Evidence for an infectious etiology of rheumatoid arthritis

Max Schüller and an alleged microbial cause of rheumatoid arthritis

Max Schüller (1843-1907) was a German surgeon and microbiologist. From 1884 to 1905, he published histopathologic and bacteriological research on bacterial infection of human joints. Beginning in 1892, he focused on a bacterium he had identified in joints of patients with rheumatoid arthritis and surmised that he had discovered the cause of this disease. He persisted in conducting various experiments, some original at the time, with which he convinced himself of the validity of his discovery, without considering the possibility that he was working on contaminants. Contemporaneous attempts to confirm Schüller's findings gave inconsistent results. A century of microbiological research with ever more sensitive techniques has not definitively answered the etiologic question. This history is a cautionary tale of the difficulty of disproving an erroneous premise.

Rheumatoid arthritis: clinical features and pathogenetic mechanisms

Rheumatoid arthritis is a disorder characterized by chronic inflammation affecting predominantly articular tissues, leading in some instances to disruption of the normal structure and function of the joint. In some patients extra-articular manifestations are also present. In the joints, the pathologic lesion consists of synovial cell proliferation with infiltration by inflammatory cells. The complex interaction among these cells is likely responsible for the connective tissue abnormalities that characterize the rheumatoid lesion. The factors responsible for inducing the inflammatory process are unknown.

Search for viruses in rheumatoid macrophage-rich synovial cell populations

The adherent cells remaining after short-term culture of synovial fluid and synovial membrane cells from rheumatoid and non-rheumatoid patients were examined for the presence of a productive virus infection and for various viral antigens. Labelling was carried out with 3H-thymidine and 3H-uridine followed by sucrose density gradient centrifugation of the culture supernatant. Only in 1 case was there incorporation of 3H-uridine into material of density 1 . 21 g/cm3. Viral antigens were tested for by indirect immunofluorescence with antisera to rubella virus, the retroviruses RD-114 and simian sarcoma associated virus, early adenovirus type 2 antigens, late adenovirus type 2 antigens, SV-40 T antigen, and in 1 case measles virus. No cell showed immunofluorescence with any antiserum except the early adenovirus type 2 antiserum, which stained the cytoplasm of about half the synovial cell cultures, some from rheumatoid and some from non-rheumatoid patients.

Studies of serum immunoglobulin binding to synovial fibroblast cell cultures from patients with rheumatoid arthritis

Using a sensitive 125I-protein A (PrA) binding assay to detect cell surface IgG, we have studied seven different synovial fibroblast cell cultures from patients with rheumatoid arthritis (RA). When these cultures were incubated in the presence of serum from 18 autologous and allogeneic RA patients (all seropositive), we were unable to detect significant IgG binding. Since IgM rheumatoid factor (RF) can block PrA binding, sera were absorbed with aggregated IgG to remove RF without affecting the results. Similar studies on three cell lines with seven rheumatoid sera were performed by antibody-dependent cell-mediated cytotoxicity. No significant cytotoxicity was observed. Since antibodies to collagen are present in rheumatoid sera, several cultures were incubated with ascorbic acid (12.5 microgram/ml) to optimize synthesis of cell surface collagen. These culture conditions did not affect serum immunoglobulin binding by the 125I-PrA assay. Thus, we can find no evidence for a direct humoral immune mediation of synovial proliferation in rheumatoid arthritis. These data do not support the hypothesis that the inflammatory process within the synovium of RA patients is an immunologic response to a fibroblast-associated antigen in the synovial membrane.

Search for viral nucleic sequences in rheumatoid cells

DNA and RNA were extracted from synovial membranes, synovial fibroblast cells, peripheral blood lymphocytes, and synovial fibroblast cells strains derived from patients with rheumatoid arthritis and other joint conditions. They were hybridised after immobilisation on nitrocellulose filters with iodinated viral nucleic acids extracted from measles, rubella virus, SV--40, and a retrovirus, RD--114. In addition, in situ-hybridisation was carried out on sections of synovial membranes by means of iodinated measles and rubella virus RNA. In no case did any hybridisation occur. Positive control systems included synovial fibroblast strains transformed with SV--40, LLC--MK2 cells chronically infected with rubella virus and RD cells infected with RD--114. It was concluded tht the synovial cells did not contain viral genomes of measles, rubella virus, SV--40, or RD--114, or at least at a level equivalent to the positive control cells.

Identification of Ureaplasma urealyticum (T-strain Mycoplasma) in patient with polyarthritis

A 10-year-old boy with confirmed congenital agammaglobulinaemia presented with polyarthritis while on gammaglobulin replacement therapy. Initial cultures of material aspirated from an abscess and of joint fluid were negative, and symptoms progressed despite antibiotic therapy. Synovial-biopsy material, cultured specifically for mycoplasmas, was positive for Ureaplasma urealyticum as were the blood, abscess fluids, throat-swab, and nasopharyngeal secretions. Therapy, based on in-vitro studies of antibiotic susceptibilities of the organism, resulted in the eradication of the infection and resolution of the arthritis. These findings suggest that U. urealyticum may be capable of inducing polyarthritis in man.

Susceptibility of rheumatoid and nonrheumatoid synovial cells to antibody-dependent cell-mediated cytotoxicity

Fibroblastic cells derived from rheumatoid (RA) and nonrheumatoid (N-RA) synovial tissue (synovial cells) were used as targets in assays of antibody-dependent cell-mediated cytotoxicity (ADCC). Synovial cells that had been pretreated with human alloantisera were rapidly lysed by normal human blood mononuclear cells. RA and N-RA synovial cells were equally susceptible to ADCC under these assay conditions. Antibody to autologous synovial cells was not detected in sera from 10 patients with RA by this method of assay.

Ultrastructural study of avian synovium infected with an arthrotropic reovirus

Three-week-old chicks were inoculated via the foot pad with the avian reovirus (strain WVU 2937) that has been proved to be the causative agent in avian viral arthritis. The initial stages of the avian reovirus infection in the synovium of the hock joint were studied by transmission electron microscopy to shed light on the early virus-host relationship of an arthrotropic virus and synovial tissue. At 48 hours postinoculation (PI), coated, partially coated, and uncoated virus particles were seen in a dense lysosome-like inclusion in the cytoplasm of subsynovial fibroblasts. Virus replication was demonstrated in fibroblasts of the subsynovium 4 days PI. Infiltration of leukocytes into the area of the virus replication was noted at the time of virus replication (4 days PI). By the second week of the infection (11 days PI), the subsynovium was heavily infiltrated with lymphocytes and plasma cells, and proliferative changes had occurred in the synovial lining cells.