Short-chain fatty acids and their gut microbial pathways distinguish rheumatoid arthritis in discordant monozygotic twins

Although genetic risk factors, such as HLA-DRB1 alleles, contribute to the pathogenesis of rheumatoid arthritis (RA), the concordance rate in monozygotic (MZ) twins is low, suggesting that other factors are involved in disease development. Further, the relative contribution of nongenetic elements in identical twins has not been characterised. Here, we aimed to characterise host and microbial biomarkers of RA by studying MZ twins discordant for disease using a multiomics approach. Eight pairs of MZ twins discordant for RA (N = 16) were enrolled in the United States (US). The gut microbiome was assessed using shotgun metagenomic sequencing. Autoantibodies, cytokines, and plasma proteins were measured in both plasma and faeces. Levels of short-chain fatty acids (SCFAs) from serum and faeces were quantified using gas chromatography mass spectrometry (GC-MS). Metagenomic data from a UK twin registry (TwinsUK) (N = 14) were used to validate findings in the US population. Although microbiome diversity and composition did not differ between twins, we observed a significant decrease in the SCFA-producing bacteria Blautia faecis and significantly lower concentrations of faecal butyrate and propionate in affected RA twins in the US. TwinsUK showed a similar reduction in the SCFA-producers Gemmiger formicilis and Faecalicatena fissicatena, as well as bacterial SCFA metabolism pathways. Multiomics biomarkers differentiate MZ twins discordant for RA. Faecal butyrate and propionate, as well as SCFA-producing bacteria, were decreased in affected twins. We found a similar decrease in SCFA-producing taxa in affected twins in a geographically distinct cohort in the UK. Our results suggest that, if further validated in larger cohorts, multiomics approaches may improve our understanding of RA pathogenesis and, potentially, contribute to more accurate diagnostics and coadjuvant therapies.