AB0089 MENDELIAN RANDOMIZATION REVEALS CAUSATIVE GUT BACTERIA FOR ANKYLOSING SPONDYLITIS

S. Chen,X. Wu,Z. Chen,M. Xiao,J. Wu,J. Gu

DOI: https://doi.org/10.1136/annrheumdis-2023-eular.275

IF: 27.973

2023-01-01

Annals of the Rheumatic Diseases

Abstract:Background Genetic, clinical and experimental evidence has mounted to suggest the implication of gut dysbiosis in the pathogenesis of ankylosing spondylitis (AS) through the gut-joint axis of inflammation[1]. A series of cross-sectional studies have unraveled the alterations of gut microbiome composition in patients with AS using metagenome sequencing. However, such observational associations might be limited to the small sample sizes and the influences from confounders. Furthermore, whether gut dysbiosis is a cause or a consequence of AS remains elusive. Mendelian randomization (MR) is an epidemiologic approach for causal inference using genetic variants as instrument variables. Objectives To assess the causal relationship between gut bacteria taxa and risk of AS using two-sample MR. Methods To select instrument variables for gut bacteria taxa, we filtered genome-wide significant (p ≤ 5 x 10-8) and independent (r 2 < 0.01 within the window size of 500kb) variants for 211 taxa from a large-scale genome-wide association study (GWAS) on gut microbiota which included 18,340 individuals from 24 cohorts[2]. The summary statistics on AS from the UK Biobank (UKB) and the FinnGen consortium were utilized for the discovery and replication analyses, respectively[3]. Under the framework of two-sample MR, the inverse variance weighted method or the Wald ratio method was used to assess the causal effect of gut microbial taxa on risk of AS. In addition, multiple sensitivity analyses and reverse MR analyses were conducted to validate the robustness of results and evaluate reverse causation, respectively. Results Despite no taxa passing the multiple-testing correction, six known taxa were identified to have causal effects on the onset of AS in the UKB cohort at the nominal significance level (p < 0.05). In the FinnGen cohort, only one bacterial genus, Anaerotruncus was replicated (UKB: βIVW = -0.0016, p = 0.04; FinnGen: βIVW = -0.3990, p = 0.04). Sensitivity analyses led no support to the presence of heterogeneity (Q_p > 0.05) and pleiotropy (Intercept_p > 0.05). Reverse MR analysis found no reverse causality from AS to Anaerotruncus. Conclusion Despite no current evidence supporting the association between Anaerotruncus and AS, this study provides the first genetic evidence for the protective effect of this short-chain fatty acid (SCFA)-producer against AS. Intriguingly, several studies have suggested a relationship between Anaerotruncus and immune-mediated diseases including rheumatoid arthritis and inflammatory bowel diseases[4]. These MR findings shed light on the causative microbiota and microbiome-modulating strategies for AS. References [1]Qaiyum Z, Lim M, Inman RD. The gut-joint axis in spondyloarthritis: immunological, microbial, and clinical insights. Semin Immunopathol 2021;43:173–92. doi:10.1007/s00281-021-00845-0 [2]Kurilshikov A, Medina-Gomez C, Bacigalupe R, et al. Large-scale association analyses identify host factors influencing human gut microbiome composition. Nat Genet 2021;53:156–65. doi:10.1038/s41588-020-00763-1 [3]S S, M K, Y T, et al. A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet 2021;53. doi:10.1038/s41588-021-00931-x [4]Thomann AK, Wüstenberg T, Wirbel J, et al. Depression and fatigue in active IBD from a microbiome perspective-a Bayesian approach to faecal metagenomics. BMC Med 2022;20:366. doi:10.1186/s12916-022-02550-7 Acknowledgements We thank the researchers in the MiBioGen consortium, UK Biobank and FinnGen biobanks for releasing their valuable GWAS datasets. Disclosure of Interests None Declared.