• autoantibodies
• autoimmunity
• rheumatoid arthritis
• infections

A study by Pianta et al published in the Journal of Clinical Investigation provides new evidence that the pathogenesis of rheumatoid arthritis (RA) may involve molecular mimicry, one of the most venerable models for the aetiology of rheumatic disease.1 Molecular mimicry represents the development of cross-reactive B or T cell responses to a component of an infecting agent. As shown in this study, the agent in question may be a commensal in the gut microbiome rather than a bacterium or virus inducing clinical disease.

While a violation of tolerance, molecular mimicry can occur due to the sharing of amino sequences by proteins from pathogen and host. If an immune response during infection goes awry, autoreactivity to the shared sequence can ensue. The most dramatic example of molecular mimicry is acute rheumatic fever (ARF), still a major problem worldwide. The scenario is now classic: within weeks of an infection with group A Streptococcus, usually pharyngitis, an inflammatory syndrome encompassing arthritis strikes a susceptible person. In ARF, the infection is obvious; the culprit bacterial antigen is the M protein.2

In a novel approach to finding molecular mimics, Pianta et al used mass spectrometry to characterise peptides bound to HLA-DR molecules of cells from synovial tissue, synovial fluid or peripheral blood of patients with RA.1 3 This analysis demonstrated the presence of DR-presented peptides derived from previously unidentified autoantigens called N-acetylglucosamine-6-sulfatase (GNS) and filamin A (FLNA). Importantly, peptides from GNS and FLNA have sequence homology to proteins of Prevotella and other gut species. The homology is notable since the microbiomes of patients with new-onset RA have expansion of Prevotella copri.4

Other studies in this paper showed that sera from patients with RA contain IgG and IgA antibodies to GNS and FLNA, with levels correlated with those of P. copri. Among patients lacking anti-citrullinated protein antibodies (ACPA) or rheumatoid factor, over half had IgG and/or IgA antibodies to GNS or FLNA. Furthermore, GNS and FLNA peptides could stimulate in vitro responses of T cells. Overall, B and T cell responses to GNS and FLNA were present in 52% and 56% of patients, respectively, providing impressive evidence for a role of gut antigens in driving events in RA.

While RA, like ARF, may involve molecular mimicry, the clinical situations of RA and ARF are vastly different. With ARF, the infection is readily apparent and the onset of disease is sudden, making it easy to postulate a link. In the case of RA, the ‘infection’ is dysbiosis, a shift in the composition in the microbiome, in this case, increased Prevotella sp. The onset of RA is gradual and years may pass before an autoantibody response to proteins, especially citrullinated versions, develops and culminates in arthralgia and arthritis. Since the microbiome is populated early in life, the basis of this long evolution is unclear.

Molecular mimicry is only one mechanism by which dysbiosis can impact the pathogenesis of autoimmunity. Studies, especially in animal models, have suggested other intriguing possibilities. Thus, genetic factors may affect the composition of the microbiome; dysbiosis may perturb the poise of the immune system; and inflammation may lead to dysbiosis. The interplay between host and organism is thus likely to be complicated and dynamic.5 6

While the gut microbiome attracts great attention, it is only one of several. The mouth and the upper airways are others. A role of the gut microbiome in RA pathogenesis must therefore be incorporated into models of RA based on a role of periodontal and pulmonary infection in inducing protein citrullination and subsequent ACPA production.7 If, as some data suggest, the gut and the oral cavity may have common organisms in their microbiomes, there can be two-way traffic of organisms in the body as well as traffic of pathogenic T cells emerging in either locale to journey into the joint.8

Practically, these studies are important in identifying new target antigens for serological assessment in RA, especially for those patients who are ACPA negative. In addition, the phenomenon of dysbiosis raises the possibility of therapy to improve this state although the changes with disease-modifying antirheumatic drugs suggest a return to more normal gut composition can occur even with current approaches.8 Figure 1 highlights these mechanisms. Whatever the direction of future therapy, this important study suggests that the boundary between foreign and self is not complete and that various encounters with bacterial organisms can lead to mimicry of the pathological kind.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The role of the microbiome in the pathogenesis of rheumatoid arthritis (RA). In the pathogenesis of RA and other autoimmune and inflammatory conditions, dysbiosis can result from the interplay of genetic and environmental factors including diet. Dysbiosis can lead to changes in the mucosal barrier; leakage of bacteria and bacterial products into the blood; immune activation; induction of cytokines; and metabolic disturbance. In RA, dysbiosis can encompass an increase in Prevotella copri (PC) as well as a decrease in Bacteroides sp. Depending on genetic factors, most prominently the HLA-DR shared epitope, exposure to bacterial antigens can induce the activation of T cells that react to proteins from Prevotella as well as cross-reactive proteins N-acetylglucosamine-6-sulfatase (GNS) and filamin A (FLNA). These T cells can provide help to B cells which produce IgG and IgA antibodies to these antigens by molecular mimicry. Both T cell and B cell autoreactivity to GNS and FLNA in the joint can drive synovitis. Disease-modifying antirheumatic drugs (DMARD) can ameliorate inflammation as well as impact on dysbiosis.