Growth factors, extracellular matrix, and oncogenes in scleroderma.
Explore the source record for details and available documents.
Biomedical subjects
Publications and source records attributed to R E Gay.
Explore the source record for details and available documents.
In view of the important role of interstitial collagenase in the pathogenesis of rheumatoid arthritis (RA), we studied the expression of fibroblast-type collagenase in rheumatoid synovium and searched for its potential transcription factors, namely the oncoprotein c-fos and the early-growth-response gene-1 (egr-1), an inducible zinc-finger encoding gene. Elevated levels of RNA sequences complimentary to c-fos and egr-1 cDNA probes could be detected in cytoplasmic extracts of collagenase-expressing synovial fibroblast-like cells when compared to equivalent RNA amounts isolated from control fibroblasts. Utilizing immunocytochemistry, immunoreactivity for c-fos oncoprotein was found in 13 of 19 joint specimens obtained from patients with active RA. These oncoprotein data were positively correlated to the collagenase expression in the same specimens. Moreover, immunohistochemical analysis confirmed the localization of both oncoprotein c-fos and fibroblast-type collagenase within synovial fibroblast-like cells attached to bone erosions.
Female MRL/lpr mice were treated with I.P. doses of methotrexate (MTX) and 10-deazaaminopterin (DAAM) in the range of 1 to 100 mg/kg body weight/week, in two equally divided doses. Treatment began at 7 weeks of age and continued to 30 weeks of age. Joint histopathology scores were tightly correlated with skin lesion-proteinuria scores at 30 weeks of age. MTX at levels of 5, 25, and 100 mg/kg body weight/week and DAAM at a level of 25 mg/kg body weight/week significantly reduced skin lesion-proteinuria scores below controls in a dose dependent manner. Animals receiving MTX at 25 mg/kg body weight/week had a significantly longer median life span and animals receiving MTX at 100 mg/kg body weight/week had a greater than 15% suppression of growth when compared with controls. Longevity and skin lesion-proteinuria scores appeared to be good indicators of drug efficacy while growth suppression appeared to be a good indicator of drug toxicity.
Typical erosions of articular joint structures in rheumatoid arthritis and in the spontaneous destructive hind-limb arthropathy of autoimmune MRL-lpr/lpr (MRL/l) mice occur predominantly in areas contiguous with proliferating synovial lining cells, suggesting release of proteolytic enzymes from these cells. Synovial lining cells were isolated from arthritic MRL/l mice, and the spontaneous expression of the interstitial procollagenase and its potential transcriptional factors, egr-1 and c-fos, was examined in vitro. The data indicate that basal collagenase RNA expression was stronger in MRL/l cells than in virus-transformed cells. Moreover, elevated RNA levels of the c-fos gene could be detected in the collagenase-expressing synovial lining cells in vitro. In a related immunohistochemical study, collagenase was detected in situ in proliferating synovial lining cells as well as in chondrocytes of the first stage of pathological changes in the MRL/l mouse arthropathy.
The mechanism of the binding of IgA to the mesangium in IgA nephropathy (IgAN) is unknown. Interactions between IgA and components of the mesangial matrix may contribute. We measured by enzyme-linked immunosorbent assay the binding of serum IgA, IgG, and IgM from patients with IgAN, human immunodeficiency virus type I (HIV) infection, and healthy controls to purified native collagen types I to VI, and to an extract of normal kidney tissue. HIV infection is an appropriate disease control because of the lack of mesangial IgA deposits, despite high serum levels of IgA and IgA1-containing immune complexes. Increased levels of IgA-binding to collagen types I and V and the kidney extract were found only in IgAN. Both IgAN and HIV-infected patients had increased IgA-binding to collagen types II, III, and VI. Preabsorption of the sera with gelatin substantially reduced the IgA-binding to collagen types I to IV, but not to types V and VI. This finding suggests that the binding to collagen type V is not fibronectin-mediated, but may reflect autoantibody formation. Thus, fibronectin-mediated IgA-collagen interactions are not specific for IgAN, and their pathogenetic role is questionable. The role of IgA anti-collagen type V antibodies requires further study.
In view of the important role of fibroblast-type collagenase in the pathogenesis of periodontal disease (PD), we investigated the expression of this metalloproteinase in primary cultures of non-stimulated fibroblasts dissected from gingival tissues of patients with generalized moderate and localized severe chronic adult PD. Enhanced hybridization signals for collagenase RNA were observed in 8/8 PD-cases when compared with equivalent RNA amounts extracted from normal fibroblasts. Since both the proto-oncogene c-fos and the "early growth response" gene egr-1 might be involved in the transcriptional regulation of the collagenase gene expression in vivo, we also compared the relative expression of both potential transcriptional factors with collagenase RNA in the same fibroblast cytoplasmic extracts. Hybridization signals indicated elevated RNA amounts for c-fos in 8/8 PD-cases and for egr-1 in 7/8 PD-cases when compared with the cells from non-inflamed tissue. In periodontitis gingival tissue specimens, immunolocalization of collagenase could be confirmed in fibroblasts, macrophages and epithelial cells in situ. Collagenase label was not widely distributed within the tissues, but concentrated at the interface between epithelium and connective tissue. The data provide the first evidence that gingival fibroblasts producing elevated levels of collagenase RNA amounts express also c-fos and egr-1 indicating a crucial role for both genes in cellular proliferation and collagenase expression in gingival and periodontal tissue destruction in vivo.
Rheumatoid arthritis (RA) is a chronic systemic disorder that is dominated by the debilitating sequelae associated with the progressive destruction of articular joints. The molecular and cellular basis of rheumatoid joint destruction is characterized by an abnormal expression of oncogenes modulating cellular proliferation and the induction of lysosomal and metalloproteinases. Based on the observation that the synovial hyperplasia in RA is associated with the proliferation of transformed-appearing synovial lining cells and an overexpression of such oncogenes, the possibility that a hitherto unknown HTLV related retrovirus is involved in the etiopathogenesis of RA is discussed.
Marrow stromal cells are important in normal myelopoiesis and support growth of leukemia/lymphoma (LL) cells in vitro. We have previously described the heterotypic adherence of a human B-lymphoblastic cell line (UTMB-460) to marrow stromal cells (MSC). We have extended these observations to a human T-lymphoblastic cell line (CEM) and characterized the heterotypic adherence of B- and T-lymphoblastic cell lines to human MSC. Electron microscopy demonstrated UTMB-460 cells were in very close apposition to the MSC, but no specific intercellular junctions were noted. Under the conditions employed, these MSC express extracellular fibronectin, collagen types I and IV, intracellular laminin, and vimentin, but no factor VIII-R antigen. In addition, the MSC had receptors for the lectin Ulex europaeus agglutinin I. UTMB-460 and CEM cells do not adhere to extracellular matrix (ECM) proteins secreted by the MSC, i.e., fibronectin, collagen types I, III, or IV, or laminin. Monoclonal antibodies (MoAbs) against CD11a, CD11b, CD18, and CD54 and a polyclonal anti-human fibronectin antibody do not inhibit attachment of either B- or T-lymphoblastic cells to MSC. Peptides GRGES and GRGDS did not inhibit adherence of UTMB-460 and CEM cells to MSC. In contrast, the anti-vascular cell adhesion molecule (VCAM)-1 MoAb (4b9) caused significant inhibition (p < 0.01) of the adherence of both UTMB-460 and CEM cells to normal human MSC monolayers. These data suggest: (1) that MSC to which lymphoblastic cells adhere are specialized mesenchymal cells; (2) that the membrane interactions between T- and B-lymphoblastic cells and MSC involve close apposition of cell membranes of MSC and the lymphoblastic cells; (3) that the heterotypic adherence between B- and T-lymphoblastic cell lines (UTMB-460 and CEM) and MSC does not involve the RGD recognition sequence of the integrin family, the B2 leukocyte integrins, CD44, LAM-1, or the ECM proteins examined; and (4) that VCAM-1 may at least be partially responsible for heterotypic adherence between human MSC and B- and T-lymphoblastic cells.
Explore the source record for details and available documents.
Based on the concept that proteolytic enzymes, like cathepsins, are associated with tissue destruction, we investigated the expression of the matrix-degrading cysteine proteinase cathepsin B in synovial tissues from the joints of patients with rheumatoid arthritis. The data indicate an enhanced transcription of cathepsin B in synovial cells when compared with normal fibroblasts, cathepsin B-producing epithelial tumor cells (SW1116), or fibroblasts derived from inflamed tonsils. Immunolocalization of cathepsin B appeared to be restricted mainly to the synovial cells attached to cartilage and bone at sites of rheumatoid joint erosion.
In order to determine whether collagen autoantibodies are present in sera from patients with vascular diseases, a highly sensitive solid-phase radioimmunoassay was utilized to evaluate the presence of antibodies against collagen types I to VI and laminin in sera of 20 patients with vasculitis and 20 patients with systemic lupus erythematosus (SLE). Autoantibodies to interstitial collagen types I and II were noted in 20 and 35% of the patients, respectively. Most significantly, 70% of the polyarteritis nodosa patients and 55% of the total vasculitis group had autoantibodies to collagen type IV. In SLE, autoantibodies to collagen type V were detected in 70% and to collagen type IV in 85% of the patient sera. These data indicate that basement membrane and basement membrane-associated collagens may become immunogenic in patients with vasculitis and SLE. The role of these autoantibodies is most likely a result of endothelial damage secondary to the initial inflammatory disease process. However, it can be concluded from the present studies that collagen types IV and V are involved in the immune response and thereby may perpetuate further vascular damage.
Monoclonal antibodies produced against pepsin-solubilized newborn rat skin type V collagen [alpha 1(V)]2 alpha 2(V), and chondrosarcoma type XI collagen [alpha 1(XI) alpha 2(XI) alpha 3(XI)] are used to localize the collagens in sections of the chondrosarcoma as well as the normal rat knee joint by indirect immunofluorescence. Immunostaining for type V collagen shows strong cellular staining of chondrocytes; while the interstitial matrix as well as the lacunae are not stained. In contrast, antitype XI stains not only chondrocytes, but the extracellular compartments as well. In ELISA, rat anti-type XI collagen reacts with its native antigen, but does not cross-react with native types I, II, III, or V collagen from rat. The distinct locations of type V and XI collagens in cartilaginous tissue suggest varied functional roles for these constituents in the tissue.
To substantiate the finding of interstitial myocardial fibrosis in the transplanted heart and to characterize the collagen profile of the transplanted heart, we studied endomyocardial biopsy specimens from 30 heart transplants and four heart-lung transplants at 1 to 82 months after transplantation. Indirect immunofluorescent techniques with affinity-purified antibodies for collagen types I, III, IV, and V were used. The degree of interstitial collagen present was scored. The amount of type I collagen was increased in transplants from distant donors (mean ischemia time, 154 minutes) compared with those from on-site donors (mean ischemic time, 59 minutes): collagen scores 1.1 and 1.7, respectively (P less than .01). There was a trend toward a positive correlation, not statistically significant, between cyclosporine dose and collagen III deposition (r = .35), collagen IV (r = .38), and collagen V (r = .56). There was a negative correlation between number of rejection episodes and mean cyclosporine dose (r = -.41) and amount of collagen III (r = -.42) or collagen IV (r = -.42). No correlations were found between collagen deposition and any other variables studied. These results confirm the mixed nature of the collagen deposited and suggest that some fibrosis is related to cyclosporine administration rather than to the number of prior healed rejection episodes.
Based on the observation that rheumatoid joint destruction is related to the presence of transformed-appearing proliferating synovial lining cells attached to cartilage and bone at the site of early destruction, we searched for the expression of proliferation- and transformation-associated oncoproteins in synovial tissues from patients with early destructive rheumatoid arthritis (RA). Immunolocalization of Ras and Myc proteins was found in about 70% of the RA cases and was restricted to the proliferating synovial lining cells. The cysteine proteinase, cathepsin L, which has been shown to be the major ras-induced protein in ras-transformed murine NIH 3T3 cells, was detected in 50% of the RA cases, predominantly in synovial cells attached to cartilage and bone at the site of joint destruction. Moreover, utilizing cytoplastic dot hybridization analysis, we demonstrated the presence of RNA sequences complementary to human cathepsin L in primary cultures of human synovial cells from RA joints and complementary to murine cathepsin L in synovial lining cells derived from MRL/l mice developing spontaneously a RA-like disease. Significant levels of ras oncogene transcripts and products in human RA synovial cells associated with an increased expression of the cathepsin L gene indicate that this collagen-degrading enzyme may contribute to the destruction of cartilage and bone in RA.
Fibroblast-type collagenase, a neutral secretory metalloproteinase capable of cleaving interstitial collagen types I-III, is expressed by a number of different cell types including fibroblasts, macrophages, osteoblasts, and keratinoyctes. To elucidate the secretory pathway of this enzyme, we examined the ultrastructural localization of this metalloproteinase in cultured human gingival fibroblasts, particularly the routing of the enzyme from the Golgi cisternae to the cell surface utilizing rabbit polyclonal antibodies raised against human fibroblast (pro) collagenase. For this purpose, one percent glutaraldehyde followed by gentle permeabilization with saponin gave superior preservation of both cellular morphology and intracellular antigenicity. At the light microscopic level, the reacting antibodies visualized by immunofluorescence and immunoperoxidase staining were localized intracellularly in the perinuclear region reflecting the Golgi apparatus. Immunoelectronmicroscopy using the pre-embedding technique and peroxidase or immunogold staining revealed electron dense label in large vacuoles indicating extended cisternae of the Golgi field. Vesicles were noted leaving the plasma membrane in long extensions. Moreover, intact vesicle containing the antibody reaction product appeared outside the membrane. In addition, most extracellular vesicular structures appeared empty of label suggesting that the collagenase had been liberated into the extracellular space. The latter observation was supported by the fact that the label was found also on the extracellular surface of the cells indicating a (re)association of collagenase with the outer cell membrane. These data demonstrate that the pathway of interstitial collagenase in human gingival fibroblasts is similar to that of other secretory proteins.
We were able to demonstrate type II collagen in synovial phagocytes of osteoarthritis (OA) and rheumatoid arthritis (RA) patients, using a monoclonal antibody to human type II collagen and immunoperoxidase staining. In addition, using immunoelectron microscopy, we demonstrated labeled fragments in synovial phagocytes of both RA and OA patients. This immunohistochemical assay may prove to be a sensitive indicator of cartilage erosion in patients with OA and RA.
A new animal model for human rheumatoid arthritis is described, and the unsolved questions regarding the mechanism of primary joint destruction are discussed. Following an analysis of the types of cells and antibodies found in joints affected by rheumatoid arthritis, it is concluded that both expression of oncogenes and the presence of retroviral sequences detectable by monoclonal antibodies to HTLV I p19 and p24 sequences are associated with early abnormal proliferation of apparently transformed cells at the site of initial cartilage and/or bone destruction.
We studied the histopathologic, immunohistochemical, and ultrastructural features of the extracellular matrix (ECM) in 20 immunologically defined large-cell lymphomas, including immunoblastic sarcoma of B cells (three patients), peripheral T cell lymphoma (five patients), large non-cleaved follicular center cell (FCC) lymphoma (nine patients), and non-marking large-cell lymphoma (three patients). Immunohistochemical studies were performed with antibodies to laminin, fibronectin, and collagen types I, III, IV, and V. The immunologically defined subgroups demonstrated characteristic differences in ECM in light microscopic appearance, composition, and ultrastructural features. Immunoblastic sarcomas of B cells showed delicate intercellular bands that were apparent only at high power but were distinct in exhibiting focal staining for basement membrane elements (laminin and type IV collagen) in addition to type I collagen and fibronectin. Electron microscopically, no basal lamina were apparent, although the collagen fibers were embedded in a dense matrix not seen in the other lymphomas. All peripheral T cell lymphomas exhibited a packeting pattern of intercellular bands and were distinguished by the frequent presence of intense pericellular staining for type V collagen as well as by focal pericellular staining for types I and III collagen. The latter finding corresponded to extensive areas of direct contact between tumor cells and the ECM by electron microscopy. The large non-cleaved FCC lymphomas and the non-marking large-cell lymphomas demonstrated both diffuse and compartmentalizing arrangements of intercellular bands that frequently coexisted and stained predominantly for fibronectin and types I and III collagen. All groups demonstrated myofibroblasts and fibroblasts partially or completely separating the ECM from tumor cells, suggesting that most of the ECM is part of a reaction to these lymphomas. These studies show more variation in light microscopic appearance, composition, and ultrastructural relationships of the intercellular and pericellular ECM than was apparent in earlier studies of cleaved FCC lymphomas.