Etiopathogenesis of the rheumatoid arthritis–like disease in MRL/1 mice. I. The histomorphologic basis of joint destruction

MRL/1 mice spontaneously develop a hindlimb arthropathy, as well as a number of immunologic abnormalities, including circulating rheumatoid factors. Although previous studies have suggested that this arthropathy is primarily an inflammatory process, we performed a comprehensive histomorphologic study which indicated that inflammation is a late manifestation of MRL/1 arthritis. The pathologic changes that occur in the joints of these mice can be divided into 3 stages. The first stage develops between the ages of 7 and 13 weeks and consists of synovial cell proliferation in the joint recesses. The second stage is characterized by continued proliferation of synovial cells which take on an appearance similar to that of transformed mesenchymal cells. The earliest destructive changes occur in the second stage and include marginal erosions, followed soon after by progressive destruction of articular and meniscal cartilage. The final stage is characterized by a diminution of synovial cell proliferation, extensive cartilage destruction, formation of scar tissue and fibrocartilage, and a very moderate infiltration of the synovial stroma by mononuclear and polymorphonuclear inflammatory cells. Throughout the disease progression there is a striking dissociation between inflammatory cell infiltration or exudation and tissue destruction. The histomorphologic similarities between human rheumatoid synovitis and the arthritis of MRL/1 mice, as well as the presence of rheumatoid factors, make this mouse strain an excellent model for studying human rheumatoid arthritis.     

Etiopathogenesis of rheumatoid arthritis-like disease in MRL/1 mice: II. Ultrastructural basis of joint destruction

MRL/1 mice develop a spontaneous hindlimb arthropathy characterized by proliferation of synovial cells and by dissociation between early destruction of articular tissue and the presence of inflammatory cell infiltration. To characterize the ultrastructural details of the synovial cells of these mice, knee joints from MRL/1, MRL/n, and BALB/c mice were examined by light and electron microscopy. Since the proliferating synovial cells of MRL/1 mice resemble the previously described proliferative synovial cells seen in histopathologic specimens from early rheumatoid arthritis, further study of these cells may provide new insights into the pathogenesis of early joint tissue destruction in human rheumatic disease.     

Contribution of synovial mesenchymal cells to the pathogenesis of rheumatoid arthritis.

Rheumatoid joint destruction is caused by (1) enzymatic digestion from articular surfaces of cartilage, (2) pannus formation, and (3) lysis of the matrix by activated chondrocytes. Pannus, a vascular and fibrous granulation tissue arising from the perichondral synovial membrane, extends onto cartilage surfaces as a layer of morphologically quiescent fibroblastic mesenchymal cells. Pannus subsequently starts invasion into cartilage matrix with the appearance of macrophagelike cells. Synovial mesenchymal cells are thought to play important roles in the pathogenesis of rheumatoid joint destruction in relation to la expression and antigen presentation as well as the elaboration of inflammatory cytokines such as interleukin (IL)-1, IL-6, and tumor necrosis factor-alpha. By experimentally inducing antigen-induced arthritis in H-2-c-fos transgenic mice whose immunoglobulin G antibody response against immunizing antigen was defective, the investigators produced destructive arthritis without lymphocyte infiltration. The only cells invading the joints in these mice were similar to a previously recognized subset of human synovial cells that had a mesenchymal appearance. These mesenchymal cells invaded the cartilage matrix upon in vitro culture. The role of synovial mesenchymal cells in the pathogenesis of rheumatoid joint destruction is discussed.     

Physiopathology of haemophilic arthropathy

Summary.  Haemophilic arthropathy, which shares some clinical and biological injury characteristics with rheumatoid arthritis, is characterized by two main features: chronic proliferative synovitis and cartilage destruction. It is the consequence of repeated extravasation of blood into joint cavities, but its exact pathogenesis, particularly with regard to early changes in the joint, is still incompletely understood. This review presents recent findings obtained in experiments performed in vitro and using animal models, which have improved our knowledge of the pathogenesis of haemophilic arthropathy. These experimental studies show that haemophilic arthropathy is a multifactorial event in which the deposit of iron in the joints appears to exert a central role. First, iron may promote the apoptosis of chondrocytes by catalysing the formation of oxygen metabolites; this may explain the fact that intra-articular blood exerts a directly harmful effect on cartilage before, and independent of synovial changes. Secondly, iron may also act on the synovial membrane by favouring its proliferation through the induction of proto-oncogenes involved in cellular proliferation and stimulation of inflammatory cytokines as well as abrogation of apoptosis. These two processes, one degenerative and cartilage-mediated, the other inflammatory and synovium-mediated could occur in parallel or sequentially. Overall, it may be expected that these experimental results will yield new therapeutic strategies capable of effectively preventing the occurrence of this still serious and common complication in patients with severe haemophilia.     

Synovial microenvironment-T cell interactions. Human T cells bind to fibroblast-like synovial cells in vitro

Synovitis in rheumatoid arthritis is characterized by infiltration of the synovium by T and B lymphocytes and monocytes, as well as by the proliferation of synovial lining cells, fibroblasts, and endothelial cells. To study synovial cell-T cell interactions in vitro, we established cultures of fibroblast-like synovial cells, and used these cells in a synovial cell-T cell binding assay. Using T cells at various stages of differentiation and activation, we found that human thymocytes and mitogen-activated peripheral blood T cells bound to fibroblast-like synovial cells, whereas fresh peripheral blood T cells did not. Moreover, activated T cells from inflammatory synovial tissue or from synovial fluid also bound to fibroblast-like synovial cells cultured in vivo. Antibodies against certain epitopes of the T cell CD2 (35.1) and synovial cell lymphocyte function-associated antigen-3 (LFA-3) (TS2/9) molecules inhibited synovial cell-thymocyte binding. However, these same anti-CD2 and anti-LFA-3 antibodies only partially inhibited synovial cell binding to activated normal peripheral blood T cells. Moreover, T cells from inflammatory synovium from rheumatoid arthritis and psoriatic arthritis patients also bound to synovial cells in vitro. These findings demonstrate that fibroblast-like synovial cells are capable of binding to human T cells in vitro, and suggest that during the course of inflammatory synovitis, synovial fibroblast-T cell interactions may occur in vivo.     

Immunohistochemical characterization of synovial cells in arthritic MRL-lpr/lpr mice

MRL-lpr/lpr (MRL/l) mice spontaneously develop an autoimmune disease associated with arthritic manifestations. We used a recently developed mild demineralization procedure, followed by immunohistochemical staining of frozen sections, to investigate cell patterns in the hindlimbs of MRL/l mice at various stages of arthritic disease. Large numbers of Mac-1 (Mas 034)-positive, macrophage-like cells were seen both within the thickened synovial lining layer and in the deeper layers of the synovial tissue in all stages of arthritis. Ia-expressing cells were scarce in the lining layer, but occurred in moderate numbers in the deeper layers of synovial tissue. Lymphocytes were totally absent in MRL/l joints in all stages of arthritis, as demonstrated by lack of staining with Ly-1, Lyt-2, GK 1,5, and antiimmunoglobulin antibodies. Our findings are discussed and related to other types of experimental arthritis and to rheumatoid arthritis in humans.     

Synovial fluid pyrophosphate and nucleoside triphosphate pyrophosphatase: comparison between normal and diseased and between inflamed and non-inflamed joints.

Deposition of intra-articular calcium pyrophosphate is associated with both aging and arthropathy; increased concentrations of free pyrophosphate (PPi) may contribute to such deposition. Free pyrophosphate and nucleoside triphosphate pyrophosphatase (NTPase) were estimated in synovial fluids from 50 subjects with normal knees and from 44 patients with rheumatoid arthritis, 61 with pyrophosphate arthropathy, and 59 with osteoarthritis. For arthropathic knees clinically assessed inflammation was classified as active or inactive using a summated score of six clinical features. The order of PPi (mumol/l) and NTPase (mumol PPi/30 min/mg protein) was pyrophosphate arthropathy greater than osteoarthritis greater than rheumatoid arthritis (median PPi, NTPase respectively: for pyrophosphate arthropathy 15.9, 0.45; for osteoarthritis 9.3, 0.25; for rheumatoid arthritis 4.4, 0.18), with significant differences between all groups. In pyrophosphate arthropathy both PPi (mumol/l) and NTPase (mumol PPi/30 min/mg protein) were higher than normal (15.9, 0.45 v 8.6, 0.2 respectively), but findings in osteoarthritis did not differ from normal. The inflammatory state of the knee had a distinct but variable effect on synovial fluid findings in rheumatoid arthritis and pyrophosphate arthropathy, but not in osteoarthritis. There was no correlation of either PPi or NTPase with age, or between PPi and NTPase in any group. This study provides in vivo data for synovial fluid PPi and NTPase. It suggests that factors other than PPi need to be considered in a study of crystal associated arthropathy. Clinical inflammation, as well as diagnosis, is important in synovial fluid studies.     

Pathogenesis of spontaneous chronic polyarthritis in MRL mice

In a sequential study the pathogenesis of chronic polyarthritis occurring spontaneously in MRL mice was analysed by light microscopy. A total of 128 MRL mice of both substrains (MRL/Mp-lpr/lpr and +/+) and different age groups was studied. In 50 lpr/lpr mice, tinctorial and histochemical methods were applied for the identification of fibrin/fibrinoid, iron compounds, amyloid, and proteoglycan. The earliest lesions seen in mice of substrain lpr/lpr at the age of 2 months were proliferation of synovial lining cells and loss of tinctorially demonstrable proteoglycan in the articular cartilage. Beginning at 3 months, severe joint destruction associated with pannus formation was encountered usually in knee, carpal and tarsal joints. Besides inflammatory processes in tendons, nerves and musculature fibrinoid-necrotising panarteritis occurred in the intra- and extraarticular tissue. Furthermore, fibrin-containing exudations and deposits of fibrinoid material, occurred in the synovium of large joints and in the periarticular connective tissue of phalangeal joints. The occurrence of these morphological changes, destructive arthritis, vasculitis and periarticular inflammatory changes, was, at the age of 3 months, associated with a highly significant increase of circulating immune complexes. In mice of substrain +/+ aged 1 year and older, arthritic changes with synovial lining cell proliferation, cartilage destruction and inflammatory periarticular lesions developed.     

Synovial microenvironment-t cell interactions

Synovitis in rheumatoid arthritis is characterized by infiltration of the synovium by T and B lymphocytes and monocytes, as well as by the proliferation of synovial lining cells, fibroblasts, and endothelial cells. To study synovial cell-T cell interactions in vitro, we established cultures of fibroblast-like synovial cells, and used these cells in a synovial cell-T cell binding assay. Using T cells at various stages of differentiation and activation, we found that human thymocytes and mitogen-activated peripheral blood T cells bound to fibroblast-like synovial cells, whereas fresh peripheral blood T cells did not. Moreover, activated T cells from inflammatory synovial tissue or from synovial fluid also bound to fibroblast-like synovial cells cultured in vivo. Antibodies against certain epitopes of the T cell CD2 (35.1) and synovial cell lymphocyte function-associated antigen-3 (LFA-3) (TS2/9) molecules inhibited synovial cell-thymocyte binding. However, these same anti-CD2 and anti-LFA-3 antibodies only partially inhibited synovial cell binding to activated normal peripheral blood T cells. Moreover, T cells from inflammatory synovium from rheumatoid arthritis and psoriatic arthritis patients also bound to synovial cells in vitro. These findings demonstrate that fibroblast-like synovial cells are capable of binding to human T cells in vitro, and suggest that during the course of inflammatory synovitis, synovial fibroblast-T cell interactions may occur in vivo.     

A new model for rheumatoid arthritis generated by engraftment of rheumatoid synovial tissue and normal human cartilage into scid mice

Objective. A new animal model was used to study the interaction between rheumatoid synovial cells and cartilage and to explore the cellular basis of rheumatoid joint destruction. Methods. Fresh synovial tissue derived from patients with rheumatoid arthritis was implanted with normal human cartilage into SCID mice, either subcutaneously or under the renal capsule, for up to 304 days. The implants were analyzed by light and electron microscopy, as well as by immunohistochemistry and in situ hybridization. Results. Human synovial tissue and cartilage implanted in SCID mice are maintained by the animals for up to 304 days. After 35 days, focal erosions occur at the site of attachment of synovial lining cells to the cartilage. After 105 days, a pannus-like formation, consisting of proliferating synovial fibroblast-like cells invading the cartilage, is observed. The fibroblast nature of these cells was supported by observation of only focal expression of the macrophage markers CD14 and CD68. Cells at the immediate site of cartilage destruction express messenger RNA for cathepsin L, whereas cathepsin D messenger RNA was detected in subsynovial regions away from the site of destruction. The human origin of the tissue involved in cartilage destruction was demonstrated using monoclonal antibodies to HLA-ABC and human type IV collagen. Conclusion. The present approach introduces a novel in vivo model of rheumatoid arthritis for the study of the molecular and cellular mechanisms of rheumatoid joint destruction at sites of synovial attachment to cartilage. In this model, the SCID mouse acts as a useful host for studying the properties of rheumatoid synovium in the absence of circulating human blood components.     

Production of angiotensin converting enzyme by rheumatoid synovial membrane.

Vascular proliferation and mononuclear cell infiltration are prominent changes observed in synovium from actively inflamed joints of patients with rheumatoid arthritis. Angiotensin converting enzyme (ACE) is a halide activated peptidase produced mainly by endothelial cells and by activated monocytes. It has been proposed that levels of ACE activity in synovial fluid might reflect changes in membrane vascularity, the degree of monocyte infiltration, or the thickness of the lining layer. In this study, ACE activity in serum and synovial fluid samples from 18 patients with inflammatory arthritis was measured and compared with levels in 12 control subjects with non-inflammatory arthritis. Although serum levels were similar in the two groups, ACE activity in synovial fluid was significantly increased in the group with inflammatory arthritis compared with controls (mean (SE) 37 (5) v 19 (3)). Staining of synovial membranes from patients with rheumatoid arthritis with a monoclonal antibody to ACE localised ACE to the endothelium and to mononuclear cells of macrophage origin. ACE activity was then measured in supernatants of synovial membrane from patients with rheumatoid arthritis after one and seven days of culture. A significant increase in ACE activity was observed after seven days of culture (mean (SE) day 1, 17 (5) v day 7, 25 (3)). Levels of ACE activity, however, did not correlate with the lining layer thickness, with the number of macrophages per square millimetre, nor with the number of blood vessels per square millimetre of synovial tissue. No correlation was observed either between levels of ACE in the supernatant of synovial membrane and levels of interleukin 1 or interleukin 6. In conclusion, ACE is produced by the synovial membrane of patients with rheumatoid arthritis and is localised to monocytes and endothelial cells. Levels of activity do not directly reflect membrane vascularity, monocyte or macrophage number, or the thickness of the lining layer.     

Retroviruses and chronic arthritis. Possible significance of some recent observations

Retroviruses have been proposed as etiologic agents for the development of chronic arthritis in humans. The arthritis seen in goats infected by caprine arthritis encephalitis virus and the spontaneous arthritis of inbred MRL/l mice illustrate how retroviruses may cause the development of a disorder closely resembling human rheumatoid arthritis. Several investigators have searched for evidence of retrovirus infection in patients with chronic arthritis, but in most cases the results have been disappointing. However, in 1983, Iversen isolated a virus-like particle from a patient with psoriasis. The particle had a buoyant density in sucrose and a protein composition that closely resembled murine and primate retroviruses. Particle proteins participate in immune complex formation in psoriasis, in psoriatic arthritis, and in ankylosing spondylitis. Particle proteins are also present in deposits in psoriatic lesions and in affected synovial tissue resembling immune complex deposits. The possible role for retrovirus-like antigens in the inflammatory process in psoriasis and seronegative arthritis is discussed.     

Laminin and vascular proliferation in rheumatoid arthritis.

Laminin is a high molecular weight basement membrane structural glycoprotein. In rheumatoid arthritis and other arthropathies immunoreactive laminin was prominent in synovial blood vessel basement membranes and acted as a marker for them. It codistributed with collagen type IV. Immunohistological reactivity to laminin showed extensive vascular proliferation in rheumatoid arthritis together with basement membrane reduplication, which was confirmed ultrastructurally. Parallel histological studies showed vascular proliferation was predominantly in the subintimal rheumatoid synovium, where it was related to connective tissue proliferation but not to the inflammatory cell infiltrate. Vascular proliferation was also seen in relation to connective tissue changes in biopsies from cases of haemophilic arthritis, osteoarthritis, and meniscal tears. We suggest connective tissue activation is non-specific reaction associated with vascular proliferation. This involves laminin and other structural proteins. It occurs in rheumatoid arthritis and other arthropathies but is distinct from inflammatory cell infiltration.     

Co-existent sickle cell disease and juvenile rheumatoid arthritis. Two cases with delayed diagnosis and severe destructive arthropathy

We describe 2 pediatric patients with sickle cell disease (SCD) who developed seropositive juvenile rheumatoid arthritis (JRA). Both patients have severe joint damage, the compound effect of both disease processes. The bone and cartilage destruction, which poses serious therapeutic challenges, highlights the difficulty of making a diagnosis of chronic inflammatory disease in the setting of SCD. There may be a correlation between increased levels of tumor necrosis factor-alpha in the synovial tissue of joints damaged by arthritis and local sickling. The resultant ischemia and corresponding inflammatory infiltrates could in turn worsen existing synovial proliferation and cartilage destruction as well as trigger further sickling.     

Cartilage proteoglycans in synovial fluid and serum in patients with inflammatory joint disease. Relation to systemic treatment

Proteoglycan concentrations in knee joint synovial fluid and in serum from patients with various inflammatory arthritides were studied using an enzyme-linked immunosorbent assay. Patients with reactive arthritis, calcium pyrophosphate arthropathy, and juvenile rheumatoid arthritis (age less than or equal to 20 years) had the highest synovial fluid concentrations. These values differed significantly (P less than 0.001) from those in patients with rheumatoid arthritis, psoriatic arthropathy, and chronic HLA-B27-associated arthropathy. Rheumatoid arthritis patients receiving low-dose prednisolone treatment had higher synovial fluid (P = 0.006) and serum (P less than 0.001) proteoglycan concentrations than did those taking nonsteroidal antiinflammatory drugs or slow-acting antirheumatic drugs. Serum proteoglycan concentrations were near the detection limits, and did not correlate with levels found in paired samples of knee joint synovial fluid. Patients with calcium pyrophosphate arthropathy had the highest mean serum level of proteoglycan. This assay of proteoglycan antigens is a useful tool in the study of proteoglycan metabolism in patients with joint disease. With its use, differences between disease groups and effects of therapy can be distinguished.     

Proliferation of synovial lining cells and fibroblasts.

After in vitro incubation with 3H-thymidine, the proliferation of synovial lining cells and fibroblasts was investigated in surgically-removed articular tissue. Under normal conditions, in nonrheumatoid arthritis and in osteoarthrosis, low proliferation rates for both cell types were observed. In rheumatoid arthritis, the rate of proliferation of both cell types was usually increased. An increased proliferation of the synovial lining cells was especially observed in cases with a hyperplastic lining cell layer and in cases with a minimal lymphocytic and plasma cellular infiltration in the synovial membrane.     

Self-associating IgG rheumatoid factors in MRL/l autoimmune mice

Previous work has shown that the intermediate complexes isolated from the plasma of patients with rheumatoid arthritis are composed of self-associating IgG rheumatoid factors. Mice of the MRL/l strain develop spontaneous autoimmune disease with arthritis that is pathologically similar to human rheumatoid arthritis. Also, the sera of MRL/l mice contain autoantibodies to nuclear antigens as well as IgM and IgG rheumatoid factors. The present studies were done to determine if the IgG rheumatoid factors isolated from these mice undergo self-association. MRL/l mouse sera were categorized into groups A and B based on serum-serum precipitin interactions. Thirteen of 13 MRL/l mice sera examined contained intermediate complexes sedimenting between the 6.6S and 19S components of normal serum by sedimentation velocity ultracentrifugation. There were no differences in the level of intermediate complexes between groups A and B. IgG rheumatoid factors were isolated from the sera of 9 other mice. Upon sedimentation equilibrium ultracentrifugation, these rheumatoid factors underwent concentration-dependent self-association similar to that described for human self-associating IgG rheumatoid factors, although the precise stoichiometry of self-association could not be determined. The IgG rheumatoid factors from group B had higher energies of self-interaction than those from group A. These studies provide additional evidence that MRL/l mice may be the best available animal model for the study of human rheumatoid arthritis.     

Autoimmune reaction to type II collagen and cartilage degeneration in MRL/Mp-<Emphasis Type="Italic">lpr/lpr</Emphasis> mouse

The early phase of cartilage degeneration was immunohistochemically examined in order to clarify the importance of autoimmune reaction against type II collagen in MRL/Mp-lpr/lpr (MRL/l) mouse in an experimental model of rheumatoid arthritis (RA). Anti-type II collagen antibodies were detectable in 3-week-old mice and preceded the appearance of rheumatoid factors. Furthermore, mesenchymal cells and tartrate-resistant acid phosphatase-positive cells began to accumulate remarkably in the periphysis, a fibrochondro-osseous area in the bone marrow vicinity. The numbers of these cells increased with mice age, together with serum levels of anti-type II collagen antibodies. Immunostaining of the periphysis revealed expression of type II collagen, IgG, C3, Mac-3, MHC class II antigen Ia, and cathepsin-L. Osteoclast-like cells and macrophage infiltration into the lesion area were confirmed by transmission electron microscopy. The results indicate that cartilage degeneration in MRL/l mouse may originate in the periphysis and progress via a pathway independent of synovial invasion.     

Osteoclast-like cells in an in vitro model of bone destruction by rheumatoid synovium.

OBJECTIVE: Osteoclasts may be involved in the process of rheumatoid bone destruction. To test this hypothesis, we developed an in vitro model of bone destruction by osteoclast-like cells derived from cultured rheumatoid synovial tissue without using any inducers. METHODS: Synovial tissues were obtained from rheumatoid arthritis and osteoarthritis patients and tissue pieces of about 2 mm(3) that contained synovial lining were cultured. Multinucleated cells derived from cultured synovial tissues were studied cytochemically and morphologically for osteoclast-specific markers. RESULTS: Fibroblast-like and macrophage-like cells from the tissue pieces proliferated in the coexistence of lymphocytes. After 14 days of culture, multinucleated cells with tartrate-resistant acid phosphatase activity appeared. These cells expressed vacuolar H(+)-ATPase, the vitronectin receptor and cathepsin K. Although binding of (125)I-labelled salmon calcitonin was very low, the cells contained ringed structures of F-actin and showed strong bone-resorbing activity on ivory slices. Proliferation of macrophage-like cells and formation of multinucleated cells continued during 6 months of culture in the presence of fibroblast-like cells. The bone-resorbing activity of multinucleated cells derived from rheumatoid synovial tissue was much higher than that of cells from osteoarthritis synovial tissue, and was related to the disease activity of rheumatoid arthritis. CONCLUSION: Our culture system reproduced in vitro the process of bone destruction by rheumatoid synovium, including the proliferation and fusion of precursor cells, polarization, activation and bone tissue resorption. This system may provide a tool for understanding the mechanisms of bone destruction in rheumatoid arthritis and for the development of new therapies to prevent bone destruction.