Background and Objectives Porphyromonas gingivalis, a major cause of chronic periodontitis (PD), has been implicated also in the aetiology of rheumatoid arthritis (RA), specifically in anti-citrullinated protein/peptide antibody (ACPA) positive RA, based on its unique property among pathogens to express a citrullinating peptidylarginine deiminase enzyme. To investigate this hypothesis further, we have analysed the antibody response to P. gingivalis in relation to known risk factors for RA, including autoantibodies, smoking and HLA-DRB1 shared epitope (SE) alleles.

Materials and Methods Serum from 66 patients with chronic PD and 60 individuals without (confirmed by dentists), as well as serum from the EIRA (Epidemiological Investigation of RA) cohort (n = 1985 RA cases and n = 400 controls), were screened by ELISA for presence of antibodies to the P. gingivalis-specific virulence factor arginine gingipainB (RgpB). Antibody levels were analysed in relation to the presence of anti-CCP antibodies, rheumatoid factor, anti-carbamylated protein antibodies and various ACPA fine-specificities, as well as smoking and SE. Differences in antibody levels were examined using Mann-Whitney U test for independent groups.

Results Significantly higher anti-RgpB antibody levels were detected in the PD subset, compared to the non-PD control subset, and in EIRA RA cases, compared to EIRA controls. Among EIRA RA cases, anti-RgpB IgG levels were significantly increased in RF + as well as anti-CCP + patients. Interestingly, anti-RgpB IgG levels associated with the presence of some ACPA fine-specificities, but not others. There was no association with anti-carbamylated protein antibodies, smoking or HLA-DRB1 SE alleles.

Conclusion Significantly increased anti-RgpB antibody levels in PD serum, compared to non-PD serum, may suggest that elevated anti-RgpB IgG levels can be used as a proxy for chronic PD, in studies where the periodontal status is of interest but unknown. A number of previous studies have demonstrated an association between anti-P. gingivalis antibody responses and ACPA positive RA. However, most of those studies have utilised bacterial lysates as capturing antigen, which are known to contain citrullinated epitopes. By using recombinant RgpB, we circumvent the potential problem of false positive results due to cross-reactive ACPAs, and we conclude that there is an association between elevated anti-RgpB IgG levels and a subset of RA that is characterised by a specific ACPA profile. Interestingly, we could not detect an association with smoking, a well-known risk factor for both PD and ACPA+ RA. These results may point to different aetiologies in different subsets of ACPA positive RA.