253 Signatures of microbial and autoantibody epitopes in idiopathic inflammatory myopathies

Abstract

Background: The idiopathic inflammatory myopathies are heterogeneous autoimmune disorders characterised by inflammation of muscle tissue and are thought to result from environmental exposures in genetically susceptible individuals. We identified recently genetic association to the chemokine receptor 5 gene in inclusion body myositis (IBM); this suggests a potential mechanism relating to disease pathogenesis whereby reduced frequency of the CCR5≤32 deletion in IBM increases viral entry into host cells. Similarly, antibodies against tripartite motif-containing proteins reduce ability to restrict infectivity and replication of retroviruses, suggesting a mechanism relating to disease pathogenesis in dermatomyositis patients with anti-TIF1 antibodies. We propose to identify disease-subtype specific signatures of past or current microbial infections and autoantibody epitopes in subgroups of individuals with IIM. Methods: To achieve this, a novel unbiased high throughput approach that allows the identification of peptide sequences against which an IgG response has been raised was used. We focused our investigation initially on pooled plasma from the following four myositis patient groups (UKMYONET) compared to age-matched healthy controls: Patients with known anti-histidyl-tRNA-synthetase (anti-Jo-1) autoantibodies as proof of principle for antigen epitope identification; pool of 10 and 20 samples. Patients with IBM who are: (a) heterozygote for the CCR5 deletion haplotype compared to (b) those that do not carry the deletion haplotype; pool of eight samples. TIF1 autoantibody positive adult dermatomyositis patients; pool of 10 and 20 samples. Ethical approval has been obtained from Multi-Centre Research Ethics Committee North West, and University of Manchester Research Ethics Committee Results: IgG was isolated from 84 serum samples and immobilised for subsequent competitive panning and clonal expansion with a bacterial random-peptide surface display system (2 x 108 random 12 amino acid peptides). After clonal expansion, plasmids were isolated and the DNA fragment which encodes for the 36 bp random peptides was amplified using specific primers. The amplification product was extracted and verified with Sanger sequencing. Pair-end DNA libraries (Nextera DNA Library Prep) were constructed and sequenced using short read pair-end high throughput sequencing (Illumina). The resulting peptide sequences are searched for homology against the entire public protein database to identify any sequences associated with microbes and/or endogenous biomarkers using a combination of alignment-based (BLAST) and alignment-free (multiple k-mers and Naïve Bayesian Classifier) approaches. Annotated microbial sequences further undergo taxonomic classification to allow analysis of the microbial diversity in multiple taxonomic levels (Family, Genus, and Species). Conclusion: We anticipate that we will identify signatures of microbial infections which are unique, or more common, in subgroups of individuals with IIM. Identification of a highly associated microbial relationship with particular forms of IIM could be of major importance to our understanding and potentially how we manage the condition, vital to improve diagnosis and stratify treatment for improved clinical outcome.