Genetics and pathophysiology of systemic juvenile idiopathic arthritis and other complex autoinflammatory diseases

National Institute Of Arthritis And Musculoskeletal And Skin Diseases

Investigators

Linked publications, trials & patents

Abstract

To enable investigations of the genetic susceptibility of sJIA, we established the International Childhood Arthritis Genetics (INCHARGE) Consortium. We assembled a collection of DNA specimens from over 1000 children with sJIA from North America, South America and Europe in whom we generated a dataset of millions of common genetic markers or single nucleotide polymorphisms (SNPs) spanning the entire genome. When combined with genome-wide SNP data from a collection of 8010 geographically matched control subjects, this case-control collection became the basis for the first genome-wide association study and a series of subsequent genomic investigations of sJIA. Using this dataset, we discovered that a class II human leukocyte antigen (HLA) region haplotype that encodes the HLA-DRB1*11 allele was the strongest genetic risk factor for sJIA (odds ratio 2.6). The class II HLA proteins encoded by this locus are essential elements of the adaptive immune system that initiate antigen-specific responses, and this study was the first to implicate adaptive immunity in sJIA. We looked beyond the HLA locus and discovered a second sJIA susceptibility locus in an intergenic region of chromosome 1 (odds ratio 2.4). Moving beyond singular gene associations, we used the sJIA GWAS dataset to perform a formal comparison of the genetic architecture of sJIA with those of the other forms of JIA. This unambiguously revealed that the genetic signature of sJIA is unique among the JIA subtypes. We then examined sJIA candidate susceptibility loci that had been reported prior to the GWAS era, which confirmed an association between sJIA and a group of regulatory SNPs in the promoter of the interleukin-1 receptor antagonist (IL1RN) gene. These SNPs are among the strongest predictors of IL1RN transcript and protein levels in humans, with low IL1RN expression correlating with increased risk of sJIA. This led us to discover that homozygosity for the high expression IL1RN genotypes was highly predictive of non-response to treatment with human recombinant IL1Ra (anakinra) in sJIA (sensitivity 0.92; specificity 0.71). We therefore proposed the use of these SNPs as biomarkers to guide selection of therapeutic agents in sJIA. 1. Immmunochip Association Study of sJIA. During the current reporting period, we have again leveraged the INCHARGE cohort to investigate genetic contributions to sJIA. Together with the Juvenile Arthritis Consortium for Immunochip (JACI), we performed a genetic study of 889 children with sJIA and >16,000 healthy subjects to examine 200 immunologically important genes. This study identified a new sJIA susceptibility locus on chromosome 2, where an sJIA-associated haplotype contains CXCR4, encoding C-X-C chemokine receptor type 4 (p = 4.3 times 10 to the negative 10; odds ratio = 1.7 per allele; 95% confidence interval 1.5 to 1.9). In silico analysis of the haplotype revealed a strong regulatory signature with evidence of participation in protein binding and chromatin looping. Using paired whole genome sequencing data and lymphoblastoid cell (LCL) RNA sequencing data from 1000 Genomes Project subjects, we identified positive correlation between the sJIA risk haplotype and CXCR4 expression in LCLs (p = 0.0004). Emphasizing the importance of this finding, in silico examination of bulk RNA sequencing data from CD14+ monocytes (GSE147608) revealed higher expression of CXCR4 in monocytes from children with sJIA than in healthy children, regardless of sJIA disease activity (p = 0.0009). To study the effect of the CXCR4 risk haplotype in primary cells, we identified individuals with haplotype configurations of interest in the NHGRI Genomic Ascertainment Cohort. CXCR4-dependent chemotaxis of 12 immune cell populations was assayed in freshly isolated peripheral blood leukocytes from 6 pairs of subjects (1 risk homozygote and 1 protective homozygote in each pair) using a spectral cytometry-based transwell system. CD14+ monocytes from subjects with the sJIA risk haplotype migrated more strongly towards the CXCR4 ligand, CXCL12, than did those from subjects without the risk haplotype (p < 0.05). We observed a similar trend in several other cell types, but these observations did not reach statistical significance. Taken together, this work suggests that the sJIA risk haplotype genetically encodes a program that results in increased CXCR4 expression and enhanced CXCR4-mediated chemotaxis. Dr. Hiroto Nakano received the 2021 NIH Fellows Award for Research Excellence for this work, which he will be presenting at the 2022 American College of Rheumatology Convergence. A manuscript describing these findings is in preparation. 2. HLA-DRB1*15 is a biomarker that strongly predicts lung disease in sJIA treated with anakinra, canakinumab or tocilizumab. We have also continued to investigate the etiology of a highly lethal form of lung disease that has emerged over the past 15 years in people with sJIA and the analogous adult condition, adult-onset Stills disease. Together with colleagues from around the globe, we identified a strong association between HLA-DRB1*15 alleles and the development of lung disease in sJIA (sJIA-LD) following treatment with a group of 3 cytokine inhibitors, anakinra, canakinumab and tocilizumab (p = 1.7 times 10 to the negative 14; odds ratio 14.1; 95% confidence interval 6.4 to 31.0). All subjects with sJIA-LD displayed peripheral eosinophilia, atypical skin rash or elevated serum transaminase levels, which in the context of the high effect HLA association was concerning for a drug hypersensitivity reaction. Using a validated classification system for hypersensitivity reactions, we showed that all subjects with sJIA-LD also fulfilled criteria for DRESS (drug reaction with eosinophilia and systemic symptoms). We also identified a group of sJIA cases without lung disease that developed peripheral eosinophilia following exposure to these agents, all of whom also fulfilled criteria for DRESS. When all sJIA-DRESS subjects were combined, we observed an association with HLA-DRB1*15 alleles (p = 7.9 times 10 to the negative 18; odds ratio 11.9; 95% confidence interval 6.3 to 22.6) that was equivalent to that observed in the sJIA-LD subset. These findings indicate that subjects with sJIA and HLA-DRB1*15 alleles are at high risk for developing both ILD and DRESS upon treatment with these medications. Therefore, we propose the use of pre-treatment HLA typing in sJIA to avoid administration of these agents to individuals at high-risk of developing sJIA-LD or DRESS. A manuscript describing this work was published in the Annals of the Rheumatic Diseases. 3. Treatment Guidelines for JIA. Over the past 5 years, we have partnered with the American College of Rheumatology and an international panel of experts in childhood arthritis to develop evidence-based guidelines for the management of juvenile idiopathic arthritis. Using the GRADE (Grading of Recommendations, Assessment, Development and Evaluation) methodology, we developed and published two sets of guidelines in 2019. In the current reporting period, we have developed and published two additional sets of guidelines. The first set of guidelines was Therapeutic Approaches for Oligoarthritis, Temporomandibular Joint (TMJ) Arthritis and Systemic Juvenile Idiopathic Arthritis and the second was Recommendations for Non-Pharmacologic Therapies, Medication Monitoring, Immunizations, and Imaging in Juvenile Idiopathic Arthritis. In both cases, these guidelines were heavily based on expert consensus, which highlights the relative paucity of good quality evidence in the published literature. These recommendations were published simultaneously in Arthritis & Rheumatology and Arthritis Care & Research.