Codistribution of heparan sulfate proteoglycan, laminin, and fibronectin in the extracellular matrix of normal rat kidney cells and their coordinate absence in transformed cells

We used antibodies raised against both a heparan sulfate proteoglycan purified from a mouse sarcoma and a chondroitin sulfate proteoglycan purified from a rat yolk sac carcinoma to study the appearance and distribution of proteoglycans in cultured cells. Normal rat kidney cells displayed a fibrillar network of immunoreactive material at the cell surface when stained with antibodies to heparan sulfate proteoglycan, while virally transformed rat kidney cells lacked such a surface network. Antibodies to chondroitin sulfate proteoglycan revealed a punctate pattern on the surface of both cell types. The distribution of these two proteoglycans was compared to that of fibronectin by double-labeling immunofluorescent staining. The heparan sulfate proteoglycan was found to codistribute with fibronectin, and fibronectin and laminin gave coincidental stainings. The distribution of chondroitin sulfate proteoglycan was not coincidental with that of fibronectin. Distinct fibers containing fibronectin but lacking chondroitin sulfate proteoglycan were observed. When the transformed cells were cultured in the presence of sodium butyrate, their morphology changed, and fibronectin, laminin, and heparan sulfate proteoglycan appeared at the cell surface in a pattern resembling that of normal cells. These results suggest that fibronectin, laminin, and heparan sulfate proteoglycan may be complexed at the cell surface. The proteoglycan may play a central role in assembly of such complexes since heparan sulfate has been shown to interact with both fibronectin and laminin.     

Concomitant loss of cell surface fibronectin and laminin from transformed rat kidney cells

Both fibronectin and laminin were found by immunofluorescence as a matrix at the surface of normal rat kidney cells. These matrices were absent from the surface of virally transformed rat kidney cells. Soluble fibronectin and laminin were detected in the culture media of the transformed as well as the normal cells. Culture supernates of the transformed cells contained even more fibronectin than the supernates of the transformed cells contained even more fibronectin than the supernates of the normal cells while laminin was present in similar amounts in both culture media. This shows that the loss of fibronectin and laminin from the surface of the transformed cells is caused by failure of the cells to deposit these proteins into an insoluble matrix and not caused by inadequate production. Fibronectins isolated from culture media of the normal and transformed cells were similar in SDS polyacrylamide gel electrophresis. Laminin isolated from culture media by affinity chromatography on heparin-Sepharose followed by immunoprecipitation was composed of three main polypeptides, one with a molecular weight of 400,000 and two with a molecular weight close to 200,000 in both cell types. Fibronectins from both cell types were equally active in promoting cell attachment. Rat fibronectin from transformed cells, like normal cells, when applied to culture dishes coated with fibronectin, readily attached and spread on the substratum, requiring approximately the same amount of fibronectin as the normal cells. On the basis of these results it seem that the failure of the transformed cells to incorporate fibronectin into an insoluble cell surface matix is not a consequence of a demonstrable change in the functional characteristics of the fibronectin molecule or in the ability of the cells to interact with fibronectin. It may depend on as yet unidentified interactions of the cell surface. Similar interactions may be needed for the deposition of laminin into the matrix, because laminin was also absent from the surface of transformed cells, despite its being synthesized by these cells.     

Laminin induces the stable expression of surface galactosyltransferase on lamellipodia of migrating cells

We have previously shown that cell surface galactosyltransferase (GalTase) mediates cell spreading and migration on basal lamina matrices by binding N-linked oligosaccharide substrates within laminin. In this study we have examined the distribution and expression of cell surface GalTase during mesenchymal cell migration on various extracellular matrices. Antisera raised against affinity-purified beta 1,4 GalTase, as well as anti-GalTase Fab fragments, inhibited cell migration on laminin-containing matrices, whereas under identical conditions, anti-GalTase IgG had no effect on the rate of cell migration on fibronectin substrates. Cells migrating on laminin had three times the level of surface GalTase, assayed by 125I-antibody binding and by direct enzyme assay, than similar cells migrating on fibronectin. On the other hand, total cellular GalTase, assayed either enzymatically or by Northern blot analysis, was similar when cells were grown on laminin or fibronectin. The laminin-dependent increase in surface GalTase was due to its expression onto the leading and trailing edges of migrating cells in association with actin-containing microfilaments assayed by double-label indirect immunofluorescence. On stationary cells, surface GalTase levels were low, but as cells began to migrate on laminin GalTase became polarized to the growing lamellipodia. GalTase was not detectable on lamellipodia or filopodia when cells migrated on fibronectin substrates. These results show that laminin-containing matrices induce the stable expression of GalTase onto cell lamellipodia and filopodia where it mediates subsequent cell spreading and migration. Since fibronectin was unable to induce GalTase expression onto lamellipodia, these studies also suggest that the extracellular matrix can selectively influence which intracellular components are maintained on the cell surface.     

Production of laminin and fibronectin by Schwannoma cells: cell-protein interactions in vitro and protein localization in peripheral nerve in vivo

We studied a rat Schwannoma cell line (RN22F) to determine if it produced the basement membrane glycoproteins laminin and fibronectin, and how it interacted with these proteins in vitro. We used antisera to laminin and fibronectin for immunoprecipitation experiments and immunocytochemical localization at the electron microscope level. Polyacrylamide gels of antilaminin immunoprecipitates of conditioned medium and solubilized Schwannoma cells contained bands of reduced Mr 200,000 and 150,000. Antilaminin immunoprecipitates of conditioned medium contained nonreduced bands of 850,000 daltons and 150,000, and immunoprecipitates of solubilized cells contained nonreduced bands of 850,000, 400,000, 200,000, and 150,000 daltons. Antifibronectin immunoprecipitates of conditioned medium contained a reduced band of 220,000 daltons, and nonreduced bands of 440,000 and 220,000 daltons. Radio-labeled protein was not detected in antifibronectin immunoprecipitates of solubilized cells. By immunocytochemistry, laminin was found along the cell surface in a continuous band, whereas fibronectin was only sparsely distributed along the cell surface. In cell adhesion assays, Schwannoma cells bound preferentially to laminin- coated substrates as compared to fibronectin or noncoated substrates. A number of Schwannoma cells displayed a curved and elongated morphology on laminin substrates, as compared to a uniformly spread morphology on fibronectin, and a round, nonspread morphology on noncoated substrates. Immunofluorescent staining showed laminin in the endoneurium and perineurium and fibronectin predominantly in the perineurium of mouse sciatic nerve in vivo. The production of laminin and fibronectin by Schwann cells may be important in the development and myelination of peripheral nerves, and the proper regeneration of axons following nerve injury.     

Expression and distribution of very late antigen-5 in mouse peritoneal macrophages upon ingestion of fibronectin-bound Staphylococcus aureus

Many pathogens colonize host tissues by binding to the extracellular matrix via their cell surface adhesion molecules, which are called MSCRAMMs (microbial surface components recognizing adhesive matrix molecules). Staphylococcus aureus expresses several of these adhesion molecules, some of which bind to fibronectin. Of these adhesion molecules, fibronectin-binding proteins play a role in the pathogenicity of S. aureus, although it is not yet clear whether they enhance its virulence. We have previously shown that fibronectin-bound S. aureus is efficiently phagocytosed by thioglycolate-induced mouse peritoneal macrophages. Bacterial ingestion is mediated by Very Late Antigen-5 (VLA-5; alpha5beta1 integrin) and is accompanied by the formation of adhesion complexes. Here we show that the expression of VLA-5 is restricted to thioglycolate-induced inflammatory macrophages and is not found in the resident macrophages. When cells were in suspension, alpha5 integrin was not expressed on the surface of either resident or inflammatory macrophages, whereas in adherent cells, this integrin was distributed on the surface of inflammatory but not resident macrophages. A high level of this integrin was present in the cytoplasmic region only in inflammatory macrophages. In agreement with this, fibronectin-mediated phagocytosis of S. aureus was observed only in the inflammatory macrophages. In inflammatory macrophages ingesting fibronectin-bound S. aureus, alpha5 integrin was concentrated close to the phagocytosed bacteria. This change in distribution was not found in macrophages ingesting untreated bacteria. Together with our previous work, these results indicate that, upon ingestion of fibronectin-bound S. aureus, VLA-5 accumulates in the area of phagocytosis in inflammatory macrophages, where it forms adhesion complexes.     

Sialilated beta1,6 branched N-oligosaccharides modulate adhesion, chemotaxis and motility of melanoma cells: Effect on invasion and spontaneous metastasis properties

B16BL6 cells, selected specifically for invasive characteristics from B16F10 mouse melanoma cells, displayed greater ability to metastasize to lungs and produced larger colonies than the parent cells, when injected intravenously. When the two cell lines were compared for surface beta1,6-branched N-oligosaccharides by flow cytometry using Leuco-Phyto-Heam-Agglutinin, B16BL6 were found to express significantly higher levels. Inhibition of the oligosaccharide expression, by treatment of the cells with swainsonine or antisense-N-acetyl glucosaminyl-transferase-V, significantly reduced metastasis and invasion (>50%). Further, inhibition of oligosaccharides on the molecules like beta1 integrin (one of the major carriers) caused 30-45% reduction in their adherence to extra-cellular-matrix components especially collagen IV and laminin, and chemotaxis towards fibronectin and matrigel. The inhibition also decreased haptotaxis by approximately 50% to fibronectin but surprisingly was enhanced towards laminin by approximately 75%. The cells on which the expression of these oligosaccharides was inhibited failed to exhibit the characteristic spontaneous metastasis and adhesion properties of B16BL6 cells. In none of the cases, however, the secretion of matrix-metallo-proteases correlated with oligosaccharide expression. Sialylation of surface oligosaccharides was found to be accompanied by even higher motility and adherence to the substrates. These results strongly support an important role of cell surface beta1,6-linked N-oligosaccharides, especially the sialylated derivatives, in the processes that influence invasion and metastasis.     

Colocalization of laminin and fibronectin in bovinelens epithelial cells in vitro

Summary The distribution and organization of the extracellular matrix (ECM) proteins laminin, fibronectin, entactin, and type IV collagen were investigated in primary colonies and secondary cultures of bovine lens epithelial cells using species-specific antisera and indirect immunofluorescence microscopy. Primary cell colonies fixed in formaldehyde and permeabilized with Triton X-100 displayed diffuse clonies. In contrast, thick bundles of laminin and fibronectin were located on the basal cellsurfaces and in between cells in the densely packed center of the colonies, and as “adhesive plaques” and fine extracellular matrix cords in the sparsely populated (migratory) outer edge of the colonies. The distribution of ECM proteins observed in secondary lens epithelial cell cultures was similar to that observed at the periphery of the primary colony. Extraction of the secondary cell cultures with sodium deoxycholate confirmed that laminin and fibronectin were deposited on the basal cell surface. Indeed, the patterns of laminin and fibronectin deposition suggested that these proteins codistribute. These results establish that lens epithelial cells in culture can be used as a model system to study the synthesis and extracellular deposition of the basement membrane proteins, laminin and fibronectin. Supported by Public Health Service grant EY05570 from the National Eye Institute Bethesda, MD.     

Macrophage cell cycling: influence of proliferative state on the antibody-mediated activities of rat resident peritoneal macrophages

Countercurrent centrifugal elutriation (CCE) was used to isolate fractions of rat resident peritoneal macrophages that were enriched in different phases of the cell cycle. The purpose was to assess the influence of the proliferative status of these cells on their antibody-mediated phagocytic activity. Autoradiographic analysis of the resident peritoneal cell population isolated 1 hr after an intravenous injection of [3H] thymidine showed that about 3% of the macrophages were in S-phase of the cell cycle. CCE yielded fractions of macrophages in which the proportions of S-phase cells ranged from 0% to about 10%. Results of flow cytometric analysis of propidium iodine-stained cells were consistent with the autoradiographic findings. Essentially all of the macrophages in the CCE fractions ingested antibody-coated particles, but there were marked differences in phagocytic capacity and in expression of Fc-receptors among discrete groups of cells. CCE fractions with the smallest cells and no S-phase macrophages ingested approximately six- to eightfold fewer particles than did macrophages from CCE fractions with the largest cells and enriched in S-phase macrophages. Similarly, smaller macrophages bound fewer antibody-coated particles than did larger macrophages. These results, which are identical to those previously reported for murine macrophage cell lines, show that the number of Fc-receptors and the phagocytic capacity of cycling resident peritoneal macrophages increase as the cells progress from G1 to G2. Thus, the proliferative state of macrophages does not determine whether they are phagocytic but rather their phagocytic capacity.     

Rat Macrophage C-Type Lectin Is an Activating Receptor Expressed by Phagocytic Cells

Macrophage C-type lectin (MCL) is a membrane surface receptor encoded by the Antigen Presenting Lectin-like gene Complex (APLEC). We generated a mouse monoclonal antibody for the study of this receptor in the rat. We demonstrate that rat MCL is expressed on blood monocytes and neutrophils, as well as on several tissue macrophage populations, including alveolar and peritoneal cavity macrophages. We also demonstrate MCL expression on a subset of resident spleen macrophages. Immunohistochemistry analysis of the spleen showed staining specifically in the marginal zone and red pulp. Exposure to pro-inflammatory mediators or to yeast cell wall extract (zymosan) increased surface MCL expression on peritoneal macrophages. We characterized a rat myeloid cell line, RMW, which expresses high levels of MCL. We found that MCL co-immunoprecipitated with the activating adaptor protein FcεRIγ in these cells. Moreover, beads coated with anti-MCL antibody increased phagocytosis in the RMW cells. Together, these observations indicate that rat MCL is a receptor that activates phagocytosis in myeloid cells under inflammatory conditions.     

Mononuclear phagocyte adherence in the presence of laminin : A possible marker of cellular differentiation

The effect of laminin on the vitro adhesive behavior of mononuclear phagocytes was investigated. Laminin significantly inhibited the adhesion of guinea pig, mouse, and rat alveolar or peritoneal macrophages and of human peripheral blood monocytes. Adhesion of these cells was unaffected by similar concentrations of fibronectin. Experiments performed with monocytes maintained in culture showed that the degree of laminin-mediated inhibition of adherence was dependent on the state of differentiation of the cells: the less mature the monocytes, the greater the degree of inhibition. Laminin also reduced the attachment capacity of polymorphonuclear leukocytes isolated from human peripheral blood. These results suggest a possible role for laminin in the regulation of the passage of cells across the basement membrane during inflammation.     

Macrophage interactions with laminin: PMA selectively induces the adherence and spreading of mouse macrophages on a laminin substratum

The ability of thioglycollate (TG)-elicited mouse peritoneal macrophages to adhere to a laminin substratum has been studied. These cells do not adhere to laminin-coated (20 micrograms/ml) surfaces, but the addition of phorbol myristate acetate (PMA; 50 ng/ml) results in their rapid adherence and spreading on this substratum. TG-elicited and PMA-activated macrophages, however, can bind soluble laminin. Macrophages adhere to fibronectin-coated surfaces and tissue culture plastic without PMA stimulation, and PMA does not increase the number of cells that adhere to these surfaces. The predominant surface proteins that bind specifically to laminin-Sepharose exhibit an Mr of 67 and 36 kD, but the expression of these proteins does not increase after PMA stimulation. Laminin receptor antibodies immunoprecipitate the 67-kD protein from radiolabled surface lysates and are capable of blocking macrophage adherence to a laminin substratum. Indirect immunofluorescence microscopy indicates that PMA stimulation does not increase receptor expression, but that it may induce the aggregation of the receptor on the cell surface. PMA stimulation also promotes macrophage spreading and induces a reorganization of the actin cytoskeleton. Taken together, these data indicate the mechanism by which PMA promotes macrophage adherence to laminin does not involve increased 67-kD receptor surface expression, but that it is related to the changes in cytoskeletal and receptor surface organization that occur in response to PMA stimulation.     

Inducible expression of murine IP-10 mRNA varies with the state of macrophage inflammatory activity

We have examined the expression of inducible inflammatory genes in murine macrophages from different tissues and at different stages of inflammatory activity. Although i.v. administration of IFN-gamma (10,000 U/mouse) strongly induced expression of IP-10 mRNA in the adherent cell population of the spleen, thioglycollate-elicited peritoneal macrophages were essentially unresponsive at the same dose. In contrast, D3 mRNA was expressed in both cell populations. This differential sensitivity of IP-10 mRNA expression was not restricted to stimulation by IFN-gamma as it was also seen when LPS (25 micrograms/mouse) was administered i.v. Expression of JE and KC mRNA, which encode cytokines related to IP-10, were also differentially expressed in elicited peritoneal macrophages from mice injected with LPS. Differential sensitivity was at least partially related to the state of macrophage activation because IP-10 mRNA was highly inducible in resident but not thioglycollate-elicited peritoneal macrophages. The eliciting agent was also an important determinant because proteose-peptone-elicited peritoneal macrophages were nearly as sensitive as splenic macrophages with respect to expression of IP-10 mRNA. IFN-gamma treatment induced IP-10 and D3 mRNA rapidly and transiently with the same time course in the spleen. IP-10 mRNA was not induced by IFN-gamma in TG-elicited macrophages regardless of the time after treatment. This differential expression of IP-10 was a consequence of different concentration requirements for IFN-gamma in the two cell types; thioglycollate-elicited macrophages required five- to 10-fold more IFN-gamma than did resident cells to achieve comparable IP-10 mRNA levels whether the agent was provided in vitro or in vivo. Thus variable sensitivity for induction of IP-10 mRNA was a characteristic of the macrophage itself and was not mediated by other cellular or molecular elements present in the inflammatory peritoneal cavity. The reduced sensitivity to IFN-gamma or LPS for expression of IP-10, JE, and KC mRNA as compared with TNF-alpha or D3 mRNA suggests that this distinct pattern of regulation may be restricted to members of these two related cytokine gene families that exhibit cell-type specific chemoattractant activity.     

Activated macrophages require T cells for xenograft rejection under the kidney capsule

Transplantation of tissues from other species has been advocated as a way to overcome the extreme shortage of human donors. Rejection, however, remains a major hurdle for clinical xenotransplantation. Although activation of macrophages by T cells is critical for the cellular rejection of xenografts, what other important interactions between these two types of cells remain less defined. When we activated macrophages of immuno-deficient mice (SCID or Rag-/-) using interferon-gamma and lipopolysacharide, xenogeneic cells were rejected by activated macrophages in the peritoneal cavity (which has an abundance of resident macrophages), but were not rejected under the kidney capsule (which requires the recruitment of effectors). This difference between the two sites implies that activated macrophages are inefficient for self-recruitment to peripheral graft sites and that T cells may still be required for the process. To test this hypothesis further, immunodeficient mice that had received xenogeneic cells were infused with peritoneal exudate cells (containing activated macrophages and activated T cells) from preimmunized immunocompetent mice. Xenogeneic cells at both the kidney capsule and peritoneal sites were rejected soon after cell transfer. However, when the exudate cells were transferred into SCID recipients that first had been injected with T cell depleting antibodies, xenograft rejection was only prominent at the peritoneal site but not kidney capsule site. These results argue that activated macrophages (as the result of T cell activation) still require T cells for xenograft rejection at peripheral sites.     

The activation dependent adhesion of macrophages to laminin involves cytoskeletal anchoring and phosphorylation of the alpha 6 beta 1 integrin

Macrophages require activation with either PMA (Mercurio, A. M., and L. M. Shaw. 1988. J. Cell Biol. 107:1873-1880) or interferon-gamma (Shaw, L. M., and A. M. Mercurio. 1989. J. Exp. Med. 169:303-308) to adhere to a laminin substratum. In the present study, we identified an integrin laminin receptor on macrophages and characterized cellular changes that occur in response to PMA activation that facilitate laminin adhesion. A monoclonal antibody (GoH3) that recognizes the integrin alpha 6 subunit (Sonnenberg, A., H. Janssen, F. Hogervorst, J. Calafat, and J. Hilgers. 1987. J. Biol. Chem. 262:10376-10383) specifically inhibited adhesion to laminin-coated surfaces. This antibody precipitated an alpha 6 beta 1 heterodimer (Mr 130/110 kD) from 125I surface-labeled macrophages. The amount of radiolabeled receptor on the cell surface did not increase after PMA activation. Thus, the induction of laminin adhesion cannot be attributed to de novo or increased surface expression of alpha 6 beta 1. By initially removing the Triton X-100-soluble fraction of macrophages and then disrupting the remaining cytoskeletal framework, we observed that 75% of the alpha 6 beta 1 heterodimer on the cell surface is anchored to the cytoskeleton in macrophages that had adhered to a laminin substratum in response to PMA. Significant cytoskeletal anchoring of this receptor was not observed in macrophages that had adhered to fibronectin or tissue culture plastic, nor was it seen in nonadherent cells. PMA also induced phosphorylation of the cytoplasmic domain of the alpha 6 subunit, but not the beta 1 subunit. Phosphorylated alpha 6 was localized to the cytoskeletal fraction only in macrophages plated on a laminin substratum. In summary, our results support a mechanism for the regulation of macrophage adhesion to laminin that involves specific and dynamic matrix integrin-cytoskeletal interactions that may be facilitated by integrin phosphorylation.     

Molecular basis of macrophage activation. Expression of the low potential cytochrome b and its reduction upon cell stimulation in activated macrophages

The expression of the novel b-type cytochrome, which is part of the superoxide anion (O2-)-generating system in phagocytes, has been investigated in population of mouse peritoneal macrophages heterogeneous in their capability to produce O2-). Reduced minus oxidized difference spectra of intact cells showed the appearance of a b-type cytochrome with major peaks in the alpha region at 558 to 559 nm and in the gamma region at 426 to 428 nm. Resident peritoneal macrophages, as well as thioglycollate broth-elicited and Corynebacterium Parvum-activated macrophages and neutrophils expressed about 50 pmol cytochrome b/10(7) cells. In intact macrophages and neutrophils, Na-dithionite reduced greater than 75% of the cytochrome b measurable in disrupted cells. No correlation was found between capability to produce O2-) by different population of macrophages and their content of cytochrome b. When stimulated in strictly anaerobic conditions with phorbol myristic acetate, macrophages activated in vivo by i.p. injection of Corynebacterium Parvum reduced approximately 40% of their total cytochrome b. In resident peritoneal macrophages that produced five times lower amounts of O2-, cytochrome b reduction was instead undetectable. Potentiometric properties of cytochrome b was investigated in macrophage subcellular particles. Both resident and Corynebacterium Parvum-activated macrophages revealed the presence of b chromophores with very low potentials of -255 and -244 mV, respectively, whose content was not different in the two populations. These results show that resident and activated macrophages express the same amount of cytochrome b, but upon stimulation with PMA, activated macrophages recruit a higher number of cytochrome b molecules in parallel with an enhanced production of O2-.     

Binding of calmodulin to the microfilament network correlates with induction of a macrophage tumoricidal response

Induction of mouse peritoneal macrophage cytotoxicity against SV3T3, a line of virally transformed mouse cells correlated with the distribution of cytoplasmic calmodulin in the macrophages. The organization of the cytoskeleton was examined by fluorescent microscopy and by transmission electron microscopy, using immunogold tagging after Triton-X-100 (TX-100) extraction of the macrophages. Macrophages that had been activated to a tumoricidal state in vivo by vaccinia virus or in vitro by lymphokine stimulation displayed cytoskeletal networks that were more extended and weblike than did resident macrophages. The organization of microfilaments and microtubules in the cytoskeleton was displayed by using either anti-actin or anti-tubulin. Immunogold labeling of tumoricidal macrophage cytoskeletons with anti-calmodulin revealed strong binding to the microfilament network and no binding to microtubules. Anti-calmodulin reacted weakly with the cytoskeletal network of resident macrophages, and this was not demonstrably greater than the reaction with normal sheep serum. However, resident macrophages displayed a high density of calmodulin (CAM) associated with unidentifiable structures in the perinuclear region when reacted with anti-calmodulin. These characteristic distributions of CAM in resident and activated macrophages was confirmed by immunofluorescence. The total and cytoskeletal-associated amounts of calmodulin per unit of protein were determined by radioimmune assay and 125I labeling followed by SDS-PAGE. No statistically significant differences were detected between resident and activated macrophages in either the total cell or cytoskeleton fractions. In summary, our results suggest that induction of tumoricidal activity of mouse peritoneal macrophages correlates with the translocation of calmodulin to the microfilament network of the cytoskeleton.     

Alterations in integrin receptor expression on chemically transformed human cells: specific enhancement of laminin and collagen receptor complexes

The abilities of malignant tumor cells to bind and migrate through basement membranes are important steps in invasion and metastasis. Malignant tumor cells would therefore be expected to express receptors on their surfaces for basement membrane and stromal components, such as collagens, laminin, and fibronectin, although the pattern of expression of these receptors on the malignant cells may be different from that on their normal progenitors. We report here that chemically transformed tumorigenic human cells express an altered pattern of integrin receptors on their cell surfaces as compared with their untransformed nontumorigenic counterparts. Specifically, N-methyl-N'-nitro-N-nitrosoguanidine transformation of HOS cells into highly tumorigenic cells results in a significant specific increase in the expression of (in descending order of level of cell surface expression) the integrins alpha 6/beta 1, alpha 2/beta 1, and alpha 1/beta 1, which are receptors for laminin, collagens, and collagen type IV and laminin, respectively. The level of expression of two fibronectin receptor integrins, alpha 5/beta 1 and alpha 3/beta 1, are, however, unaltered, whereas the level of expression of vitronectin receptor integrin, alpha v/beta 3, is drastically reduced on the transformed cells. Consistent with the increased expression of laminin and collagen receptors and the decreased expression of vitronectin receptors on the transformed cells, these cells attached three- to fivefold more strongly to laminin and collagen but attached very poorly to vitronectin. The MNNG-HOS cells were also found to have a greater potential for invasion through reconstituted basement membrane, matrigel, the major components of which are laminin and type IV collagen. The invasion of both the HOS and MNNG-HOS cells was inhibited 45-50% by a polyclonal anti-fibronectin receptor antibody. However, although the invasion of HOS cells could be inhibited up to 75% by an anti-alpha 6 monoclonal antibody, a similar concentration of this antibody had no effect on the alpha 6-overproducing MNNG-HOS cells. A fivefold higher concentration of this antibody did result in partial inhibition of MNNG-HOS invasion. These data indicate a critical role for the alpha 6/beta 1 laminin receptor in the invasion of these cells through basement membranes and demonstrate that chemical transformation of nontumorigenic human cells to highly tumorigenic cells is associated with an altered pattern of integrin expression which may play a direct role in the increased capacity of these cells to bind and invade through basement membranes.     

Induction of extracellular matrix gene expression in normal human keratinocytes by transforming growth factor beta is altered by cellular differentiation

Changes in epithelial substrate have been related to the cellular capacity for proliferation and to changes in cellular behavior. The effect of TGF beta 1 on the expression of the basement membrane genes, fibronectin, laminin B1, and collagen alpha 1 (IV), was examined. Northern analysis revealed that treatment of normal human epidermal keratinocytes with 100 pM TGF beta 1 increased the expression of each extracellular matrix (ECM) gene within 4 h of treatment. Maximal induction was reached within 24 h after treatment. The induction of ECM mRNA expression was dose dependent and was observed at doses as low as 1-3 pM TGF beta 1. Incremental doses of TGF beta 1 also increased cellular levels of fibronectin protein in undifferentiated keratinocytes and resulted in increased secretion of fibronectin. Squamous-differentiated cultures of keratinocytes expressed lower levels of the extracellular matrix RNAs than did undifferentiated cells. Treatment of these differentiated cells with TGF beta 1 induced the expression of fibronectin mRNA to levels seen in TGF beta-treated, undifferentiated keratinocytes but only marginally increased the expression of collagen alpha 1 (IV) and laminin B1 mRNA. The increased fibronectin mRNA expression in the differentiated keratinocytes was also reflected by increased accumulation of cellular and secreted fibronectin protein. The inclusion of cycloheximide in the protocol indicated that TGF beta induction of collagen alpha 1 (IV) mRNA was signaled by proteins already present in the cells but that TGF beta required the synthesis of a protein(s) to fully induce expression of fibronectin and laminin B1 mRNA. The differential regulation of these genes in differentiated cells may be important to TGF beta action in regulating reepithelialization.     

Laminin B1 expression is required for laminin deposition into the extracellular matrix of PC12 cells.

The extracellular matrix of rat pheochromocytoma PC12 cells was shown by indirect immunofluorescence to consist of a network of fibronectin. The matrix did not contain laminin. The cells synthesized messenger RNA for fibronectin, laminin B2, and s-laminin but not for entactin or the B1 and A chains of laminin. Laminin B2 but not laminin B1 was detectable in the culture medium and in cell lysates. A full-length cDNA clone for the B1 chain of laminin was constructed in the plasmid p-444, which contains the neomycin-resistance marker and human beta-actin promoter. PC12 cells were transfected with this recombinant plasmid, and stable neomycin-resistant clones were isolated and characterized. Clones that synthesized laminin B1 messenger RNA were found to deposit a laminin-containing matrix. In many of these clones the deposition of the fibronectin matrix was greatly diminished. The laminin matrix was predominantly localized in the intercellular spaces forming a honeycomb pattern. The morphology of the laminin-synthesizing transfected cells was markedly different from the parental cells. The cells grew in tight clusters that were resistant to dissociating agents. It is concluded that the B1 chain of laminin contains information that is required for the formation of a stable laminin-containing extracellular matrix network either by interaction with cell surface receptors or other extracellular matrix components. Furthermore, expression of the laminin B1 gene may be a central regulatory point in determining extracellular matrix composition during embryogenesis.