Genomic Mutations in Rheumatoid Arthritis

Abstract

Background Rheumatoid arthritis is a complex and genetically heterogeneous autoimmune disease characterised by chronic systemic inflammation and joint destruction. Once established, RA transitions from episodes of remission to relapse or flare, as well as to a phase of complete drug resistance, known as refractory RA. The pathophysiological pathways behind these transitioning stages are largely unknown. In this PhD project, the landscape of somatic and rare pathogenic germline variants was explored. Based on the evidence, that severe immune dysregulation in inflammatory disorders occur due to clonal somatic mutations in haematopoietic stem cells, the somatic variant search was focused on clonal haematopoiesis (CH)-associated somatic mutations. Arthritis is also considered as a presenting symptom of primary immunodeficiency (PID) therefore, rare pathogenic germline variants were searched in genes associated with primary immunodeficiency (PID) as well as immune function genes that are associated with the risk of RA by genome-wide association studies. Methods Somatic variants: Genomic DNA from 136 RA samples (remission=49, flare=47, refractory =40) was used for whole exome sequencing (WES) at 90-fold coverage. WES data was pre processed using standard GATK4 pipeline, followed by somatic variant calling using Mutect2 pipeline in tumour-only mode against a panel of normal. A customised list of 210 CH gene was to identify CH-associated somatic mutations. Germline variants: WES data from 43 refractory arthritis patients and 98 non-refractory arthritis controls was pre-processed using standard GATK4 pipeline, followed by germline variant calling using Haplotype caller. Rare pathogenic variants in 545 PID genes and 330 RA GWAS genes were identified with the following criteria: MAF <0.0005 or 0.05% in Exome Aggregation Consortium (ExAC) database and Combined Annotation Dependent Depletion (CADD) score of >20. Results CH-associated somatic mutations: Overall, CH was prevalent in 33% of the RA cohort, with individual percentage prevalence of 40%, 28% and 30% in flare, remission and refractory RA cohorts, respectively. The incidence of CH increased exponentially with age, affecting 41.8% of the patients between these ages of 60-69 years. The genes predominantly mutated were DNMT3A, ASXL1 and TET2, however, PABPC1 was identified as a potential novel driver of CH. One patient with SFRB1 clone at 25% VAF subsequently developed myelodysplastic syndrome (MDS). Rare, pathogenic germline variants: In the rare variant analysis, MLH1 was found significantly associated with refractory RA (Fisher’s exact test, p=0.02), affecting more patients in cases than controls. Another gene that was found highly mutated in refractory arthritis patients was NAV2 as compared to controls. At least three genes that were exclusively mutated in refractory arthritis had variants of potential clinical relevance, which included POLR3E, NAF1, and NOX1. Conclusions This is the first large-cohort study reporting both CH-associated somatic mutations as well rare pathogenic germline variants in RA. CH was detected at 33% prevalence in RA, which is higher than previous reports. Furthermore, in rare-variant association analysis, MLH1 was found significantly associated with refractory RA.