Functional activation of integrin alpha V beta 3 in tumor cells expressing membrane-type 1 matrix metalloproteinase.

Matrix metalloproteinases (MMPs) and integrins have been implicated in a variety of processes involved in tumor progression. To evaluate the individual roles of integrin alphavbeta3 and membrane-type 1 matrix metalloproteinase (MT1-MMP), as well as the effects of their joint expression on tumor cell functions, MCF7 breast carcinoma cells were transfected stably with either the MT1-MMP, the beta3 integrin subunit or both MT1-MMP and beta3 cDNAs. MT1-MMP expression is accompanied by the functional activation of integrin alphaVbeta3, thereby increasing vitronectin-mediated adhesion and migration of MCF7 cells transfected with MT1-MMP and integrin alphaVbeta3. MT1-MMP-dependent functional activation of alphaVbeta3 correlates with modification(s) of the beta3 subunit, including its higher electrophoretic mobility and affected the LM609-binding site. MCF7 cells jointly expressing MT1-MMP and alphaVbeta3 were the most efficient in adhesion to the recombinant C-terminal domain of MMP-2 as well as in generating soluble and cell surface associated mature MMP-2 enzyme. These findings suggest a mechanism of selective docking of MMP-2 at tumor cell surfaces, specifically at the sites that include MT1-MMP and activated integrin alphaVbeta3. These mechanisms may provide a link between spatial regulation of focal proteolysis by the cell surface associated MMPs and the regulation of integrin-mediated motility of tumor cells.

Remodeling of collagen matrix by human tumor cells requires activation and cell surface association of matrix metalloproteinase-2.

We assessed the functional significance of tumor cell-associated matrix metalloproteinase (MMP)-2 in extracellular matrix remodeling compared with that of the soluble enzyme by evaluating the contraction of three-dimensional collagen lattices by human glioma U251.3 and fibrosarcoma HT-1080 cell lines. In this model, the constitutive synthesis and activation of the MMP-2 proenzyme were modulated by stable transfections of tumor cells with cDNA encoding membrane type 1-MMP (MT1-MMP). The efficiency of transfected cells in contracting collagen lattices was shown to be dependent on the MT1-MMP-mediated activation of MMP-2 accompanied by cell surface association of activated MMP-2, on the cell-matrix interactions controlled by collagen-specific integrins, and on the integrity of actin and microtubule cytoskeletons. Each one of these mechanisms was essential but was not sufficient by itself in accomplishing gel contraction by MT1-MMP-transfected cells. Both MMP-2 activation and gel contraction by transfected glioma cells were inhibited by tissue inhibitor of metalloproteinase (TIMP)-2 and the recombinant COOH-terminal domain of MMP-2. However, the kinetics and mechanisms of their inhibitory effects were different, because TIMP-2 and the COOH-terminal domain of MMP-2 preferentially inhibited the MT1-MMP-dependent and autocatalytic steps of MMP-2 activation, respectively. By contrast, TIMP-1, an efficient inhibitor of soluble MMP-2 activity, failed to affect gel contraction. In addition, soluble MMP-2 activated by either organomercurials or cells was not able to induce the contraction of collagen lattices when added to transfected cells. Therefore, soluble activated MMP-2, sensitive to TIMP-1 inhibition, does not mediate collagen gel contraction by tumor cells, whereas the activity of cell surface-associated MMP-2 plays a critical role in remodeling of the extracellular matrix in vitro. These mechanisms of functional and spatial regulation of MMP-2 may also be applicable to different aspects of tissue reorganization in vivo, including cell migration and invasion, angiogenesis, and wound healing.

Expression of tenascin-C splice variants in normal and bullous keratopathy human corneas.

PURPOSE: To characterize the expression patterns of tenascin-C (TN-C) splice variants in normal corneas and in those affected by pseudophakic-aphakic bullous keratopathy (PBK-ABK). METHODS: Alternatively spliced variants of TN-C mRNA from normal and age-matched human corneas with PBK-ABK were analyzed by semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) and Southern blot hybridization, using beta2-microglobulin as a housekeeping gene to normalize the samples. Normal and PBK-ABK corneas were studied by immunofluorescence and western blot analysis with antibodies to specific fibronectin type III-like (FN-III) repeats of TN-C. RESULTS: Tenascin-C mRNA expression was detected in epithelial, stromal, and endothelial cells of normal and PBK-ABK central corneas, although the protein was seen only in diseased corneas. Assessed by RT-PCR, PBK-ABK corneas expressed approximately three times more total TN-C mRNA than did normal corneas. Four major TN-C mRNA variants (with no FN-III insertional repeats or with retained insertional repeats D, A1, or A1+D) and three minor variants (with retained repeats A1+A2, A1+A2+D, or A1+A2+B+D) were much more abundant in PBK-ABK than in normal corneas. Repeat A1 was more abundant in PBK-ABK TN-C protein than repeats A2, A3, B, or D. Major TN-C variants in PBK-ABK corneas were in the range of 190 kDa to 240 kDa. CONCLUSIONS: Expression of TN-C mRNA and protein is higher in PBK-ABK corneas than in normal corneas. This increase mainly concerns relatively small TN-C splice variants that may affect corneal cell adhesion and migration and contribute to the exacerbation of PBK-ABK.

Novel splice variants of human tenascin-C mRNA identified in normal and bullous keratopathy corneas.

PURPOSE: Pseudophakic/aphakic bullous keratopathy (PBK/ABK) human corneas accumulate an extracellular matrix glycoprotein tenascin-C (TN-C), an important modulator of cell adhesion and migration. Here, the purpose was to identify specific TN-C mRNA splice variants in normal and PBK/ABK human corneas. METHODS: Conventional and semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) with primers to alternatively spliced (insertional) and constitutive fibronectin type II-like repeats of TN-C was used. Splice variants were identified by cloning and sequencing of RT-PCR products or by Southern blot analysis. RESULTS: The majority of corneal TN-C mRNA species corresponded to relatively small forms of the protein. Four previously unidentified TN-C mRNA splice variants were found in normal and PBK/ABK corneas that contained insertional repeats A1+A2+B+D, A1+A2+D, A1+B+D, or A1+D. Variants with insertional repeats A1+A2 or A1, previously described in mouse and rat, were also identified in human corneas. Semiquantitative RT-PCR showed that novel TN-C mRNA variants were dramatically elevated in PBK/ABK compared to normal corneas. CONCLUSION: TN-C protein was found in PBK/ABK but not in normal corneas; however, both normal and diseased corneas contained mRNA for 15 different TN-C isoforms. PBK/ABK corneas had elevated levels of six relatively small TN-C mRNA variants including five novel ones. These specific isoforms may adversely affect adhesion and migration of corneal cells thus contributing to the exacerbation of PBK/ABK.

Neutralizing anti-vascular endothelial growth factor antibody completely inhibits angiogenesis and growth of human prostate carcinoma micro tumors in vivo.

BACKGROUND: Neovascularization mediated by growth factors produced by tumors is critical for the growth of tumors. Vascular endothelial growth factor (VEGF) is one such growth factor. A neutralizing anti-VEGF antibody (A4.6.1) was recently shown in vivo to inhibit tumor angiogenesis and growth of the human rhabdomyosarcoma cell line A673. The antibody profoundly changed the growth characteristics of the tumor line from a rapidly growing malignancy to a dormant microcolony. METHODS: In the present study, we evaluated the effects of A4.6.1 (100 microg twice weekly, i.p.) on growth and angiogenic activity of spheroids of the human prostatic cell line DU 145 (diameter 700 microm at implantation) implanted in dorsal skinfold chambers in nude mice (n = 11). An antibody of the same isotype (n = 5) or saline (n = 5) was used as control. Tumor cells were prelabeled with a fluorescent vital dye (CMTMR), which allowed measurement of size of the implanted tumor spheroids throughout a two week observation period. FITC-dextran was used for plasma enhancement to visualize angiogenic activity. RESULTS: Tumors of control animals induced a neo-vasculature with high vascular density (350+/-12 cm[-1]). In animals treated with the anti-VEGF antibody, there was complete inhibition of neovascularization of the micro tumors and complete inhibition of tumor growth after the initial prevascular angiogenesis independent growth phase. CONCLUSIONS: These results demonstrate that inhibition of the key regulatory paracrine growth factor for endothelial cells, VEGF, results in complete suppression of prostate cancer induced angiogenesis and prevents tumor growth beyond the initial prevascular growth phase.

Matrix metalloproteinase-2 activation modulates glioma cell migration.

Stable transfection of U251.3 glioma cells with cDNA encoding MT-MMP-1 resulted in increased cell surface expression of MT-MMP-1 and TIMP-2, constitutive activation of MMP-2 proenzyme and increased collagen degradation. In tumor spheroid outgrowth assays, cell migration of MT-MMP-1 transfectants relative to control was enhanced on collagen and decreased on vitronectin and fibronectin. These effects were reversed by TIMP-2 and were not associated with any substantial changes in cell adhesion. Binding of U251.3 cells to the C-terminal domain of MMP-2 was specifically inhibited by anti-(alpha)vss3 integrin blocking antibody indicating that MMP-2 interacts with (alpha)vss3 through the enzyme's C-terminal portion at or near the integrin's matrix adhesion sites. We propose that these mechanisms could govern directed matrix degradation in the tumor cells' microenvironment by sequestration of active MMP-2 on the cell surface. Our data suggest that activation of MMP-2 and its proteolytic activity localized to the cell surface could differentially modulate tumor cell migration in response to particular matrix proteins by altering both composition of the extracellular matrix and expression of adhesion receptors on the cell surface.

Tumor cell invasion through matrigel is regulated by activated matrix metalloproteinase-2.

We tested the hypothesis that there is a correlation between tumor cell efficiency in activation of matrix metalloproteinase-2 (MMP-2) and invasion through basement membrane-like Matrigel barriers. To generate cells capable of MMP-2 activation, we stably transfected three human tumor cell lines, HT-1080 fibrosarcoma, MCF7 breast carcinoma, and U251.3 glioma with cDNA encoding the full length human membrane-type matrix metalloproteinase-1. Our results show a bimodal correlation between the extent of MMP-2 activation and Matrigel invasion by tumor cells. Cell transfectants characterized by a partial activation of MMP-2 were the most invasive while those with an extensive conversion of MMP-2 proenzyme into enzymatically active forms were the least efficient in invading Matrigel. Modulation of MMP-2 activation by exogenous TIMP-2 reverted the rate of Matrigel invasion by cell transfectants to control levels. We conclude that the regulation of activated MMP-2 in the tumor cells, microenvironment may be critical in facilitating tumor cell invasiveness.

IL-6 signals inhibition of cell adhesion in melanoma A375-C6.

IL-6 has been found to be a potent inhibitor of melanoma A375-C6 cell adhesion, in addition to its known action in arresting cells at G1/G0 phase of the cell cycle IL-6 treated melanoma cells were found to round up and to lose the ability to adhere to fibronectin, laminin, collagen, and tenascin over 72 to 96 hours of IL-6 treatment, a time course similar to that seen for cell cycle inhibition. Cell cycle inhibition and loss of adhesion were found, however, to be independent effects of IL-6. Analysis of cell surface integrins indicated significant changes in the expression of several integrins including downregulation of a3 and av beta 5 and upregulation of a3. However, the changes in integrin expression did not correlate with loss of adhesion to relevant ligands. Three A375 melanoma clones varying in metastatic potential also demonstrated inhibition of both cell proliferation and matrix adhesion by IL-6.

A novel monoclonal antibody, L1A3, is directed to the functional site of the alpha v integrin subunit.

We have generated a monoclonal antibody (MAb) L1A3 directed to the alpha v integrin subunit as shown by competitive binding with other anti-alpha v-specific MAbs and immunodepletion. MAb L1A3 is a function-blocking antibody inhibiting cell adhesion to the extracellular matrix proteins, fibronectin and vitronectin. Adherence to vitronectin of all cells studied including normal dermal microvascular endothelial cells and three tumor cell lines was inhibited in the presence of MAb L1A3. However, the contribution of the alpha v integrin subunit in mediating adhesion to fibronectin was dependent on the cell line, as indicated by differences in the inhibition of cell adhesion with MAb L1A3 and alpha 5 beta 1 integrin subunit blocking MAb P1D6. Glioma U251.3 cell adhesion to fibronectin was blocked by either MAb L1A3 or MAb P1D6 while fibrosarcoma HT1080 cells were blocked with MAb P1D6 only. Tumor cell migration mediated by vitronectin and fibronectin is blocked by MAb L1A3 in the two-dimensional spheroid outgrowth assay. Microvascular endothelial cell transwell membrane migration onto the fibronectin was also blocked by MAb L1A3. Comparison of the integrins involved in U251.3 cell migration on fibronectin or tenascin using a panel of integrin blocking MAbs including MAb L1A3 showed that only a subset of integrins participating in cell adhesion is essential for cell migration and these integrins appear to be ligand specific. Fibronectin-mediated tumor cell migration was critically dependent on alpha v integrins as shown by L1A3 blocking of migration while the beta 1 integrins were absolutely necessary for tenascin-mediated cell migration.

Tenascin mediates human glioma cell migration and modulates cell migration on fibronectin.

The role of tenascin in mediating tumor cell migration was studied using two cell migration models. In migration/invasion Transwell assays U251.3 glioma cells rapidly migrated through the 8 mu m pore size membranes onto tenascin- and fibronectin-coated surfaces. In this assay the number of cells migrating onto tenascin was 52.2 +/- 9.6% greater than on fibronectin within 4 hours. To assess cell migration rates and cell morphology, U251.3 migration was examined in a two-dimension spheroid outgrowth assay. The radial distance migrated by U251.3 cells from tumor spheroids was found to be 53.8 +/- 4.9% greater on tenascin than on fibronectin. Cells migrating on tenascin display a very motile appearance, while cells migrating on fibronectin spread and maintain close intercellular contacts. Cell migration in the presence of integrin blocking antibodies demonstrated that migration on tenascin and fibronectin is mediated by distinct integrins, alpha2beta1 and alphavbeta5/alphavbeta3, respectively. Since tenascin is coexpressed in malignant tumor matrices with fibronectin, we assessed the effects of tenascin on U251.3 cell migration mediated by fibronectin. Tenascin was found to provide a positive effect on fibronectin-mediated migration by altering cell morphology and enhancing cell motility. These effects of tenascin on fibronectin-mediated cell migration were inhibited by blocking beta1 and alpha2beta1 integrins. The results suggest that tenascin may play a significant role in promoting tumor cell migration and invasiveness by modulating cell responses to normal matrix components.

Identification of a growth factor for primary murine stroma as macrophage colony-stimulating factor.

Knowledge of the stromal microenvironment is crucial for understanding the hematopoietic system. We took advantage of an assay that permits analysis of primary stroma-initiating cells (SICs) on the clonal level, and further characterized SICs and the factors that regulate SICs. Stroma formation in this assay is dependent on a high-molecular-weight factor secreted by the stromal cell line AC3.U. Here we show that this factor is identical to macrophage colony-stimulating factor (M-CSF), and that purified M-CSF is sufficient for induction of stroma formation. M-CSF, isolated from the line AC3.U, as well as from L929 cells and COS cells transfected with an expression vector encoding M-CSF, migrated in two peaks as 160- and 650-kD species after gel filtration. These molecular-weight species encompassed all stroma-inducing activity, and both stimulated macrophage colony formation. Affinity chromatography and blocking studies with antibodies specific for M-CSF and c-fms confirmed M-CSF as the sole factor in the supernatant of the stromal cell line AC3.U that promotes stroma formation. Culture of marrow, for as little as 1 week, depleted M-CSF-dependent SIC while increasing the incidence of replatable, factor-independent SIC. This suggests that culture changes the properties of SICs, perhaps by inducing differentiation into mature stromal cells. Thus, our results show a novel function of M-CSF as an important modulator of stroma formation.

L-selectin and very late antigen-4 integrin promote eosinophil rolling at physiological shear rates in vivo.

Adherence of eosinophils to vascular endothelium and their accumulation at sites of allergen challenge are hallmarks of allergic inflammation. However, the molecular mechanisms mediating eosinophil adhesion under conditions of blood flow are not well understood. The present studies were performed to identify the receptors on human eosinophils involved in initiating adhesion to activated endothelium at physiologic shear rates in vivo. We have compared the relative contribution of L-selectin, VLA-4 (CD49d), and CD18 integrins in mediating eosinophil adhesion to microvascular endothelial cells in the rabbit mesentery by using intravital video microscopy. Eosinophils were found to roll in venules, but not arterioles, and this rolling could be stimulated by activation of endothelium with IL-1. In contrast to neutrophil rolling, which is predominantly L-selectin-dependent, eosinophil rolling was mediated by L-selectin, and also VLA-4. mAbs to L-selectin and VLA-4 alpha, but not CD18, significantly inhibited eosinophil rolling in vivo. The inhibition of VLA-4-mediated eosinophil rolling was not caused by modulation of eosinophil L-selectin or CD18 expression. This inhibition also was not caused by nonspecific inhibitory effect of the Abs studied, because the anti-VLA-4 mAbs inhibited eosinophil (VLA-4+) but not neutrophil (VLA-4-) rolling in the mesenteric venules. These results demonstrate that early events of eosinophil adhesion, i.e., rolling, are mediated by multiple adhesion receptors, including L-selectin and VLA-4, at physiologic shear rates in vivo.

The integrin alpha 9 beta 1 mediates cell attachment to a non-RGD site in the third fibronectin type III repeat of tenascin.

We have previously reported the sequence of the integrin alpha 9 subunit, a partner of the beta 1 subunit that is expressed in basal keratinocytes, hepatocytes, airway epithelial cells, and smooth and skeletal muscle. In the present study, we have stably expressed alpha 9 beta 1 on the surface of the human embryonic kidney cell line 293 and the human colon carcinoma cell line SW480 and used these transfected cells lines to identify ligand(s) for this integrin. Transfected cells did not appear to utilize alpha 9 beta 1 for attachment to the extracellular matrix proteins fibronectin, laminin, vitronectin, fibrinogen, thrombospondin, or type I or IV collagen. However, in contrast to mock transfectants, both 293 cells and SW480 cells expressing alpha 9 beta 1 adhered to intact chicken tenascin. By utilizing a variety of recombinant fragments of tenascin, we were able to localize the binding site for alpha 9 beta 1 to the third type III repeat. This repeat contains the arginine-glycine-aspartic acid (RGD) tripeptide that has been shown to serve as a binding site in tenascin for alpha v-integrins. However, the RGD site does not appear to be the binding site for alpha 9 beta 1, as the attachment of alpha 9 transfectants to this fragment was not inhibited by RGD peptide, nor by changing the RGD site to RAD or RAA.

Clonal analysis of primary marrow stroma: functional homogeneity in support of lymphoid and myeloid cell lines and identification of positive and negative regulators.

Stromal cells play an important role in regulating early hematopoiesis. How stromal cells exert their different functions and the factors that regulate stromal cells themselves remain to be elucidated definitively, however. We describe here a limiting dilution assay for primary stroma colonies from murine marrow. This system permits a critical analysis of stromal cell function and regulation on the clonal level. We report that stroma formation was dependent on an activity secreted by the long-term cultured stromal line AC-3.U. Differential ultrafiltration of AC-3.U supernatant (SN) suggests that this potentially novel activity is represented by molecules with apparent molecular weights (m.w.) of > 100 < 300 kD and > 300 kD. In contrast to the AC-3.U activity, hydrocortisone (HC) acts as a negative regulator of stroma colony formation. We used the stroma colony assay to analyze potential stromal cell heterogeneity. We found that most, if not all, primary stromal colonies supported expansion of both myeloid and lymphoid cell lines. In contrast, long-term cultured stromal cell lines differed not only among lines, but also on the level of sublines, in their ability to sustain myeloid and lymphoid cells. This intraclonal variation suggests that the heterogeneity of cell lines can be a reflection of ongoing culture adaptation. The functional homogeneity of primary stromal colonies, together with their susceptibility to regulators, indicates that the performance of primary stroma is subject to external control. The establishment of a clonal assay system has paved the way to analyze the molecules that regulate primary stroma and thereby hematopoietic cells.

Endothelial cell attachment and spreading on human tenascin is mediated by alpha 2 beta 1 and alpha v beta 3 integrins.

Human umbilical vein endothelial cells were found to attach and partially spread on human tenascin. The attachment of endothelial cells to tenascin results in elongated cells with interconnecting processes and is distinct from the flattened appearance of endothelial cells on fibronectin, collagen, vitronectin or laminin substrata, suggesting a role for tenascin in modulating cell adhesion and motility. Endothelial attachment to tenascin was partially inhibitable by the SRRGDMS peptide derived from human tenascin and completely inhibitable by anti-integrin antibodies to alpha 2 beta 1 and alpha v beta 3. Endothelial cell attachment to tenascin could be inhibited up to 80% with anti-alpha 2 and anti-beta 1 monoclonal antibodies P1E6 and P4C10, respectively, and this was associated with a complete loss in cell spreading. In contrast, pretreatment of endothelial cells with the anti-alpha v beta 3 monoclonal antibody LM609, resulted in a 35% inhibition in cell attachment but did not alter cell spreading. In combination the anti-alpha 2 and anti-alpha v beta 3 antibodies, could completely abrogate cell spreading and attachment to tenascin-coated surfaces. Affinity purification of 125I-labeled endothelial cell extract on a tenascin matrix column followed by immunoprecipitation with monoclonal antibodies to different integrin alpha and beta subunits resulted in the identification of alpha 2 beta 1 and alpha v beta 3 integrins, respectively, as tenascin binding receptors. Collagen affinity-purified alpha 2 beta 1 receptor from endothelial cells bound not only to collagen and laminin but also to tenascin in a radio receptor binding assay. The results demonstrate that alpha 2 beta 1 and alpha v beta 3 mediate distinct endothelial cell interactions with tenascin; cell spreading and cell binding, respectively. Binding by alpha v beta 3 is mediated by the SRRGDMS site on tenascin, whereas the alpha 2 beta 1 binding site remains undefined. The interaction of alpha 2 beta 1 and alpha v beta 3 with tenascin may be regulated in a cell type-specific manner as evidenced by the binding of endothelial cell alpha 2 beta 1 and alpha v beta 3 to tenascin, and the lack of binding by the same receptors on osteosarcoma MG63 to tenascin.

A novel tenascin type III repeat is part of a complex of tenascin mRNA alternative splices.

Sequence analysis of two human tenascin encoding cDNA clones from a cDNA library of the U251 glioblastoma cell line revealed the presence of a novel 276 bp tenascin type III fibronectin like repeat. This alternatively spliced type III repeat designated AD1 is located between the previously identified repeats 10 and 11 and has sequence homology with human, chicken and mouse tenascin type III repeats. These results show that tenascin has at least 16 consecutive fibronectin like type III repeats. PCR amplification of random primed mRNA with specific type III repeat primers revealed a pattern of multiple alternative splices of AD1 and flanking type III repeats. The alternative splice variants were confirmed by direct sequencing. Differences were observed in the expression of the various alternative splices of tenascin mRNA between tumor and normal cells and may thus indicate differences in tenascin isoform expression and function in normal and tumor cells. PCR and Southern analysis of genomic DNA indicate that AD1 is coded by a single exon present in both human and mouse genome.

Tenascin expression in human glioma cell lines and normal tissues.

Tenascin expression was evaluated in 21 human glioma cell lines and in normal adult tissue extracts by Western and Northern blotting. The cell lines differed in their relative expression of tenascin in the cell-associated and supernatant compartments. Glioma cell line tenascin production was not uniformly stimulated by changes in fetal bovine serum concentration in the growth media. In most glioma cell lines and normal tissue extracts, reducing Western blots and Northern blots revealed two tenascin species, respectively: a major 340 kDa polypeptide and a 9 kb RNA transcript accompanied by a less intense 250 kDa polypeptide and 7 kDa RNA species. In U-87 MG and in normal adult kidney extracts, however, the 250 kDa band and 7 kb transcript were more prominent. Quantitation of tenascin in the glioma lines revealed variable levels that were significantly higher than those in the tissue extracts.

Analysis of glycosaminoglycan substitution in decorin by site-directed mutagenesis.

Posttranslational glycosaminoglycan attachment to decorin, a chondroitin/dermatan sulfate proteoglycan, was studied by expression of a wild-type decorin cDNA and several mutagenized forms in two types of mammalian cells. Transfection of the wild-type cDNA resulted in the synthesis of an authentic chondroitin/dermatan sulfate proteoglycan similar to the decorin molecule synthesized by cultured human fibroblasts. Conversion of the serine residue that serves as the attachment site for the sole glycosaminoglycan chain in decorin to a threonine residue greatly reduced the efficiency of the glycosaminoglycan substitution. Less than 10% of the threonine-mutated core protein acquired a glycosaminoglycan chain, whereas most of the core protein was secreted without such substitution. Expression of cDNA in which an alanine residue had been introduced into the substituted serine position resulted in the secretion of core protein with no detectable glycosaminoglycan. Conversion to alanine of either one of the glycine residues that are adjacent to the substituted serine yielded the proteoglycan form of decorin. These results show that the xylosyltransferase responsible for the initiation of the glycosaminoglycan chain on the core protein can use a threonine residue for this substitution instead of a serine residue, but that such substitution is only partial, creating a "part-time" proteoglycan. Moreover, variations are possible in the sequence context of a glycosaminoglycan-substituted serine residue without loss of glycosaminoglycan substitution. The conformation of the substitution site may therefore be important for xylosyltransferase recognition.

Tenascin mediates cell attachment through an RGD-dependent receptor.

Tenascin is an extracellular matrix glycoprotein expressed in association with mesenchymal-epithelial interactions during development and in the neovasculature and stroma of undifferentiated tumors. This selective expression of tenascin indicates a specific role in cell matrix interactions. We now show that tenascin can support the adhesion of a variety of cell types, including various human tumor cells, normal fibroblasts, and endothelial cells, all of which can attach to a substrate coated with tenascin. Detailed studies on the mechanism of the tenascin-promoted cell attachment were carried out with the human glioma cell line U251MG. The attachment of these cells and others to tenascin were inhibited specifically by peptides containing the RGD cell attachment signal. Affinity chromatography procedures similar to those that have been used to isolate other adhesion receptors yielded a heterodimeric cell surface protein which bound to a tenascin affinity matrix in an RGD-dependent fashion. One of the subunits of this putative tenascin receptor comigrates with the beta subunit of the fibronectin receptor in SDS-PAGE and cross reacts with antibodies prepared against the fibronectin receptor in immunoblotting. These results identify the tenascin receptor as a member of the fibronectin receptor family within the integrin superfamily of receptors. The cell attachment response on tenascin is distinctly different from that seen on fibronectin, suggesting that cell adhesion and motility may be modulated at those sites where tenascin is expressed in the extracellular matrix.