Two distinct mechanisms for the interaction of cells with fibronectin substrata

Soluble or bound laminin elicit in human neuroblastoma cells short- or long-term potentiation of a K+ inwardly rectifying current: relevance to neuritogenesis

Changes in the resting potential (VREST) and in the underlying ionic conductances were measured by the patch-clamp technique in SH-SY5Y human neuroblastoma cells exposed to substrate-bound or soluble Laminin (bLN; sLN), as compared to integrin-independent substrates (polylysine (PL); bovine serum albumin (BSA)). While PL and BSA were ineffective, both forms of LN caused an early (5-15 min) activation of a peculiar type of Inwardly Rectifying K+ current (IIR) characterised by a voltage-dependent inactivation in the range of membrane potentials around -50/0 mV. IIR was blocked by Cs+ ions and by the antiarrhythmic drug E-4031, a specific inhibitor of the HERG-codified channels. In cells adherent to bLN, IIR potentiation (85%) persisted for 90-120 min and was accompanied by a similar, but transient, increase in the leakage conductance (GL). Successively, the persistence of a high IIR conductance and the decrease of GL progressively bring VREST from -12 to approximately -30 mV in about 120 min. On the other hand, in cells adherent to PL, exposure to sLN produced a similar but not persistent activation of IIR, without any increase in GL: this caused a rapid, transient hyperpolarisation of VREST. The effects of bLN and sLN were mimicked by antibodies raised against the integrin beta 1 subunit, suggesting a specific integrin-mediated mechanism. In fact, when bound to the culture dishes, these antibodies simultaneously activated the IIR and GL, whereas in soluble form they only activated IIR. Cells adherent to bLN emitted neurites, a process impaired by the block of IIR by E-4031 and Cs+. On the whole data suggest that the integrin-mediated activation of IIR plays a crucial role in the commitment to neuritogenesis of neuroblastoma cells, independently on the effects of this activation on VREST.

Effect of the conformation and orientation of adsorbed fibronectin on endothelial cell spreading and the strength of adhesion

The effect of surface hydrophobicity upon the conformation of the cell binding domain of fibronectin (Fn) and the influence of Fn conformation on bovine aortic endothelial cell (BAEC) adhesion were examined. The free sulfhydryl group of Fn located near the cell binding domain was selectively labeled with acrylodan, a polarity sensitive fluor. Fluorescence emission was monitored in solution and upon adsorption to hydrophilic glass and hydrophobic silanized glass. The acrylodan-labeled Fn emission maximum shifted to longer wavelengths upon adsorption and the shift was greater for acrylodan-labeled Fn adsorbed to hydrophilic glass than hydrophobic silane, suggesting that the acrylodan was in a more solvent accessible environment on glass than silane. BAEC, suspended in serum-free medium, attached for 15 or 120 min onto glass or silane surfaces containing preadsorbed Fn, after which cell spreading and the strength of adhesion in a parallel plate flow chamber were measured. Cell spreading was similar on both surfaces after 15 min attachment, but BAECs were more spread on glass than silane after 120 min. At low surface concentrations of Fn, BAECs were more adherent on glass than silane. At higher surfaces concentrations, adhesion was similar. After a 2-h incubation in serum-free medium, cells on glass showed more extensive development of focal contacts as determined by immunofluorescent staining for vinculin. Cell adhesion under flow was reduced on silane by inhibition of protein synthesis with cycloheximide, suggesting that cell attachment to silane was promoted by cellular synthesis of Fn. The results indicate that changes in the conformation of the Fn cell binding domain affect Fn affinity for its cell surface receptor.

Stable heparin-producing cell lines derived from the Furth murine mastocytoma

Stable cell lines that synthesize heparin have been established from the Furth murine mastocytoma. The parental line (MST) divides in suspension every 14-18 h in growth medium supplemented with fetal bovine serum or defined growth factors. Adherent subclones were selected by adhesion to plastic culture vessels. Both adherent and nonadherent cells contain about 0.4 micrograms of glycosaminoglycan hexuronic acid per 10(6) cells, composed of 80% heparin and 20% chondroitin sulfate E. Deaminative cleavage of MST heparin by HNO2 at pH 1.5 released disaccharides that were similar in composition to those obtained from commercial heparin, except that disaccharides containing 3,6-O-desulfated GlcN units were not found. Greater than 90% of the glycosaminoglycans were stored in cytoplasmic granules, and challenge of the cells with dinitrophenylated bovine serum albumin and anti-dinitrophenyl IgE released a portion of the stored material. Growth studies of subclones showed that MST cells tolerate a 10-fold variation in glycosaminoglycan content. Incubation of cells with sodium chlorate reduced glycosaminoglycan sulfation by > 95% without affecting cell growth. Thus, granule glycosaminoglycans appear to be nonessential for growth of MST cells.

Phorbol ester stimulation of fibronectin-mediated cell adhesion

Stimulation of Chinese hamster ovary cells with 12-O-tetradecanoyl 13-acetate increases the rate of adhesion to fibronectin-coated substratum. The EC50 of the phorbol ester that initiates the change in kinetics of adhesion is approximately 8 nM and is specific to those phorbol esters which activate protein kinase C. When compared to control cells, cells stimulated with active phorbol esters require a significantly lower amount of fibronectin to support their adhesion, and exhibit 50% adhesion inhibition by a log-fold higher concentration of PB1, a monoclonal antibody which specifically blocks fibronectin-mediated adhesion. These results indicate that stimulation of cells with phorbol esters results in a modification of the fibronectin receptor leading to an apparent increase in the interaction of the receptor with fibronectin.

Monoclonal antibodies to distinctive epitopes on the alpha and beta subunits of the fibronectin receptor

Monoclonal antibodies (MAbs) have been developed that can recognize epitopes that are unique to either the alpha or beta subunit of the fibronectin receptor (FnR). MAbs 11B4 and 7A8 immunoblot the alpha subunit of FnR either in purified form from Chinese hamster ovary (CHO) cells or in nonionic detergent extracts of cells of human and rodent origin electrophoresed under reducing or nonreducing conditions. The MAbs seem to be more reactive to the subunit when it has been electrophoresed under reducing conditions, suggesting that the epitope may be partially masked by the conformation conferred by disulfide bonding. A second set of MAbs, 7E2 and 7F9, is directed to an epitope on the beta subunit that is conformationally dependent upon disulfide bonding, as reduction of the subunit leads to loss of reactivity with both MAbs. Further, 7E2/7F9 immunoblots of nonionic detergent extracts of CHO cells, run under nonreducing conditions, reveal the presence of a third band (90-kDa), immunologically related to the beta subunit, which is not surface-labeled with 125I in intact cells and which does not copurify with the alpha and beta subunits isolated by immunoaffinity purification of FnR using the MAb PB1. The 90-kDa component is not found associated with a plasma membrane fraction prepared by crude cell fractionation, but is abundant in a low-speed pellet containing nuclei and intracellular membranes. This finding suggests that the 90-kDa component is a precursor to the beta subunit. Finally, the epitope of 7E2/7F9 is unique to CHO cells, as cross-reactivity to other cell types cannot be demonstrated by either immunoblotting or immunoprecipitation.

Adhesion of glycosaminoglycan-deficient chinese hamster ovary cell mutants to fibronectin substrata

We have examined the role of cell surface glycosaminoglycans in fibronectin-mediated cell adhesion by analyzing the adhesive properties of Chinese hamster ovary cell mutants deficient in glycosaminoglycans. The results of our study suggest that the absence of glycosaminoglycans does not affect the initial attachment and subsequent spreading of these cells on substrata composed of intact fibronectin or a fibronectin fragment containing the primary cell-binding domain. However, in contrast to wild-type cells, the glycosaminoglycan-deficient cells did not attach to substrate composed of a heparin-binding fibronectin fragment. Furthermore, the wild-type but not the glycosaminoglycan-deficient cells formed F-actin-containing stress fibers and focal adhesions on substrata composed of intact fibronectin. We propose, therefore, that cell surface proteoglycan(s) participate in the transmembrane linking of intracellular cytoskeletal components to extracellular matrix components which occurs in focal adhesions.

Interaction of embryonic chick calvarial bone cells with collagen substrata; attachment characteristics and growth behaviour

Freshly isolated bone cells from embryonic chick calvariae were found to attach to films of native monomeric type I or II collagen, to gels of native type I collagen and, to a lesser extent, to air-dried films of these collagens, in a fibronectin-independent manner, and with no requirement for collagen synthesis. Adhesion to type III collagen was relatively poor. Films of native monomeric type I or II collagen were preferred to the corresponding air-dried films as substrata for proliferation. Cis-hydroxyproline inhibited growth upon both plastic and type I and type II collagenous substrata, but this effect could largely be overcome by providing bone cell-conditioned medium, suggesting the possibility of a requirement in the growth process for synthesis of a collagen other than type I, the major form in bone matrix.

Association between fibronectin receptor and the substratum: spare receptors for cell adhesion

We have investigated the association of the recently described 140-kDa cell membrane receptor for fibronectin with the cytoskeleton or with substratum-bound fibronectin. Using a monospecific polyclonal antibody to the 140-kDa receptor, we have demonstrated that most of the receptor molecules are soluble in nonionic detergent either in suspension culture CHO cells or in CHO cells attached to and spread on a fibronectin-coated substratum. This may suggest that putative linkages of the receptor either to fibronectin or to detergent-insoluble cytoskeletal components are labile to nonionic detergent and thus are rather weak. Alternatively, it may mean that only a small fraction of the cell's receptors are needed to mediate adhesion. In order to explore this latter concept, we have coated substrata with various concentrations of PB1, a monoclonal antibody with a high affinity for fibronectin receptor. We demonstrate that coating the substratum with increasing concentrations of PB1 results in increasing amounts of 140-kDa receptor becoming bound to the substratum in detergent-insoluble form. However, the amount of receptor bound does not necessarily correlate with the degree of cell adhesion and spreading. Thus, coating the substratum with 5 micrograms/ml of PB1 results in essentially complete attachment and spreading of CHO cells, but only a small fraction of the 140-kDa receptor becomes substratum bound. Coating with 50 micrograms/ml of PB1 produces no further increase in cell adhesion and spreading, but results in the detergent-stable association of a large fraction of the total receptor pool with the substratum. These observations suggest the possibility of there being "spare" receptors for cell adhesion processes.

Evidence for the recycling nature of the fibronectin receptor of macrophages

Plasma fibronectin (pFN) has been shown to mediate phagocytosis of several types of artificial particles and tissue debris by macrophages. In the present investigation some of the dynamic aspects of this receptor-mediated cellular process have been studied. Plasma fibronectin did not bind specifically to fibronectin (FN)-receptors of rat peritoneal macrophages at either 4 degrees C or 37 degrees C. On the other hand, pFN aggregated on the surface of gelatin-coated latex beads (gLtx) and 125I-labeled pFN covalently coupled to latex beads (pFN-Ltx) bound strongly to macrophages at both temperatures. Both of these particles were also internalized at 37 degrees C. Treatment of macrophages by chymotrypsin, thermolysin, or trypsin in a protein-free tissue culture medium did not affect either of the above reactions; however, pronase treatment strongly reduced both the binding and internalization of the pFN-coated particles. The pronase-treated macrophage monolayers in time regained their ability to bind and internalize pFN-gLtx when incubated in fresh tissue culture medium. Such recovery, however, did not take place when the medium contained cycloheximide. On the other hand, phagocytosis of pFN-gLtx was not affected directly by cycloheximide with untreated macrophages; this suggests that the FN-receptor recycles during sustained phagocytosis. This assumption was substantiated by the observations that some of the established lysosomotropic amines--i.e., chloroquine, dansylcadaverine, and dimethyldansylcadaverine--caused total inhibition of internalization without affecting the binding of particles to macrophages. Furthermore, chloroquine protected the FN-receptors against destruction by pronase. Together these results suggest that macrophage receptors for FN are protein, present both on the cell surface and intracellularly, and recycle between the plasma membrane and intracellular sites during phagocytosis.

Expression and function of a putative cell surface receptor for fibronectin in hamster and human cell lines

We have previously reported the use of monoclonal antibodies to identify a 140-kD cell surface glycoprotein in mammalian cells that is specifically involved in fibronectin-mediated cell adhesion. We now report the purification of this molecule using immunoaffinity chromatography and the subsequent generation of polyclonal antibodies that selectively immunoprecipitate 140-kD putative fibronectin receptor glycoprotein (gp140) extracted from rodent or human cells; these antibodies also specifically block fibronectin-mediated cell adhesion but not adhesion mediated by other factors in serum. Expression of gp140-like molecules was detected on the surfaces of several adherent human cell lines (HDF, WISH, and EFC) but not on erythrocytes; however, gp140 was also detected on a nonadherent human lymphoid line (DAUDI). Analysis of gp140 on nonreducing SDS gels revealed two closely migrating bands. Protease digestion and peptide mapping suggests that the two bands are closely related polypeptides.

Enhanced levels of attachment of fibronectin-primed Treponema pallidum to extracellular matrix

Freshly extracted Treponema pallidum organisms treated with exogenous human fibronectin (Fn) (Fn-primed treponemes) showed a 6- to 15-fold increased level of attachment to Fn-coated cover slips and to extracellular matrix (ECM) when compared with unprimed treponemes. Treponemes primed with collagen or laminin showed no similar enhanced binding to immobilized Fn or ECM. Preexposure of immobilized Fn and ECM to anti-Fn serum but not to anti-collagen or anti-laminin serum prevented treponemal adherence. Also, the presence of proteoglycanlike molecules such as dextran sulfate or heparan sulfate inhibited Fn-primed treponemal attachment to Fn or ECM. In contrast Fn-primed treponemes did not exhibit elevated levels of attachment to eucaryotic cell monolayers. To understand the increased tropism of Fn-primed T. pallidum organisms for Fn and ECM-like surfaces, we radiolabeled freshly extracted treponemes with [35S]methionine and examined them for the presence of surface immunoreactive Fn. Magnetic protAspheres and glass beads coated with monospecific anti-Fn serum bound only 20 to 30% of radiolabeled treponemes. Nonadherent treponemes failed to bind to gelatin-agarose, further confirming the absence of surface Fn or Fn-like material. Fn-free organisms, however, did attach to Fn-coated cover slips and to cell monolayers like treponemes of the original population. Incubation of Fn-free treponemes with human Fn resulted in almost total binding of organisms to anti-Fn antibody on glass beads and also produced increased attachment to Fn-coated cover slips and ECM. These results suggest that enhanced interactions between T. pallidum and the host are dependent on the presence of Fn on syphilis spirochetes and the specific location and orientation of Fn in vivo.

Plasma membrane potential of murine erythroleukemia cells: approach to measurement and evidence for cell-density dependence

The plasmamembrane potential (delta psi p) of murine erythroleukemia (MEL) cells has been determined by measuring the distribution of the lipophilic cation tetraphenylphosphonium (TPP+) across the plasmamembrane. TPP+ accumulation within the cells (experimental accumulation ratio, AR exp) was measured by adding 2 microM TPP+ directly to the culture flasks, avoiding any other perturbation of the experimental system. The mitochondrial contribution to AR exp, evaluated by adding carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) or 2,4-dinitrophenol (DNP), was apparently negligible in standard cultures, AR exp being substantially the same in either the absence or presence of these uncouplers. However, the addition of oligomycin produced a strong AR exp enhancement, which was abolished by FCCP, suggesting that MEL cell mitochondria are in state 3. The aspecific TPP+ binding was estimated by a new mathematical approach worked out to fit AR exp values measured in the presence of valinomycin at various extracellular K+ concentrations and plotted against the ratio of intracellular to extracellular K+ concentration ([K+]i/[K+]e). This binding was found to be close to zero in MEL cells. By applying the Nernst equation directly to AR exp values, delta psi p of these cells was then measured; this potential varying from -65 mV to -16 mV (inside negative) is inversely related to the cell density on the culture surface on which the cells sediment (cells/cm2; CD). The dependence of delta psi p on CD is practically eliminated by valinomycin and appears to be related to a cell interaction with the culture surface of the flasks, suggesting that in the immediate environment of MEL cells one or more factors are produced that modulate the K+ plasma membrane permeability (PK).