The chemotactic attraction of human fibroblasts to a lymphocyte-derived factor.

A quantitative assay that measures fibroblast chemotaxis in vitro is described. Application of this technique has revealed that peripheral blood lymphocytes stimulated by antigen or mitogen in vitro produce a factor that is chemotactic for human dermal fibroblasts. This lymphocyte-derived chemotactic factor for fibroblasts (LDCF-F) is different from the lymphokine that is chemotactic for monocytes or macrophages. Macrophages are required for the generation of LDCF-F by T lymphocytes stimulated by phytohemagglutinin. The fibroblast chemotactic factor is heat stable (56 degrees C for 30 min), trypsin sensitive, and neuraminidase resistant. LDCF-F could function to attact connective tissue fibroblasts to sites at which cell-mediated immune reactions are occurring in vivo.

Collagen-and collagen peptide-induced chemotaxis of human blood monocytes.

The ability of collagen and collagen-derived peptides to act as chemotactic stimuli was investigated by in vitro chemotaxis assays. Native human and chick skin collagen (type I) and alpha-chains obtained from purified chick skin collagen were each chemotactic for human peripheral blood monocytes. In addition, smaller peptides obtained either by digesting native collagen with bacterial collagenase or by degrading purified alpha-chains with cyanogen bromide or pepsin were also chemotactic for monocytes. In contrast, native collagen, alpha-chains, and smaller collagen-derived peptides were not chemotactic for human neutrophils. Since collagen is degraded at sites of tissue damage and inflammation, our findings suggest the possibility that such collagen-derived degradation products might directly serve as chemotactic stimuli for human peripheral blood monocytes in vivo.

Autoimmunity to type II collagen an experimental model of arthritis.

We have found that intradermal injection of native type II collagen extracted from human, chick or rat cartilage induces an inflammatory arthritis in approximately 40% of rats of several strains whether complete Freund's adjuvant or incomplete Freund's adjuvant is used. Type I or III collagen extracted from skin, cartilage proteoglycans and alpha1(II) chains were incapable of eliciting arthritis, as was type II collagen injected without adjuvant. The disease is a chronic proliferative synovitis, resembling adjuvant arthritis in rats and rheumatoid arthritis in humans. Native type II co-lagen modified by limited pepsin digestion still produces arthritis, suggesting that type-specific determinants residing in the helical region of the molecule are responsible for the induction of disease. Since homologous type II collagen emulsified in oil without bacterial preparations regularly causes the disease, this new animal model of arthritis represents a unique example of experimentally-inducible autoimmunity to a tissue component.

Induction of fibroblast chemotaxis by fibronectin. Localization of the chemotactic region to a 140,000-molecular weight non-gelatin-binding fragment.

Plasma and cell-derived fibronectin are potent chemoattractants for human dermal fibroblasts in vitro. The chemotactic property of fibronectin resides in a major 140,000-mol wt non-gelatin-binding fragment of the native molecule. Human monocytes and neutrophils do not recognize fibronectin as a chemotactic stimulus. These findings suggest that fibronectin and perhaps certain fragments of fibronectin may function in vivo as a specific chemoattractant for fibroblasts and could, therefore, induce directional migration of fibroblasts to sites of tissue injury, remodeling or morphogenesis.

Stimulation of the chemotactic migration of human fibroblasts by transforming growth factor beta.

Transforming growth factor beta (TGF-beta) is a potent chemoattractant in vitro for human dermal fibroblasts. Intact disulfide and perhaps the dimeric structure of TGF-beta is essential for its ability to stimulate chemotactic migration of fibroblasts, since reduction with 2-ME results in a marked loss of its potency as a chemoattractant. Although epidermal growth factor (EGF) appears to be capable of modulating some effects of TGF-beta, it does not alter the chemotactic response of fibroblasts to TGF-beta. Specific polyvalent rabbit antibodies to homogeneously pure TGF-beta block its chemotactic activity but has no effect on the other chemoattractants tested (platelet-derived growth factor, fibronectin, and denatured type I collagen). Since TGF-beta is secreted by a variety of neoplastic and normal cells including platelets, monocytes/macrophages, and lymphocytes, it may play a critical role in vivo in embryogenesis, host response to tumors, and the repair response that follows damage to tissues by immune and nonimmune reactions.

Identification of an immunosuppressive epitope of type II collagen that confers protection against collagen-induced arthritis.

We have previously reported that collagen-induced arthritis can be suppressed by intravenous injection of native type II (CII) but not type I collagen. We have now identified denatured fragments of CII capable of suppressing collagen-induced arthritis and inducing tolerance. Purified CII was cleaved with cyanogen bromide (CB), and the major resulting peptides were isolated. Female DBA/1 mice were administered OVA, native CII, or one of the CB peptides, intravenously, before immunization with native CII, 6 wk after immunization, mice tolerized with CII and CB11 had a markedly lower incidence of arthritis compared with controls. There was a correlation between the overall antibody response and the incidence of arthritis. In addition, animals tolerized with either CII or CB11 had a decreased antibody response not only to CII, but also to each of the other CB peptides tested. To identify the epitope involved in suppression of arthritis, five synthetic peptides, 21-26 amino acids in length, corresponding to selected regions of CB11, were generated. Each of the peptides was injected intravenously into mice before immunization. Only one of these, CB11 122-147, was capable of suppressing arthritis. In addition, mice given the synthetic peptide CB11 122-147 neonatally were suppressed for arthritis and antibody responsiveness when immunized with CII at 8 wk of age. Thus, we have identified CB11 122-147 as an epitope of CII important in induction of tolerance and suppression of disease. Further experiments narrowing down the pivotal amino acids for the immunogenicity of this epitope and the role this epitope plays in induction and regulation of disease will enhance our understanding of how the immune response to collagen affects autoimmune arthritis.

Human HLA-DR beta gene hypervariable region homology in the biobreeding BB rat: selection of the diabetic-resistant subline as a rheumatoid arthritis research tool to characterize the immunopathologic response to human type II collagen.

Collagen arthritis (CA), an autoimmune model of rheumatoid arthritis (RA), has been studied in various animals. However, it has not been studied in an animal with a genetic background relevant to RA. We selected rats from a diabetic-resistant (DR) subline of the diabetic BB rat because they have an autoimmune disease-prone background, but not the immunodeficiencies of the diabetic BB rat, and the third hypervariable region (HVRIII) of the BB RT1.D beta gene appeared to encode a nucleotide sequence of the human HLA DR beta gene, which has been reported to be associated with susceptibility to RA. We synthesized oligonucleotide primers flanking the RT1.D beta HVRIII, cloned polymerase chain reaction-amplified DNA into M13mp18, and confirmed the presence of the susceptibility sequence (SS) (RRRAA) by the dideoxy sequencing method in a colony of DR BB/Wor-UTM rats. When immunized with human type II collagen (CII) in incomplete Freunds adjuvant (IFA), arthritis developed rapidly by day 10 with 100% incidence. Light and electron microscopy revealed an unusually severe and aggressive, bidirectional pattern of cartilage resorption by synovial and subchondral mononuclear and multinucleated inflammatory cells. These findings coincided with a predominant humoral response to the cyanogen bromide (CB) 11 fragment of the human CII molecule by the pathogenic IgG2a isotype. This study provides further support to the role of CA as a relevant RA model, the specific roles of the CB11 fragment as a major site of arthritogenic epitopes, and of antibody mechanisms in the pathogenesis of CA. Furthermore, the identification of an RA SS in an immune response gene of the DR BB rat presents a novel opportunity to determine with an animal model the role of other antigens as well as this SS in RA.

Interleukin 1 stimulation of collagenase production by cultured fibroblasts.

Interleukin 1 is a monokine that exerts biological effects on a variety of target cells in vitro. In this report, interleukin 1 has been found to be capable of stimulating collagenase production by cultured dermal fibroblasts. The concentrations of interleukin 1 that stimulate fibroblast collagenase production are similar to those that stimulate mouse thymocyte proliferation. Analyses by high performance liquid chromatography indicate that interleukin 1, rather than a contaminating monokine, is responsible for this effect on fibroblasts. Interleukin 1, released in vivo by macrophages infiltrating sites of tissue damage or inflammation, may function to stimulate the release of collagenase by connective tissue fibroblasts.

Formation of multinucleated giant cells from human monocyte precursors. Mediation by a soluble protein from antigen-and mitogen-stimulated lymphocytes.

Multinucleated giant cells are associated with granulomas arising from immunological and nonimmunological inflammatory reactions. They are an integral part of the host immune response to chronic infectious diseases. In the present study we have demonstrated that human lymphocytes when stimulated by specific antigens of T cell mitogens produce a soluble factor that causes peripheral blood monocytes to fuse and form multinucleated giant cells in vitro. Production of the giant cell factor by antigen-stimulated peripheral blood lymphocytes correlates with the existence of cell-mediated immunity to specific antigen. Monocyte-depleted blood lymphocytes, but not purified blood monocytes, produce the giant cell factor when cultured with antigens or T cell mitogens. Gel filtration and physiochemical studies indicate that the lymphocyte-derived giant cell factor is a heat-labile protein of approximately 60,000 mol wt. These findings suggest that multinucleated giant cells in granulomas may be formed by fusion of circulating monocytes in response to the release of a 60,000-mol wt protein from antigen-stimulated T lymphocytes.

Type II collagen-induced arthritis in rats. Passive transfer with serum and evidence that IgG anticollagen antibodies can cause arthritis.

We have found that serum from rats with type II collagen-induced arthritis, when fractionated with 50% ammonium sulfate and concentrated, would transfer arthritis to nonimmunized recipients. The arthritis in recipients developed within 18-72 h and displayed all of the major histopathologic characteristics of the early lesion in immunized animals but was transient and less severe. Although consideration was given to the possibility that a circulating immune complex was involved, no evidence of such a complex was detected. Further fractionation of the serum yielded an IgG anticollagen antibody that was fully active in transferring disease. The antibody's reaction was inhibited by the native bovine type II collagen used for immunization of donors and the antibody strongly cross-reacted with homologous type II collage but not with denatured collagen. These studies demonstrate that arthritis in rats can be induced with anti-type II collagen antibodies and suggest that an autoimmune process is involved. Because antibodies to collagen have also been detected in human rheumatic diseases, further investigation of the characteristics of collagen antibodies capable of inducing arthritis seems warranted.

Auricular chondritis in rats. An experimental model of relapsing polychondritis induced with type II collagen.

Outbred Wistar rats immunized with native type II collagen developed ear lesions resembling those of human relapsing chondritis. As in human disease, these lesions were characterized by intense chondritis, positive immunofluorescence reactions to IgG and C3, and circulating IgG reactive with native type II collagen. Furthermore, electron-dense deposits were seen near the surface of chondrocytes and corresponded with deposits of IgG and C3. These observations suggest a causal relation between humoral immunity to type II collagen and auricular chondritis in the rat and support the hypothesis than human relapsing polychondritis is an autoimmune disease mediated by immunity to type II collagen.

Alterations in collagen production in mixed mononuclear leukocyte-fibroblast cultures.

The cell-cell interactions between fibroblasts and mononuclear leukocytes (MNL) which promote alterations in collagen accumulation were examined using a system of co-culture of human fibroblasts and peripheral blood MNL. The stimulation of collagen production was optimal after 48 h of co-culture and the increase in collagen correlated directly with the number of MNL added. The enhancement of collagen production was seen in both autologous and allogeneic co-cultures. Stimulation of non-collagenous protein was also noted. Co-culture supernatants contained soluble substances that were capable of stimulating collagen production, although they stimulated collagen production to a lesser degree than direct co-culture. Fractionation of these supernatants on Sephadex G-200 revealed a predominant area of stimulatory activity at 160,000 mol wt. Lesser areas of activity were noted at molecular weights of 80,000 and 25,000. Determination of the types of collagen produced by fibroblasts during co-culture with MNL showed that the ratio of type I:III collagen was decreased. These alterations in both the quantitative and qualitative accumulation of collagen mimic the changes often seen in wound healing and early inflammation suggesting that cellular interactions between fibroblasts and MNL may be important in the modulation of collagen production in normal and pathologic states.

Heterogeneity among human collagenases demonstrated by monoclonal antibody that selectively recognizes and inhibits human neutrophil collagenase.

The heterogeneity of human collagenases has been examined using a monoclonal antibody to neutrophil collagenase. This antibody inhibited collagenase activity and, when covalently coupled to Sepharose, bound both latent and active enzyme. Although human neutrophil collagenase was inhibited by the antibody, the activity of human skin and rheumatoid synovial collagenase was not significantly diminished in the presence of the antibody. Competitive inhibition studies also differentiated between these collagenases. Only human neutrophil collagenase effectively blocked the antibody in a competitive enzyme-linked immunosorbent assay while skin and rheumatoid synovial collagenase again failed to interact with the antibody. The unequivocal recognition of neutrophil collagenase as an immunologically distinct entity from other collagenases supports the hypothesis that neutrophil collagenase is a separate gene product from fibroblast or synovial collagenase.

Induction of arthritis in monkeys by immunization with type II collagen.

Immunization of two cynomolugus and three rhesus monkeys with purified type II collagen resulted in the development of polyarthritis. Arthritis first became clinically apparent 7 wk after primary immunization and persisted for 16 mo. Radiologic examination of the limbs demonstrated soft tissue swelling with severe joint destruction including loss of cartilage and bone. Involved joints eventually became ankylosed with permanent loss of some motion. All of the monkeys developed a response to the immunizing collagen as determined by ELISA of serum for antibodies. Arthritis was associated with weight loss and constitutional symptoms, including lethargy and refusal to eat. One monkey became so debilitated that it was necessary to euthanize it. Histologic examination of the joints showed synovial hypertrophy with pannus formation. A control monkey immunized with type I collagen suffered no apparent ill effects.

Purification and properties of M protein extracted from group A streptococci with pepsin: covalent structure of the amino terminal region of type 24 M antigen.

M protein was extracted from type 24, group A streptococci with pepsin at pH 5.8 and was further purified by ammonium sulfate precipitation, ribonuclease digestion, ion-exchange chromatography, and isoelectric focusing. The purified pepsin extract of M (pep M) protein was shown to be free of nontype-specific immunoreactivity in (a) complement fixation tests with heterologous M antiserum, (b) skin tests in normal adult guinea pigs, and (c) passive hemagglutination tests for the presence of lipoteichoic acid sensitizing or antigenic activity. The pep M24 was highly immunogenic; two of three rabbits developed opsonic antibody titers of 1:256 and the third a titer of 1:32 6 wk after a single injection of 100-pg doses of pep M24 emulsified in complete Freund's adjuvant. The antisera lacked nontype-specific antibodies and produced single precipitin lines in agar gel diffusion tests against crude HC1 extracts of the homologous M protein. Thus, the type-specific antigenic determinant(s) of type 24 M protein appears to be separable from immunotoxic, cross-reactive antigens without loss of immunogenicity in rabbits. The mobility of pep M24 upon electrophoresis in 10 percent sodium dodecyl sulfate pelyacrylamide gel was consistent with an average mol wt of 33,500 daltons. Amino acid analysis demonstrated a predominance of alanine, followed by glutamic acid, lysine, leucine, and aspartic acid. Pep M24 contained an estimated six to seven methionine residues and approximately ten phenylalanine residues per molecule. No other aromatic amino acids were detected. Automatic Edman degradation of pep M24 yielded the sequence of the first 29 amino acids (the amino terminal amino acid being valine) of the amino terminal region of the molecule. The detection of only one new amino acid at each step of Edman degradation confirmed the homogeneity of the purified pep M24.

Collagen-induced arthritis in mice. Localization of an arthritogenic determinant to a fragment of the type II collagen molecule.

Purified chick type II collagen was cleaved with cyanogen bromide (CB), and the resulting peptides isolated and renatured. Sera from arthritic DBA/1 mice, immunized with chick type II collagen, were tested for reactivity with each peptide. The sera preferentially recognized peptides 11, 10, and 8, in that order. Some reactivity was also detected to peptides 9, 7, and 12. Because arthritis depends upon binding of antibody to autologous type II collagen in the joint, sera were also tested for reactivity with mouse type II collagen. There was a strong positive correlation between reactivity with peptide 11 and reactivity with mouse collagen, but no correlation was found with any of the other peptides. Peptides 11, 10, and 8 were also used for immunization. Antibodies were detected in response to each of these peptides, but arthritis developed only in mice immunized with peptide 11. We conclude that a major immunogenic and arthritogenic epitope on type II collagen resides in the region of the molecule represented by CB peptide 11.

An HLA-DR1 transgene confers susceptibility to collagen-induced arthritis elicited with human type II collagen.

Rheumatoid arthritis (RA) is an autoimmune disease that is strongly associated with the expression of several HLA-DR haplotypes, including DR1 (DRB1*0101). Although the antigen that initiates RA remains elusive, it has been shown that many patients have autoimmunity directed to type II collagen (CII). To test the hypothesis that HLA-DR1 is capable of mediating an immune response to CII, we have generated transgenic mice expressing chimeric (human/mouse) HLA-DR1. When the DR1 transgenic mice were immunized with human CII (hCII), they developed a severe autoimmune arthritis, evidenced by severe swelling and erythema of the limbs and marked inflammation and erosion of articular joints. The development of the autoimmune arthritis was accompanied by strong DR1-restricted T and B cell responses to hCII. The T cell response was focused on a dominant determinant contained within CII(259-273) from which an eight amino acid core was defined. The B cell response was characterized by high titers of antibody specific for hCII, and a high degree of cross-reactivity with murine type II collagen. These data demonstrate that HLA-DR1 is capable of presenting peptides derived from hCII, and suggest that this DR1 transgenic model will be useful in the development of DR1-specific therapies for RA.

Modulation of fibroblast functions by interleukin 1: increased steady-state accumulation of type I procollagen messenger RNAs and stimulation of other functions but not chemotaxis by human recombinant interleukin 1 alpha and beta.

Interleukin-1 (IL-1) is synthesized by and released from macrophages in response to a variety of stimuli and appears to play an essential role in virtually all inflammatory conditions. In tissues of mesenchymal origin (e.g., cartilage, muscle, bone, and soft connective tissue) IL-1 induces changes characteristic of both destructive as well as reparative phenomena. Previous studies with natural IL-1 of varying degrees of purity have suggested that it is capable of modulating a number of biological activities of fibroblasts. We have compared the effects of purified human recombinant (hr) IL-1 alpha and beta on several fibroblast functions. The parameters studied include cell proliferation, chemotaxis, and production of collagen, collagenase, tissue inhibitor of metalloproteinase (TIMP), and prostaglandin (PG) E2. We observed that hrIL-1s stimulate the synthesis and accumulation of type I procollagen chains. Intracellular degradation of collagen is not altered by the hrIL-1s. Both IL-1s were observed to increase the steady-state levels of pro alpha 1(I) and pro alpha 2(I) mRNAs, indicating that they exert control of type I procollagen gene expression at the pretranslational level. We found that both hrIL-1 alpha and beta stimulate synthesis of TIMP, collagenase, PGE2, and growth of fibroblasts in vitro but are not chemotactic for fibroblasts. Although hrIl-1 alpha and beta both are able to stimulate production of PGE2 by fibroblasts, inhibition of prostaglandin synthesis by indomethacin has no measurable effect on the ability of the IL-1s to stimulate cell growth or production of collagen and collagenase. Each of the IL-1s stimulated proliferation and collagen production by fibroblasts to a similar degree, however hrIL-1 beta was found to be less potent than hrIL-1 alpha in stimulating PGE2 production. These observations support the notion that IL-1 alpha and beta may both modulate the degradation of collagen at sites of tissue injury by virtue of their ability to stimulate collagenase and PGE2 production by fibroblasts. Furthermore, IL-1 alpha and beta might also direct reparative functions of fibroblasts by stimulating their proliferation and synthesis of collagen and TIMP.

Collagen molecules: distribution of alpha chains.

To ascertain the distribution of alpha chains within the collagen molecule, intramolecular cross-links were introduced into tropocollagen by reacting formaldehyde with dilute homogeneous molecular dispersions of collagen extracted from lathyritic guinea pig skin. Denaturation, chromatography on carboxymethyl cellulose, measurement of molecular weight, and analyses of the amino acids of the cross-linked product indicate that most, if not all, of the collagen molecules consist of two alpha1 chains and one alpha2 chain [(alpha1)(2)(alpha2)(1)].

Collagen-induced platelet aggregation: involement of an active glycopeptide fragment (alpha1-CB5).

It is widely held that the tertiary structure of collagen is essential for induction of platelet aggregation. However, we have found that the purified alpha1 chain prepared from denatured chick skin collagen aggregates platelets. This activity appears to be confined to a distinct region of the molecule representing less than 4 percent of the length of the alpha1 chain. Of all of the cyanogen bromide peptides of the alpha1 chain tested, only one (alpha1-CB5) was active. This glycopeptide, devoid of any ordered tertiary structure, contains only 36 amino acids and one residue of O-alpha-D-glucopyranosyl-(1 --> 2)-O-beta-D-galactopyranosyloxy-(1 --> 5)-lysine (Glc-Gal-Hyl). Blocking experiments strongly suggest that the Glc-Gal-Hyl is one of the structural determinants involved in collagen-induced platelet aggregation.