Abstract
INTRODUCTION: Severe COVID-19 illness is characterized by an overwhelming immune hyperactivation. Autoantibodies against vascular, tissue, and cytokine antigens have been detected across the spectrum of COVID-19. How these autoantibodies correlate with COVID-19 severity is not fully defined. METHODS: We performed an exploratory study to investigate the expression of vascular and non-HLA autoantibodies in 110 hospitalized patients with COVID-19 ranging from moderate to critically ill. Relationships between autoantibodies and COVID- 19 severity and clinical risk factors were examined using logistic regression analysis. RESULTS: There were no absolute differences in levels of expression of autoantibodies against angiotensin II receptor type 1 (AT1R) or endothelial cell proteins between COVID-19 severity groups. AT1R autoantibody expression also did not differ by age, sex, or diabetes status. Using a multiplex panel of 60 non- HLA autoantigens we did identify seven autoantibodies that differed by COVID-19 severity including myosin (myosin; p=0.02), SHC-transforming protein 3 (shc3; p=0.07), peroxisome proliferator-activated receptor gamma coactivator 1-beta (perc; p=0.05), glial-cell derived neurotrophic factor (gdnf; p=0.07), enolase 1 (eno1; p=0.08), latrophilin-1 (lphn1; p=0.08), and collagen VI (coll6; p=0.05) with greater breadth and higher expression levels seen in less severe COVID-19. DISCUSSION: Overall, we found that patients hospitalized with COVID-19 demonstrate evidence of auto-reactive antibodies targeting endothelial cells, angiotensin II receptors, and numerous structural proteins including collagens. Phenotypic severity did not correlate with specific autoantibodies. This exploratory study underscores the importance of better understanding of the role of autoimmunity in COVID-19 disease and sequelae.