Autoantibody Repertoires in Healthy Individuals Vary According to HLA Class II Genotype
Suseno, R.; Boquett, J. A.; Dandekar, R.; Ituarte, T.; Alvarenga, B. D.; Vierra-Green, C.; Spellman, S.; Maiers, M.; DeRisi, J. L.; Wilson, M.; Hollenbach, J. A.
Show abstract
The link between HLA genotype and the presence of pathogenic autoantibodies has previously been established across different autoimmune disorders. However, the functional process linking specific antibodies to specific HLA remains unclear. Additionally, autoantibodies - usually associated with autoimmune disease - are also present in healthy individuals. To this end, we sought to determine the spectrum of self-antigen antibody specificity (the "autoreactome") in healthy individuals, stratified by HLA genotypes. We utilized Phage ImmunoPrecipitation Sequencing (PhIP-Seq), a programmable phage display for interrogation of antibody specificity, which encompasses over 700,000 peptides tiled across the entire human proteome. Serum from 741 donors without diagnosed autoimmune disease were grouped by their five homozygous HLA-DRB1 genotypes and analyzed for differences in autoreactivity profiles. We applied a custom filter to the PhIP-Seq normalized data and obtained a set of enriched peptides for each HLA-DRB1 genotype. We found that the autoreactome in healthy individuals with different HLA-DRB1 genotypes are generally distinct. Binary logistic regression successfully identified whether a sample belongs to a specific HLA-DRB1 genotype (HLA-DRB1*01, HLA-DRB1*03, HLA-DRB1*04, HLA-DRB1*07, and HLA-DRB1*15) with the following accuracies: 96%, 92%, 90%, 94%, and 90%, and multinomial models predicted HLA-DRB1 genotype with up to 90% accuracy. Finally, gene-level analysis suggests that individuals with specific HLA autoimmune risk alleles may harbor potentially pathogenic autoantibodies in the absence of, or prior to the establishment of, overt disease. Our analysis demonstrates that autoreactivity profiles in healthy people vary according to HLA class II genotype, and may provide insight into the pathological processes associated with development of autoimmune and other immune-mediated diseases.
Matching journals
The top 9 journals account for 50% of the predicted probability mass.