PART 11.

CHARACTERISTICS OF A HIGH-MOLECULAR-WEIGHT GLYCOPROTEIN IN CONNECTIVE TISSUEAND BASEMENT MEMBRANES

ORGANIZATION OF EXTRACELLULAR PROTEINS O N T H E CONNECTIVE TISSUE CELL SURFACE: RELEVANCE T O CELL-MATRIX INTERACTIONS I N VITRO A N D IN VIVO* Paul Bornstein, Dan Duksin, Gary Balian,? Jeffrey M. Davidson, and Ed Crouch Department of Biochemistry University of Washington Seattle, Washington 981 95

The fluid mosaic model for the structure of the plasma membrane’ postulates the existence of integral proteins that, within limits, are free t o diffuse laterally in the plane of the membrane formed by a fluid lipid bilayer, and the occurrence of interactions of such proteins with peripheral proteins either on the cytoplasmic or on the external surface of the cell ~ n e m b r a n e . ~It- ~is thought that cytoplasmic microfilaments are linked directly or indirectly to integral proteins via peripheral proteins subjacent t o the cell membrane and that many vellular properties, including adhesion, shape, and motility, are controlled by dynamic interactions between these protein component^.^-^ As a consequence of these concepts and of the studies they engendered, it has been possible to achieve a better understanding of the binding of drugs and hormones to membrane receptors,7 the clustering of receptors by antibodies and l e ~ t i n s , ~ *and ~ * *cell-cell interactions.’ Considerable light has also been shed o n alterations in the properties of neoplastic and transformed cells, many of which are reflected in changes in cell surface In contrast, the interactions of the connective tissue cell surface with the extracellular matrix and the consequences of this relationship for cellular function have received less attention. These interactions are likely to account, in part, for such diverse phenomena as directed cellular migration, maintenance of a characteristic shape, the ability of the cell to organize a complex matrix, adherence to surfaces, and the role of the cell surface in tissue differentiation. We have developed a model for the organization of the external cell surface (FIGURE1) designed to account for the nature of cell-matrix interactions and for the cytoplasmic control of these interactions. We propose that in a connective 1 ,A), fibronectin (see below) serves as tissue cell attached to a substratum (FIGURE an external peripheral protein, one of whose functions is to bind collagen o r collagen precursors, or both. The length of the collagen molecule (-3000 A ) and its tendency to aggregate promote clustering of fibronectin-binding integral proteins, resulting in a meshwork on the external cell surface. This protein meshwork may be linked, in turn, to microfilaments via transmembrane integral proteins and cytoplasmic peripheral proteins. Collagen, in its role as a “secondary” *Supported by National Institutes of Health Grants AM 11248, DE 02600, and H L 18645 and Training Grant G M 07266 (to E.C.). ?Established Investigator of the American Heart Association.

93 0077-8923/78/0312-0093 M)I .75/0 01978, The New York Academy of Sciences

P

W

FIGUREI . The molecular organization of the cell surface i n a connective tissue cell attached to a substratum (A) and in a similar cell detached from the surface by the action of proteases, in mitosis, or after viral transformation ( 8 ) .The model is based. in part, on the fluid mosaic structure of the cell membrane proposed by Singer and Nicolson I and on the relation of the internal cytoskeleton, which consists of microfilaments and microtubules, to the cell surface as discussed by Edelman3 and Nicolson.4 The relative dimensions of cell membrane constituents and submembranous structures have been distorted to illustrate processes that involve membranous proteins; the implications of depicting these various elements to scale were presented by Loor.13 The quantity of membrane proteins, relative to lipid, has also been reduced for the purpose of clarity. I n a cell attached to a substratum (A), integral membrane proteins are clustered by a cell surface protein complex, which consists of fibronectin (shown as a capsule-shaped molecule) and collagen. The dimeric nature of fibronectin permits clustering of integral proteins; interactions are depicted to involve the oligosaccharide side chains of integral proteins, but carbohydrate side chains of fibronectin may also participate. or amino acid side chains in both proteins may be primarily involved. Binding of fibronectin to the cell membrane may involve divalent cations. The tendency of collagen to aggregate serves to form a collagen-fibronectin meshwork on the surface of the cell. This meshwork is linked via membrane proteins to microfilaments ( M F ) and

FIGURE1, continued microtubules (MT), thus enveloping the bilayer membrane in a two-ply system. This linkage is shown to be direct but could function indirectly through ionic fluxes or changes in cyclic nucleotides. In cells detached from a surface by proteases or in mitotic or transformed cells (B), fibronectin and collagen are missing from the cell surface, or are markedly reduced, and the system of microtubules and microfilaments is disorganized. Fibronectin-binding integral proteins are more dispersed. Some oligosaccharide side chains are shown to lack normal determinants. Interactions between the external protein meshwork and the internal cytoskeleton are thought to be cooperative, so that disturbances in either component could lead to disruption of the other. This model depicts protease-treated, mitotic, and transformed cells identically to emphasize common surface features of these cells. Clearly, important differences among these three types of cells must exist.

96

A n n a l s N e w York Academy of Sciences

peripheral protein (secondary in the sense that it does not interact directly with an integral membrane protein), adds cohesiveness to the external meshwork and mediates some of the interactions of the cell with its extracellular environment. Additional stabilization may be provided by formation of disulfide bonds between fibronectin monomers and between fibronectin and other cell surface prot e i n (not ~ ~ ~ shown in Figure). Alterations in any of the components of the system by genetic or environmental changes would be expected to disrupt the integrity of the entire meshwork (FIGURE 1,B). Thus, changes in components of the internal cytoskeleton caused by drugs, mitosis, viral transformation, failure to synthesize or secrete collagen or fibronectin normally, or failure to assemble these proteins on the cell surface should all have major predictable consequences for cell adhesion, shape, and motility and matrix organization. In the following discussion, we summarize the data that led to the development of this model and examine some of its consequences. FIBRONECTIN I S INVOLVED

IN CELL-SUBSTRATE INTERACTIONS

Recently, a glycoprotein, or a group of glycoproteins, variously known as cell surface protein, l 5 fibroblast surface antigen,16 LETS (large external transformation-sensitive) protein,” galactoprotein a,I8 Z protein,” and fibronectin,20*2ihas been found on the surface of fibroblast^,'^.^^ smooth muscle cells,23glial cells,24and both epithelial” and endothelia126cells. The protein has been identified primarily by cell surface labeling27and exists as a dimer with a subunit molecular weight of approximately 230,000 daltonslO; its linkage to the cell surface is known to be sensitive to trypsin and other proteases. Fibronectin (the term used here) is reduced or absent from the surfaces of many transformed and malignant cells of mesenchymal and is thought t o be involved in the attachment of a cell to a matrix (see below). Of considerable interest is the recent finding that fibronectin also exists as an important structural component of the extracellular matrix .28,29 While intercellular interactions are likely to involve complex specificities that function i n cellular recognition, the molecular specificities that govern cell-matrix interactions may be more limited in number, since motile connective tissue cells, such as fibroblasts, interface primarily with collagen fibrils and proteoglycan aggregates and, to lesser extents, with elastic fibers, extracellular connective tissue microfibrils, and other cells. Several studies implicate fibronectin in cell-substrate interactions: ( I ) During attachment and spreading, cells in culture deposit a “microexudate” on the artificial surface that consists of hyaluronic acid, collagen, and several glycoproteins, one of which is similar in size t o f i b r ~ n e c t i n . ’32~ (2) Mutagenized Balb/c 3T3 cells, selected for reduced adhesiveness to a plastic surface, were examined for cell surface proteins by lactoperoxidase-catalyzed iodination; some of these cells were shown to lack several proteins, including f i b r ~ n e c t i nThe . ~ ~ defect appeared to be primarily in adhesion, because addition of dibutyryl cyclic adenosine monophosphate to the culture medium caused rounded mutant cells to flatten and to revert partially t o the morphology of the wild-type cells but did not increase the adhesion of these cells t o the substratum. (3) Fibronectin isolated from embryonic chick or N I L hamster fibroblasts restored a fibroblast-like morphology and increased adhesiveness and contact inhibition of movement in several transformed cell lines.34-36Fibronectin can also agglutinate forrnalinized sheep erythrocytes, and this reaction was inhibited by agents that chelated calcium.37The latter observation is consistent with the ability of chelating

Bornstein et al.: Cell Surface Proteins

97

agents to detach cells from a substratum. (4)Tunicamycin, an inhibitor of lipid carrier-dependent protein g l y c o ~ y l a t i o ninhibited , ~ ~ ~ ~ ~the secretion of fibronectin by embryonic chick fibroblastsa and 3T3 cells4' in culture. Mouse 3T3 and virally transformed 3T3 cells showed dramatic changes in surface morphology and reduced adhesiveness in the presence of t ~ n i c a m y c i n Transformed .~~ cells were far more sensitive to the drug, while effects with normal 3T3 cells were seen primarily when cells were in the logarithmic growth phase. Cell surface labeling demonstrated that tunicamycin interfered with insertion of fibronectin on the cell surface. The increased tendency of transformed 3T3 cells to detach from a surface in the presence of tunicamycin is consistent with a marked reduction of cell surface fibronectin in these cells in comparison with most normal cells (see below).

A REDUCTIONIN FIBRONECTIN I S ASSOCIATED WITH DECREASED CELL-SUBSTRATE ADHESION

Numerous studies have documented that fibronectin is either absent or reduced on the surfaces of virally transformedlo-l2and neoplastic cell^.*^*^^ A good correlation was observed between lack of fibronectin, as detected by surface immunofluorescence, and oncogenicity in several virally transformed cells and in cells derived from spontaneously occurring neoplasms.43 Virally transformed and neoplastic cells differ from normal cells in cell-cell and cell-substrate adhesive proper tie^.^^,^^ Indeed, reduction in adhesion to an extracellular matrix may play a role in metastatic spread of malignant tumors. Although the factors involved in these cell surface modifications are complex,".'2 there are increasing indications that loss of fibronectin from the surface of transformed cells contributes to reduced cell-substrate adhesion. Perhaps the most direct evidence for this hypothesis is the restoration of adhesiveness in transformed cells induced by addition of f i b r ~ n e c t i n . ~ ~ . ~ ~ Cells in mitosis possess certain of the altered cell surface properties displayed by transformed cells, including reduced adhesion to a surface.46 In mitosis, a marked reduction of lactoperoxidase-catalyzed labeling of fibronectin was observed in hamster NIL and mouse L-929 cells.47A8Mitotic cells also demonstrate reduced cell surface, but not intracellular, fibronectin concentrations when examined by immunofluorescence m i c r o ~ c o p yConversely, .~~ cells arrested in the G 1 phase of the cycle, as a result of density-dependent inhibition of growth or serum deprivation, show high levels of cell surface f i b r o n e ~ t i n . ~ ~ ~ ~ ~

FIBRONECTIN BINDS BOTH TO

THE CELL SURFACE AND TO COLLAGEN

The binding of fibronectin to the cell surface is apparent from its localization by i m m u n o f l u o r e ~ c e n c e and ~ l ~ ~from ~ its ability to restore some normal surface properties to transformed A surface location is supported by the finding that antifibronectin serum is cytotoxic to fibroblasts in the presence of complement.s2 We believe that there is also sufficient evidence to implicate binding of cell surface fibronectin to collagen. A high-molecular-weight protein in serum, thought to be related to fibronectin, mediates attachment of SV40-transformed 3T3 and CHO cells to collagen-coated surfaces and requires divalent cations and cellular ~ ~ , ~factor ~ binds to collagen, dissociated metabolic energy for its f u n ~ t i o n . This

98

Annals N e w Y o r k A c a d e m y of Sciences

collagen a-chains, and to a cyanogen bromide-produced collagen fragment, a l(I)-CB7." An interaction between collagen o r gelatin and cold-insoluble glob-

ulin (CIg), a serum protein similar in molecular weight and immunologically related to fibronectin, has been shown both i m m ~ n o c h e m i c a l l yand ~ ~ by direct binding experiment^.^^.^'.^^ It is of interest that collagen chains and the al(1)-CB7 fragment enhance attachment of myoblasts to a plastic surface.'" Fusion of myoblasts and myotube formation in vitro are promoted by interaction of the cells with collagen o r collagen-related protein^^^,^* and depend on the presence of calcium ions.59 Fibronectin appears to play a central role in this process, because there are alterations in the distribution of surface fibronectin a t the time of fusion,60 and virally transformed myoblasts, which had markedly reduced levels of surface fibronectin, failed to fuse.6' COLLAGEN is

A

CELLSURFACE COMPONENT OF ATTACHED CONNECTIVE TISSUECELLS

Studies dealing with a variety of cells attached to a s u b ~ t r a t u m ~and ~ ~ "in~ suspensionw indicate that antibodies to collagen, o r to a collagen-related synthetic polytripeptide, induce complement-mediated cytotoxicity. In some of these and other studies,65collagen was located on the cell surface by indirect immunofluorescent staining. Recently, Bornstein and Ash23 demonstrated collagen or procollagen, or both, on the surface of monolayer normal rat kidney ( N R K ) fibroblasts by use of affinity-purified antibodies to collagen. As judged by indirect immunofluorescence, the protein was distributed in a reticular fashion on the cell surface, and antibody-induced translational movement of this network in the plane of the membrane was severely restricted. Similar distribution and behavior were observed for fibronectin in f i b r o b l a s t ~and ~ ~ ,in~ smooth ~ muscle cells.23When N R K fibroblasts were dissociated with trypsin and examined in suspension, neither collagen nor fibronectin was observed on the cell surface; mitotic figures observed in monolayer cultures of these cells also showed markedly reduced or absent immunofluorescent staining with antibodies to these proteins.23 These observations suggest that the external protein meshwork is released or shielded when cells are detached from a substratum.

MODELFOR

THE ON THE

ORGANIZATION OF EXTRACELLULAR PROTEINS CONNECTIVE TISSUEC E L L SURFACE

We propose that connective tissue cells in vitro and in vivo synthesize and organize a complex extracellular cell surface meshwork that contains both fibronectin and collagen and that can interact with other cell surface proteins, with the internal cytoskeleton, and with the extracellular matrix. It should be emphasized that fibronectin and collagen are fundamentally viewed as extracellular matrix glycoproteins. Under the constraint of two-dimensional growth in the fluid environment of cell culture, they are the major secreted proteins of connective tissue cells. However, the considerations discussed below have led us to reconsidef the conventional concept that a distinct dividing line separates the cell surface and the matrix. The interaction of fibronectin with collagen (FIGURE1,A) probably involves on divalent discrete regions of the collagen m o l e ~ u l e and ~ ~ may ~ ~ ~be~ dependent *~

Bornstein et al.: Cell Surface Proteins

99

cationss4; cations may also mediate binding of fibronectin to the cell surface. We are currently unable to exclude the possibility that procollagen rather than collagen (or perhaps a combination of the two proteins) binds to fibronectin, because antisera t o rat procollagen also stain N R K fibroblast^.,^ If so, the nontriplehelical domains at both the NH, and COOH termini of procollagen may interact separately with other cell surface binding sites. We have also not distinguished among the several known collagen types66in assigning a role for collagen as a cell surface protein. Very possibly, different connective tissue cells have different collagen types on their surfaces, and this difference may contribute to the characteristic properties of these cells. In cells known to possess a structurally distinct basal lamina, the cell surface-associated collagen may serve as a scaffold for this extracellular structure. The similar, nonrandom immunofluorescent patterns observed for cell surfaceassociated fibr~nectin*’,~~ and collagen,23 the interaction between these macromolecules, and the relatively restricted movement, at the light microscope level, induced by divalent a n t i b ~ d yall ~ ~suggest . ~ ~ that fibronectin and collagen form a discrete meshwork on the surface of normal cells attached to a substratum. Preliminary experiments performed with separate fluorescein and rhodamine labels for fibronectin and collagen indicate that these proteins are partially coextensive on the surfaces of chicken and rat embryo fibroblasts.68 Additional support for the nonrandom distribution of fibronectin on the cell surface is provided by the segregation of the major part of the externally labeled protein into a high-density subclass of plasma membrane vesicles after homogenization by nitrogen cavitat i ~ nDirect . ~ ~ evidence for the immobility of cell surface fibronectin has also come from fluorescence photobleaching recovery experiment^.^' We further propose that this external protein meshwork interacts, via transmembrane and cytoplasmic peripheral proteins, with microfilaments subjacent to the plasma membrane and that, together, these assemblies envelop the plasma membranein a two-ply system. It is of interest that the addition of fibronectin to transformed cells was associated with the appearance of well-defined actin cables that were not evident in the untreated transformed cells.36 We envision that the integrated function of the internal cytoskeleton and external surface meshwork dictates several properties of normal cells, including adhesion to a substratum, and that the interactions with neighboring cells are manifest in the phenomenon of density-dependent inhibition of m ~ v e m e n t . ~ ’ While we have focused on two protein constituents of this putative external protein meshwork, it is likely that other proteins contribute t o the complex. Fibrinogen and CIg interact,72v73 and plasma transglutaminase (blood coagulation factor XIII), which can cross-link fibronectin on the surface of human fibroblasts,,’ is also capable of forming hybrid polymers of fibrin and CIg.73 Other components of the extracellular matrix, including proteoglycans, may also interact with the cell surface meshwork. Virally transformed cells, cells in mitosis, and cells dissociated by proteolytic enzymes have certain surface characteristics in common (FIGURE 1,B). The internal cytoskeleton in these cells is partially disassembled o r organized differentl~.’~.~’ These alterations could lead to a reduced level of surface fibronectin if, as we suggest, the organization of integral fibronectin-binding proteins is dependent o n the organization of the internal cytoskeleton. On the other hand, the external protein meshwork could also exert a stabilizing influence o n the internal cytoskeleton. However, measurement of synthetic rates for f i b r ~ n e c t i nand ~ ~ the finding of reduced levels of m R N A for both fibronectin and procollagen in transformed suggest that reduced synthesis of these proteins contributes to the altered

1 00

A n n a l s New York Academy of Sciences

cell surface characteristics that are observed in these cells. Cells dissociated with proteases may lack fibronectin as a consequence of the protease sensitivity of the protein, and the disassembly of actin-containing microfilament bundles may, in turn, result from this surface prote~lysis.'~The interrelation of these events is underscored by the findings that cell surface proteolytic activity is increased after transformation12 and that transformed cells demonstrate altered membraneassociated cytoskeletal elements and transmembrane ~ o n t r o l . ~ . ' These observations suggest that the external protein meshwork and elements I(A). The very close of the internal cytoskeleton interact as suggested in FIGURE apposition of cytoplasmic cortical microfilaments with the cell membrane7' is consistent with linked protein interactions. Alternatively, alterations in surface proteins may lead to ionic fluxes or changes in cyclic nucleotides that secondarily affect the assembly of microfilaments and microtubules. Enzymatically dissociated cells and cells in mitosis lack cell surface-associated collagen,*3 a finding in accord with the model proposed in FIGUREl(B). Our model predicts that virally transformed cells will also have reduced cell surfaceassociated collagen. Since transformed cells generally synthesize far less collagen than d o their normal counterparts,80 a distinction between reduced synthesis and binding may be difficult to make.

'

OF A FI~KONECTIN-COLLAG~N POSSIBLEFUNCTIONS CELL SURFACE MESHWORK

There are many functions that might be well served by a collagen-fibronectin complex on the connective tissue cell surface, in addition to adhesion to a substratum, density-dependent inhibition of movement, and effects related to cell shape and motility. In specialized tissues, such as the cornea, a precise arrangement of collagen fibrils is observed; intricate spatial orientation of extracellular structures of this type could be effected, in part, by the ability of the cell to use its surface as a nucleation site for a growing fibril. Directed migration of connective tissue cells over considerable distances occurs in the developing embryo." Such movement could take advantage of cell surface collagen-collagen fiber interactions. Indeed, the inductive properties of collagena2 may be mediated by such interactions. Mobility would require transient interactions. This property could be provided by subjecting the extracellular protein meshwork to transmembrane control by cellular metabolic changes that act via the internal cytoskeleton. Finally, the ability of the external protein meshwork to interact with other proteins, such as fibrin, could serve an important role in the localization of fibroblasts in a healing wound. Collagen chains and fragments have been reported to be chemotactic for both fibroblasts83 and m o n ~ c y t e s The . ~ ~ ability to bind collagen may therefore not be limited to connective tissue cells. Collagen is known to be capable of binding to platelets, resulting in platelet aggregation and release of granular contents. The effectiveness of collagen in this process is enhanced markedly by prior aggregation of collagen The requirement for polymeric collagen may reflect the need to redistribute fibronectin-like platelet membrane proteins to elicit the physiologic response. Platelets, and possibly other cells, bind the Clq component of complement, a protein known to contain a collagen-like domain.86It would he of interest to determine whether fibronectin or a related protein is involved in these interactions and whether a process analogous to the formation of a cell surface meshwork might occur in these cells.

Bornstein el al.: Cell Surface Proteins

101

SUMMARY A model has been developed that proposes a cell surface-associated protein meshwork, composed in part of fibronectin and collagen, for a connective tissue cell attached to a substratum. In support of this model are the observations that collagen and fibronectin interact and that these proteins are similarly distributed on the fibroblast cell surface. We suggest that this external meshwork interacts directly o r indirectly with the internal cytoskeleton and with the extracellular matrix and thereby mediates several cellular properties, including adhesion, shape, and motility. Loss of cell surface fibronectin as a result of viral transformation, or due to treatment of normal cells with tunicamycin, an inhibitor of protein glycosylation, may contribute to the reduced adhesion and altered morphology observed in these circumstances. We therefore predict that the changes in these properties observed with virally transformed cells, mitotic cells, and cells treated with proteolytic enzymes are related to alterations in the external protein meshwork.

ACKNOWLEDGMENT

We thank Virginia Brooks for assistance with the illustrations. REFERENCES I. 2. 3. 4. 5. 6. 7.

8. 9.

10.

II.

12.

SINGER,S. J. & G. L. NICOLSON. 1972. The fluid mosaic model of the structure of cell membranes. Science 175: 720-731. SINGER,S. J. 1974. The molecular organization of membranes. Annu. Rev. Biochem. 43: 805-833. EDELMAN, G. M . 1976. Surface modulation in cell recognition and cell growth. Science 192: 218-226. NICOLSON,G. L. 1976. Transmembrane control of the receptors on normal and tumor cells. 1. Cytoplasmic influence over cell surface components. Biochim. Biophys. Acta 457: 57-108 POSTE,G., D. PAPAHADJOPOULOS & G. L. NICOLSON.1975. Local anesthetics affect transmembrane cytoskeletal control of mobility and distribution of cell surface receptors. Proc. Nat. Acad. Sci. USA 72: 4430-4434. REES,D. A., C. W. LLOYD& D. THOM.1977. Control of grip and stick in cell adhesion through lateral relationships o f membrane glycoproteins. Nature (London) 267: 124-128. CUATRECASAS, P. & M. D. HOLLENBERG. 1976. Membrane receptors and hormone action. Adv. Protein Chem. 30: 251-451. SINGER,3. J . 1974. Molecular biology of cellular membranes with applications to immunology. A h . Immunol. 19: 1-66. SINGER, S. J. 1976. The fluid mosaic model of membrane structure. Some applications to ligand-receptor and cell-cell interactions. In Surface Membrane Receptors: Interface between Cells and Their Environment. R. Bradshaw, Ed.: 1-24. Plenum Publishing Corporation. New York, N.Y. HYNES,R . 0. 1976. Cell surface proteins and malignant transformation. Biochim. Biophys. Acta 458: 73-107. NICOLSON, G. L. 1976. Trans-membrane control of the receptors on normal and tumor cells. 11. Surface changes associated with transformation and malignancy. Biochim. Biophys. Acta 458: 1-72. ROBBINS, J. C. & G. L. NICOLSON. 1975. Surfaces of normal and transformed cells. In Cancer: A Comprehensive Treatise. Volume 4, Biology of Tumors: Surfaces, Im-

102

13. 14. 15. 16. 17. 18.

19.

20. 21. 22. 23. 24. 25. 26. 27. 28.

20. 30. 31. 32. 33.

34. 35.

Annals New York Academy of Sciences munology and Comparative Pathology, F. F. Becker, Ed.: 3-54. Plenum Publishing Corporation. N e w York, N . Y . LOOR,F. 1976. Cell surface design. Nature (London) 264: 272-273. HYNES,R. 0. & A. DESTREE.1977. Extensive disulfide bonding at the mammalian cell surface. Proc. Nat. Acad. Sci. USA 74: 2855-2859. YAMADA,K . M . & J . A. WESTON.1974. Isolation of a major cell surface glycoprotein from fibroblasts. Proc. Nat. Acad. Sci. USA 71: 3492- 3496. P. KUUSELA & E. LINDER.1973. Fibroblast surface antiKUOSLAHTI, E., A . VAHEKI, gen: a new serum protein. Biochim. Biophys. Acta 322: 352 358. HYNES,R. 0. 1973. Alteration of cell-surface proteins by viral transformation and by proteolysis. Proc. Nat. Acad. Sci. USA 70: 3170-3174. C. ti. & S.-I. HAKOMORI. 1975. Surface carbohydrates of hamster fibroGAHMBERG, blasts. 11. Interaction of' hamster NIL cell surfaces with ricinus communis lectin and concanavalin A as revealed by surface galactosyl label. J . Biol. Chem. 250: 2447-245 I . BLUMBERC, P. M. & P. W. ROBBINS.1975. Effect of proteases o n activation of resting chick embryo fibroblasts and on cell surface proteins. Cell 6: 137-147. K U U S E L A , P., E. RUOSLAHTI, E. ENGVALL & A. VAHBRI. 1976. Immunological interspecies cross-reactions of fibroblast surface antigen (fibronectin). Imrnunochemistry 13: 639-642. K ~ S K I - O JJ., A ,D. F. MOSHER& A. VAIIERI. 1976. Cross-linking of a major tibrohlast surface-associated glycoprotein (fibronectin) catalyzed by blood coagulation factor XIII. Cell 9 : 29 35. WARTIOVAARA, J., E. LINDER.E. RUOSLAHTI & A. VAHERI. 1974. Distribution O f fibroblast surface antigen. Association with fibrillar structures of normal cells and loss upon viral transformation. J. Exp. Med. 140: 1522--1533. BORNSTEIN, P. & J. F. ASH. 1977. Cell surface-associated structurai proteins in connective tissue cells. Proc. Nat. Acad. Sci. USA 74: 2480-2484. & J. PONTEN.1976. A common cell-type VAHERI, A,. E. RUOSLAHTI, B. WESTERMARK specific surface antigen in cultured human glial cells and fibroblasts: loss in malignant cells. J . Exp. Med. 143: 64-72. CHEN,L. B., N. MAITLAND, P. H . GALLIMORE & J . K . MCDOUGALL. 1977. Detection of the large external transformation-sensitive protein on some epithelial cells. Exp. Cell Res. 106: 39-46. BORNSTFIN, P. & S. SCHW.ARTZ. 1977. Unpublished observations. I). R . , J. A . WYKE& R . 0. HYNES.1976. Cell surface and metabolic CRITCIILEY, labelling of the proteins of normal and transformed chicken cells. Biochim. Biophys. Acla 436: 3 3 5 - 3 5 2 . I . I N I ) E R , E., A . V A H E R I ,E. RUOSLAHTI & J. WARTIOVAARA. 1975. Distribution of fibroblast surface antigen in the developing chick embryo. J. Exp. Med. 142: 41-49. S. & A . VAHERI. 1977. Distribution of tibronectin a new connective tissue STENMAN, protein-in normal human tissues. Uppsala J . Med. Sci. 82: 154. TERKY A ,. H. & L. CULP. 1974. Substrate-attached glycoproteins from nornial and virus-transformed cells. Biochemistry 13: 414-425. K . J . , R . E. BRANSON, P. B. HEWCLEY & L. W. C U N N I N C I I A1977. M . The LEMBACH, synthesis of macromolecular 3-hydroxyproline by attaching and confluent cultures of human fibroblasts. Eur. J. Biochem. 72: 379-383. CULP, L. A. 1976. Electrophoretic analysis of substrate-attached proteins from normal and virus-transformed cells. Biochemistry IS: 4094-4104. POUYSSEGUR, J. M. & I . PASTAN.1976. Mutants of' Balb/c 3T3 fibroblasts defective in adhesiveness to substratum: evidence for alteration in cell surface proteins. Proc. Nat. Acad. Sci. USA 73: 544 548. K . M., S . S . YAMADA & I. PAs1.m. 1976. Cell surface protein partially reYAMADA, stores morphology, adhesiveness. and contact inhibition of movement to transformed fibroblasts. Proc. Nat. Acad. Sci. USA 73: 1217-1221. K. M., S. H. OHANIAN & I . PASTAN.1976. Cell surface protein decreases YAMADA, microvilli and ruffles o n transformed mouse and chick cells. Cell 9: 241-245. ~

Bornstein et af.: Cell Surface Proteins

I03

36. ALI, I. U., V. MAUTNER,R. LANZA& R. 0.HYNES.1977. Restoration of normal morphology, adhesion and cytoskeleton in transformed cells by addition of a transformation-sensitive surface protein. Cell 11: 115- 120. K. M., S. S. YAMADA & 1. PASTAN.1975. The major cell surface glyco37. YAMADA, protein of chick embryo fibroblasts is an agglutinin. Proc. Nat. Acad. Sci. USA 72: 3158-3162. A. & G . TAMURA. 1971. Effect of tunicamycin on the synthesis of macro38. TAKATSUKI, molecules in cultures of chick embryo fibroblasts infected with Newcastle disease virus. J. Antibiot. 24: 785-794. 39. TKACZ,J. S. & J. 0. LAMPEN.1975. Tunicamycin inhibition of polyisoprenyl Nacetylglucosaminyl pyrophosphate formation in calf-liver microsomes. Biochem. Biophys. Res. Commun. 65: 248-257 1977. Impaired conversion of procollagen t o collagen by 40. DUKSIN,D. & P. BORNSTEIN. fibroblasts and bone treated with tunicamycin, an inhibitor of protein glycosylation. J. Biol. Chem. 252: 955-962. 41. DUKSIN,D. & P. BORNSTEIN.1977. Changes in surface properties of normal and transformed cells caused by tunicamycin, an inhibitor of protein glycosylation. Proc. Nat. Acad. Sci. USA 74: 3433-3437. & P. BORNSTEIN. 1978. Cell surface morphology and ad42. DUKSIN, D., K. HOLBROOK hesive properties of normal and virally transformed cells treated with tunicamycin, an inhibitor of protein glycosylation. Exp. Cell Res. In press. & J . K. MCDOUGALL. 1976. Correlation between tu43. CHEN,L. B., P. H. GALLIMORE mor induction and the large external transformation sensitive protein on the cell surface. Proc. Nat. Acad. Sci. USA 73: 3570-3574. R. & K. POLLOCK.1974. The adhesion of BHK and PyBHK cells to the 44. SHIELDS, substratum. Cell 3: 31-38. 45. SACHS,L. 1974. Regulation of membrane changes, differentiation, and malignancy in carcinogenesis. Harvey Lect. 68: 1-35. D. F., E. C . ANDERSON & R. A. TOBEY. 1968. Mitotic cells as a source of 46. PETERSEN, synchronized cultures. I n Methods in Cell Physiology. D. M. Prescott, Ed. Vol. 3: 347-370. Academic Press, Inc. New York, N.Y. 47. HYNES.R. 0. & J. M. BYE. 1974. Density and cell cycle dependence of cell surface proteins in hamster fibroblasts. Cell 3: 113-120. 48. HUNT,R. C., E. GOLD& J . C. BROWN.1975. Cell cycle dependent exposure of a high molecular weight protein o n the surface of mouse L cells. Biochim. Biophys. Acta 413: 453-458. S., J . WARTIOVAARA & A. VAHERI.1977. Changes in the distribution of a 49. STENMAN, major fibroblast protein, fibronectin, during mitosis and interphase. J. Cell Biol. 74: 453-467. E. & M. D. WATERFIELD. 1974. Metabolic studies on '*51-labeled baby 50. PEARLSTEIN, hamster kidney cell plasma membranes. Biochim. Biophys. Acta 362: I 12. 1974. Organization of glycolipids and glycoC. G . & S. I. HAKOMORI. 51. GAHMBERG, proteins in surface membranes: dependency on cell cycle and on transformation. Biochem. Biophys. Res. Commun. 59: 283-291. E. & H. 0. SIOGREN.1976. Specific anti-fibroblast cytotoxicity of anti52. RUOSLAHTI, bodies to fibroblast surface antigen. Int. J. Cancer 18: 375-378. 53. KLEBE,R. J. 1974. Isolation of a collagen-dependent cell attachment factor. Nature (London) 250: 248-25 1. 54. KLEBE,R. J. 1975. Cell attachment to collagen: the requirement for energy. J. Cell. Physiol. 86: 231-236. H. K., E. B. MCGOODWIN & R . J. KLEBE.1976. Localization of the cell 55. KLEINMAN, attachment region in types 1 and I1 collagens. Biochem. Biophys. Res. Commun. 72: 426--432. S. 1).& N. K. WHITE.1972. Studies of myogenesis in vitro. I. Temporal 56. HAUSCHKA, changes in the proportion of muscle-colony-forming cells during the early stages of limb development. 11. Myoblast-collagen interaction: molecular specificity required. Excerpta Med. Int. Congr. Ser. 240: 53-71. S. D. & I. R. KONIGSBERG. 1966. The influence of collagen on the de57. HAUSCHKA, velopment of muscle clones. Proc. Nat. Acad. Sci. USA 55: 119-126. ~

104 58. 59. 60. 61.

62. 63. 64. 65. 66. 67. 68. 68a. 69. 70.

71. 72. 73. 74. 75. 76. 77.

78.

79. 80.

Annals New York Academy of Sciences KETL.EY, J . M., R . W. ORKIN& G. R. MARTIN. 1976. Collagen in developing chick muscle in vivo and in vitro. Exp. Cell Res. 99: 261-268. 1972. Myosin synthesis in cultures of differentiating PATTERSON, B. & R. C. STROHMAN. chicken embryo skeletal muscle. Dev. Biol. 29: 113-138. C H E N ,L. B. 1977. Alteration in cell surface LETS protein during myogenesis. Cell 10: 393-400. D. R . CRITCHLEY & C. J. EPSTEIN.1976. HYNES,R . O., G. S. MARTIN,M. SHEARER, Viral transformation of rat myoblasts: effects on fusion and surface properties. Dev. Biol. 48: 35-46. LUSTIG,L. 1970. Cytotoxic action of an antiserum to soluble collagen on tissue culture of fibroblasts. Proc. Soc. Exp. Biol. Med. 133: 207-21 I . DUKSIN,D., A. MAOZ & S. FUCHS.1975. Differential cytotoxic activity of anticollagen serum on rat osteoblasts and fibroblasts in tissue culture. Cell 5 : 83-86. J. R., E. A. BAUER,R. HOYT & H. J . WEDNER.1976. Immunologic LICHTENSTEIN, characterization of the membrane-bound collagen in normal human fibroblasts: identification of a distinct membrane collagen. J . Exp. Med. 144: 145- 154. PAGEFAULK,N., L. B. CONOCHIE & A. TEMPLE.1975. Immunobiology of membranebound collagen on mouse fibroblasts. Nature (London) 256: 123-125. MILLER,E. J. 1976. Biochemical characteristics and biological significance of the genetically-distinct collagens. Mol. Cell. Biochem. 2: 165- 192. HYNES,R. O., A. T. DESTREE& V. M A U T N ~ 1976. R . Spatial organization at the cell surface. In Membranes and Neoplasia: New Approaches and Strategies. V. T. Marchesi, Ed.: 189-201. Alan R . Liss, Inc. New York, N.Y. VON DERMARK,K. 1977. Personal communication. 1978. In preparation. BORNSTEIN, P., J. F. ASH & G. BALIAN. GRAHAM J ., M.,R. 0. HYNES,E. A. DAvlDsoN & D. F. BAINTON. 1975. The location of proteins labeled by the ~2~I-lactoperoxidase system in the NIL 8 hamster fibroblast. Cell 4: 353 365. SCHLESSINGER, J., L. s. BARAK, G. G. HAMMES, K. M. YAMADA, I. PASTAN, W. w. WEBB& E. L. ELSON. 1977. Mobility and distribution of a cell surf’ace glycoprotein and its interaction with other membrane components. Proc. Nat. Acad. Sci. USA 74: 2909-2913. ABERCROMBIE, M. 1970. Contact inhibition in tissue culture. In Vitro 6: 128-142. EDSALL, J. T., G. A. GILBERT & H. A. SCHERAGA.1955. The non-clotting component of the human plasma fraction 1-1 (“cold insoluble globulin”). J. Amer. Chem. Soc. 77: 157-161. MOSHER,D. F. 1975. Cross-linking of cold-insoluble globulin by fibrin-stabilizing factor. J. Biol. Chem. 250: 6614-6621. POLLACK, R., M . OSRORN& K . WEBBR.1975. Patterns of organization of actin and myosin in normal and transformed cultured cells. Proc. Nat. Acad. Sci. USA 72: 994 998. R. & D. RIFKIN. 1975. Actin-containing cables within anchorage-dependent POLLACK, rat embryo cells are dissociated by plasmin and trypsin. Cell 6 : 495-506. OLDEN, K . & K . M . YAMADA. 1977. Mechanism of the decrease in the major cell surface protein of chick embryo fibroblasts after transformation. Cell 11: 957-969. ADAMS,S. L., M. E. SOBEL,B. H. HOWARD,K . OLDEN,K. M. YAMADA, B. DE CROMBRUCGHE 82 1. PASTAN. 1977. Levels O f translatable mRNAs for cell surface protein, collagen precursors, and two membrane proteins are altered in Rous sarcoma virus transformed chick embryo fibroblasts. Proc. Nat. Acad. Sci. USA 74: 3399-3403. ROWE,D. W., R. C. MOEN,J. M. D A ~ I D S O IP.WH. , BYERS,P. BORNSTEIN 82 R. D. PALMITER. 1978. Correlation of procollagen m R N A levels in normal and tranaformed chick embryo fibroblasts with different rates o f procollagen synthesis. Biochemistry. In press. T. D. & R. R. WEIHINC.1974. Actin and myosin and cell movement. POLLAKD. C R C Crit. Rev. Biochem. 2: 1-65. PETERKOFSKY, B . & W . B. PRATHER.1974. Increased collagen synthesis in Kirsten

Bornstein et al.: Cell Surface Proteins

81. 82. 83. 84. 85. 86. 87. 88.

105

sarcoma virus-transformed ‘BALB 3T3 cells grown in the presence of dibutyryl cyclic AMP. Cell 3: 291-299. TRELSTAD, R. L., E. D. HAY & J . P. REVEL.1967. Cell contact during early morphogenesis in the chick embryo. Dev. Biol. 16: 78-106. HAY,E. D. & S. MEIER. 1976. Stimulation of corneal differentiation by interaction between cell surface and extracellular matrix. 11. Further studies o n the nature and site of transfilter “induction.” Dev. Biol. 52: 141- 157. A. E., R. SYNDERMAN & A. H. KANG.1976. The chemotactic attracPOSTLETHWAITE, tion of human fibroblasts to a lymphocyte-derived factor. J. Exp. Med. 144: 11881203. POSTLETHWAITE, A. E. & A , H. KANG.1976. Collagen and collagen-peptide induced chemotaxis of human blood monocytes. J . Exp. Med. 143: 1299-1307. BRASS,L. F. & H . B. BENSUSAN.1974. Role of collagen quaternary structure in platelet-collagen interaction. J. Clin. Invest. 54: 1480-1487. REID, K. B. M. & R. R. PORTER.1976. Subunit composition and structure of subcomponent of Clq of the first component of human complement. Biochem J. 155: 19-23. DESSAU, W. & F. JILEK.1977. Interaction of the serum-derived AGF and denatured collagen. Uppsala J. Med. Sci. 82: 128. ENGVALL, E. & E. RUOSLAHTI.1977. Binding of soluble form of fibroblast surface protein, fibronectin, to collagen. Int. J . Cancer 20: 1-5.