha chain production. Dot blots of total mRNA from a representative clone expressing 20% FnR showed reduced levels of material hybridizing to an 0.97-kb hamster FnR alpha chain cDNA probe as compared with wild type, while mRNA from a representative clone expressing 2% FnR had no detectable hybridizable RNA; this seems to agree well with the results obtained by immunoblotting. Thus, the defect in the variant clones seems to be due to reduced levels of alpha chain mRNA leading to a deficit of mature FnR and consequent alterations in cell adhesion and motility on fibronectin substrata.     

Dynamic cytoskeleton-integrin associations induced by cell binding to immobilized fibronectin

We have examined the early events of cellular attachment and spreading (10-30 min) by allowing chick embryonic fibroblasts transformed by Rous sarcoma virus to interact with fibronectin immobilized on matrix beads. The binding activity of cells to fibronectin beads was sensitive to both the mAb JG22E and the GRGDS peptide, which inhibit the interaction between integrin and fibronectin. The precise distribution of cytoskeleton components and integrin was determined by immunocytochemistry of frozen thin sections. In suspended cells, the distribution of talin was diffuse in the cytoplasm and integrin was localized at the cell surface. Within 10 min after binding of cells and fibronectin beads at 22 degrees C or 37 degrees C, integrin and talin aggregated at the membrane adjacent to the site of bead attachment. In addition, an internal pool of integrin-positive vesicles accumulated. The mAb ES238 directed against the extracellular domain of the avian beta 1 integrin subunit, when coupled to beads, also induced the aggregation of talin at the membrane, whereas ES186 directed against the intracellular domain of the beta 1 integrin subunit did not. Cells attached and spread on Con A beads, but neither integrin nor talin aggregated at the membrane. After 30 min, when many of the cells were at a more advanced stage of spreading around beads or phagocytosing beads, alpha-actinin and actin, but not vinculin, form distinctive aggregates at sites along membranes associated with either fibronectin or Con A beads. Normal cells also rapidly formed aggregates of integrin and talin after binding to immobilized fibronectin in a manner that was similar to the transformed cells, suggesting that the aggregation process is not dependent upon activity of the pp60v-src tyrosine kinase. Thus, the binding of cells to immobilized fibronectin caused integrin-talin coaggregation at the sites of membrane-ECM contact, which can initiate the cytoskeletal events necessary for cell adhesion and spreading.     

(-)-Epigallocatechin-3-gallate, a polyphenolic compound from green tea, inhibits fibroblast adhesion and migration through multiple mechanisms

It is increasingly evident that the stromal cells are involved in key metastatic processes of melanoma and some malignant solid tumors. (-)-Epigallocatechin-3-gallate (EGCG), a polyphenolic compound from green tea, has been shown to have anti-tumor activity, inhibiting adhesion, migration, and proliferation of tumor cells. However, little attention has been paid on its effects on stromal cells. In the present study, we determined the effects of EGCG on stromal fibroblasts. We showed that fibroblast adhesion to collagen, fibronectin, and fibrinogen were inhibited by EGCG. One of the possible mechanisms is binding of EGCG to fibronectin and fibrinogen but not to collagen. We then focused how EGCG affected fibroblast adhesion to collagen. EGCG treatment attenuated the antibody binding to fibroblast's integrin alpha2beta1, indicating EGCG may affect the expression and affinity of integrin alpha2beta1. Moreover, intracellular H2O2 level was decreased by EGCG treatment, suggesting that the tonic maintenance of intracellular H2O2 may be required for cell adhesion to collagen. In parallel, collagen-induced FAK phosphorylation, actin cytoskeleton reorganization in fibroblasts, migration and matrix metalloproteinase(s) (MMPs) activity were also affected by EGCG. Tubular networks formed by melanoma cells grown on three-dimensional Matrigel were also disrupted when fibroblasts were treated with EGCG in a non-contact coculture system. Taken together, we provided here the first evidence that EGCG is an effective inhibitor on behaviors of the stromal fibroblasts, affecting their adhesion and migration. The inhibitory activity of EGCG may contribute to its anti-tumor activity. The findings and concepts disclosed here may provide important basis for a further experiment towards understanding tumor-stroma interaction.     

The extracellular matrix ligands fibronectin and tenascin collaborate in regulating collagenase gene expression in fibroblasts.

Tenascin (TN) is a large oligomeric glycoprotein that is present transiently in the extracellular matrix (ECM) of cells and is involved in morphogenetic movements, tissue patterning, and tissue repair. It has multiple domains, both adhesive and anti-adhesive, that interact with cells and with fibronectin (FN) and other ECM macromolecules. We have studied the consequences of the interaction of TN with a FN matrix on gene expression in rabbit synovial fibroblasts. Fibroblasts plated on a mixed substrate of FN and TN, but not on FN alone, upregulated synthesis of four genes: collagenase, stromelysin, the 92-kDa gelatinase, and c-fos. Although the fibroblasts spread well on both FN and FN/TN substrates, nuclear c-Fos increased within 1 h only in cells that were plated on FN/TN. TN did not induce the expression of collagenase in cells plated on substrates of type I collagen or vitronectin (VN). Moreover, soluble TN added to cells adhering to a FN substrate or to serum proteins had no effect, suggesting that TN has an effect only in the context of mixed substrates of FN and TN. Collagenase increased within 4 h of plating on a FN/TN substrate and exhibited kinetics similar to those for induction of collagenase gene expression by signaling through the integrin FN receptor. Arg-Gly-Asp peptide ligands that recognize either the FN receptor or the VN receptor and function-perturbing anti-integrin monoclonal antibodies diminished the interaction of fibroblasts with a mixed substrate of FN, TN, and VN, but had no effect on the adhesion of fibroblasts to a substrate of FN and VN, suggesting that both receptors recognize the complex. Anti-TN68, an antibody that recognizes an epitope in the carboxyl-terminal type III repeats involved in the interaction of TN with both FN and cells, blocked the inductive effect of the FN/TN substrate, whereas anti-TNM1, an antibody that recognizes an epitope in the amino-terminal anti-adhesive region of epidermal growth factor-like repeats, had no effect. These data suggest that transient alteration of the composition of ECM by addition of proteins like TN may regulate the expression of genes involved in cell migration, tissue remodeling, and tissue invasion, in regions of tissue undergoing phenotypic changes.     

A gene for rabbit synovial cell collagenase: member of a family of metalloproteinases that degrade the connective tissue matrix

We have determined the nucleotide sequence of a collagenase mRNA from rabbit synovial cells from which the primary structure of the encoded protein was deduced. This proteinase is 51% homologous to the enzyme that activates it from the zymogen form, rabbit synovial cell activator/stromelysin. Rabbit collagenase and activator/stromelysin thus share comembership in a gene family that includes human skin collagenase; the human and rabbit metalloproteinase, activator/stromelysin; and an oncogene-induced proteinase from rat named transin. The mRNA sequence of collagenase enabled us to completely map the structure of its gene, which is 9.1 kilobases and is composed of 10 exons and 9 introns. This is the first report of the structure of a collagenase gene. We show that it has striking similarity to additional members of this metalloproteinase gene family, transin genes I and II of rat. We have further sequenced genomic DNA flanking the collagenase gene and have identified nucleic acid elements of possible importance in gene regulation.     

Membrane-type 1 matrix metalloproteinase modulates focal adhesion stability and cell migration

Membrane-type 1 matrix metalloproteinase (MT1-MMP) plays an important role in extracellular matrix-induced cell migration and the activation of extracellular signal-regulated kinase (ERK). We showed here that transfection of the MT1-MMP gene into HeLa cells promoted fibronectin-induced cell migration, which was accompanied by fibronectin degradation and reduction of stable focal adhesions, which function as anchors for actin-stress fibers. MT1-MMP expression attenuated integrin clustering that was induced by adhesion of cells to fibronectin. The attenuation of integrin clustering was abrogated by MT1-MMP inhibition with a synthetic MMP inhibitor, BB94. When cultured on fibronectin, HT1080 cells, which endogenously express MT1-MMP, showed so-called motile morphology with well-organized focal adhesion formation, well-oriented actin-stress fiber formation, and the lysis of fibronectin through trails of cell migration. Inhibition of endogenous MT1-MMP by BB94 treatment or expression of the MT1-MMP carboxyl-terminal domain, which negatively regulates MT1-MMP activity, resulted in the suppression of fibronectin lysis and cell migration. BB94 treatment promoted stable focal adhesion formation concomitant with enhanced phosphorylation of tyrosine 397 of focal adhesion kinase (FAK) and reduced ERK activation. These results suggest that lysis of the extracellular matrix by MT1-MMP promotes focal adhesion turnover and subsequent ERK activation, which in turn stimulates cell migration.     

Discoordinate expression of stromelysin, collagenase, and tissue inhibitor of metalloproteinases-1 in rheumatoid human synovial fibroblasts. Synergistic effects of interleukin-1 and tumor necrosis factor-alpha on stromelysin expression.

Primary and passaged human synovial fibroblasts isolated from rheumatoid pannus were treated with recombinant interleukin-1 (IL-1) alpha or beta, tumor necrosis factor-alpha (TNF), or phorbol myristate acetate (PMA) to determine the effects of these stimuli on the relative expression of stromelysin, collagenase, and tissue inhibitor of metalloproteinases (TIMP). The steady-state mRNA levels for these genes and glyceraldehyde-3-phosphate dehydrogenase were determined on Northern blots. Immunoblot analyses of the conditioned media using monoclonal antibodies generated against recombinant human stromelysin, collagenase, or TIMP showed that protein levels reflected the corresponding steady-state mRNA levels. The results revealed that 1) stromelysin and collagenase were not always coordinately expressed; 2) IL-1 was more potent than TNF or PMA in the induction of stromelysin expression; 3) neither IL-1 nor TNF significantly affected TIMP expression; 4) PMA induced both metalloproteinase and TIMP expression; and 5) the combination of IL-1 plus TNF had a synergistic effect on stromelysin expression. Dose response and time course experiments demonstrated that the synergistic effect of IL-1 plus TNF occurred at saturating concentrations of each cytokine and lasted for 7 days. In summary, the ability of IL-1 and TNF to preferentially induce stromelysin and collagenase expression, versus TIMP, may define a pivotal role for these cytokines in the pathogenesis of rheumatoid arthritis.     

Commitment to expression of the metalloendopeptidases, collagenase and stromelysin: relationship of inducing events to changes in cytoskeletal architecture

Agents that alter the morphology of rabbit synovial fibroblasts induce synthesis of the metalloendopeptidases, collagenase and stromelysin. We studied the relationship of cytoskeletal changes to the commitment to expression of these metalloendopeptidases. Cells treated with cytochalasin B (CB) or 12-O-tetradecanoylphorbol-13-acetate rounded, and only cells that had lost their stress fibers expressed collagenase and stromelysin, as determined by immunofluorescence. We concentrated on the effects of CB because of its rapid reversibility. When CB was added for 1-24 h, then removed, the cells respread within 30-60 min. The minimum period of CB treatment that committed cells to the subsequent synthesis of collagenase and stromelysin was 3 h. After initial treatment with 2 micrograms/ml CB for 3-24 h, or with various concentrations of CB (0-2 micrograms/ml) for 24 h, both enzyme activity and biosynthesis of the proenzymes showed a graded increase when measured at 24 h. Even after treatment with 2 micrograms/ml CB for only 3 h, greater than 85% of all cells were positive for both collagenase and stromelysin when cells were monitored by immunofluorescence. In contrast, when the dependence of collagenase and stromelysin expression on the inducing concentration of CB was examined, there was a dose-dependent increase in the number of cells positive for collagenase and stromelysin, as determined by immunofluorescence. Thus, at low concentrations of CB (less than 0.5 micrograms/ml), a heterogeneous population response was observed. These results suggest that the commitment of fibroblasts to induction of the metalloproteinases is a stochastic process in which a second signal that correlates with the disruption of the actin cytoskeleton may be rate-limiting for collagenase and stromelysin gene expression.     

The Fibrillin-1 RGD Integrin Binding Site Regulates Gene Expression and Cell Function through microRNAs

Fibrillins are the major components of microfibrils in the extracellular matrix of elastic and non-elastic tissues. Fibrillin-1 contains one evolutionarily conserved RGD sequence that mediates cell–matrix interactions through cell-surface integrins. Here, we present a novel paradigm how extracellular fibrillin-1 controls cellular function through integrin-mediated microRNA regulation. Comparative mRNA studies by global microarray analysis identified growth factor activity, actin binding and integrin binding as the most important functional groups that are regulated upon fibrillin-1 binding to dermal fibroblasts. Many of these mRNAs are targets of miRNAs that were identified when RNA from the fibrillin-1-ligated fibroblasts was analyzed by a miRNA microarray. The expression profile was specific to fibrillin-1 since interaction with fibronectin displayed a partially distinct profile. The importance of selected miRNAs for the regulation of the identified mRNAs was suggested by bioinformatics prediction and the interactions between miRNAs and mRNAs were experimentally validated. Functionally, we show that miR-503 controls p-Smad2-dependent TGF-β signaling, and that miR-612 and miR-3185 are involved in the focal adhesion formation regulated by fibrillin-1. In conclusion, we demonstrate that fibrillin-1 interaction with fibroblasts regulates miRNA expression profiles which in turn control critical cell functions.     

H-ras oncogene-transformed human bronchial epithelial cells (TBE-1) secrete a single metalloprotease capable of degrading basement membrane collagen

H-ras-transformed human bronchial epithelial cells (TBE-1) secrete a single major extracellular matrix metalloprotease which is not found in the normal parental cells. The enzyme is secreted in a latent form of 72 kDa, which can be activated to catalyze the cleavage of the basement membrane macromolecule type IV collagen. The substrates in their order of preference are: gelatin, type IV collagen, type V collagen, fibronectin, and type VII collagen; but the enzyme does not cleave the interstitial collagens or laminin. This protease is identical to gelatinase isolated from normal human skin explants, normal human skin fibroblasts, and SV40-transformed human lung fibroblasts. Based on its ability to initiate the degradation of type IV collagen in a pepsin-resistant portion of the molecule, it will be referred to as type IV collagenase. This enzyme is most likely the human analog of type IV collagenase detected in several rodent tumors, which has the same molecular mass and has been linked to their metastatic potential. Type IV collagenase consists of three domains. Two of them, the amino-terminal domain and the carboxyl-terminal domain, are homologous to interstitial collagenase and human and rat stromelysin. The middle domain, of 175 residues, is organized into three 58-residue head-to-tail repeats which are homologous to the type II motif of the collagen-binding domain of fibronectin. Type IV collagenase represents the third member of a newly recognized gene family coding for secreted extracellular matrix metalloproteases, which includes interstitial fibroblast collagenase and stromelysin.     

Internalization of the fibronectin receptor is a constitutive process

Using a monoclonal antibody specific for the hamster fibronectin receptor (FnR), we have demonstrated that a portion of the CHO cell FnR population is constitutively endocytosed. Three independent techniques were used to study the internalization: 1) after saturation binding of an anti-FnR antibody (PB1) to cells at 4 degrees C, internalization was initiated by warming to 37 degrees C, and then acid/salt elution of membrane-bound ligand was used to quantitate the internalized 125I-PB1; 2) cell vesicular traffic was pharmacologically disrupted with monensin or chloroquine, and the subsequent reduction of the cell surface pool of FnR was monitored; and 3) selective immunoprecipitation was used to separate surface and internalized 125I-labeled FnR. These experiments indicate that about 30% of the cell surface FnR is endocytosed with a t1/2 of 7 min and that this internalization occurs regardless of the ligation state of the receptor. Other observations indicate that the larger fraction of the cell surface FnR pool (70-75%) is apparently shed from the cell upon ligation with antibody at 37 degrees C. This process occurs much more slowly than receptor internalization and leads to an overall reduction in the amount of cell surface FnR. Our results suggest physically or chemically distinct populations of FnR, one of which is unavailable for internalization and recycling.     

Kinetic and morphological evidence for endocytosis of mammalian cell integrin receptors by using an anti-fibronectin receptor beta subunit monoclonal antibody

Monoclonal antibody (mAb) 7E2.2, which recognizes the beta subunit of the hamster fibronectin receptor (FnR) (Brown, P.J., and Juliano, R. L. (1988) Exp. Cell Res. 177. 303), was used to examine the distribution of and to quantify the internalization of the FnR and possibly related integrins on adherent fibroblasts. Purified 7E2.2 IgG was iodinated and used in binding and internalization studies. Binding to Chinese hamster ovary cells was saturable with a Km of 0.3 nM and an estimated total number of cell surface beta subunits at 2 x 10(5) per cell. The FnR colocalized with fibronectin at cell adhesion contact sites and also was distributed evenly over the dorsal cell surface as discrete clusters. By using a direct immunocolloidal gold approach, the FnR was not associated with coated pits at 4 degrees C until internalization followed warming of the labeled cells to 37 degrees C. A proportion of the FnRs were endocytosed with a half-time of 6.5 min and, consistent with clathrin-mediated uptake, this was sensitive to hypertonic conditions. Receptor-immunocomplexes rapidly became localized within coated pits, small diameter tubules, and peripheral endosomes but the majority remained at the cell surface. At subsaturating concentrations of bound 7E2.2, approximately one-fourth of the total cell receptor population resided intracellularly at any one moment following steady-state; however, appreciable degradation of the iodinated mAb was not detected following accumulation for 4 h at 37 degrees C. These data showed that at least a portion of the FnR are endocytosed via a receptor-mediated pathway and suggested that these receptors do not immediately enter a degradative compartment.     

Anti-integrin antibodies induce type IV collagenase expression in keratinocytes

During wound healing, pericellular proteolysis is thought to be essential for the detachment of keratinocytes from basement membrane and in their migration into the wound bed. We have characterized integrin-type cell adhesion/migration receptors in human mucosal keratinocytes and examined their function in the regulation of type IV collagenase gene expression. Two major integrins of the β1 class, α2β1 and αβ1, were found to function as collagen and fibronectin receptors, respectively. Antibodies against β1 and α3 integrin subunits were found to stimulate the expression of the 92 kDa type IV collagenase severalfold in a dosedependent manner. Keratinocytes expressed also the 72 kDa type IV collagenase, the synthesis of which remained, however, unchanged in keratinocytes treated with anti-integrin antibodies. Stimulation of 92 kDa enzyme was found to be caused directly by antibody binding to integrins, since Fab-fragments of anti-β1 antibodies alone were able to induce collagenase expression in the absence of secondary, clustering antibodies. Antibodies against α2β1 integrin caused no stimulation. Keratinocytes seeded on different substrata (plastic, collagen, fibronectin, laminin, or vitronectin) showed equal induction of type IV collagenase expression. Expression of 92 kDa type IV collagenase could not be induced by peptides (GRGDS, GRGES), proteins (fibronectin, laminin, fibrinogen., albumin), or antibodies to fibronectin. We suggest that proteolytic processes around keratinocytes can be regulated by extracellular factors signalling through integrin-type receptors. © 1993 Wiley-Liss, Inc.