Triggering type 1 diabetes post-covid: molecular mimicry?

Abstract

Abstract Objective To evaluate the possible similarity between the AA sequences of human insulin and human glutamic acid decarboxylase-65 (GAD65) with the SARS-CoV-2/COVID proteins to explain the possible trigger of DM1. Methods AA sequences of human insulin, GAD65 and SARS-CoV-2 were obtained from the Protein Data Bank archive information database (RCSB PDB). NetMHCpan v4.1 was used for epitope prediction. Sequences were compared using BLAST for epitope comparison and Pairwise Structure Alignment to assess protein similarity. The AA sequences of human insulin (4F0N) and GAD65 (2OKK) were compared with the sequences of the following SARS-CoV-2 proteins: SARS-Cov2 S protein at open state (7DDN), SARS-Cov2 S protein at close state (7DDD), SARS CoV-2 Spike protein (6ZB5), Crystal structure of SARS-CoV-2 nucleocapsid protein N-terminal RNA binding domain (6M3M), Crystal structure of SARS-CoV-2 nucleocapsid protein C-terminal RNA binding domain (7DE1), Crystal structure of NSP1 from SARS-CoV-2 (7K3N), and SARS-CoV-2 S trimer (7DK3)). Results The percent similarity between epitopes ranged from 45 to 60% (P 0.048) between both human insulin and SARS-CoV2 and for GAD 65 and SARS-CoV2, while the AA similarity of the evaluated samples ranged from 5.00–45.45% between human insulin and SARS-CoV2 and from 10.45–22.22% between GAD65 and SARS-CoV2. Conclusion Immunoinformatics data suggest a potential pathogenic link between SARS-CoV-2/COVID and DM1. Thus, by molecular mimicry, we found that sequence similarity between epitopes and AA sequence between SARS-CoV-2 / COVID and human insulin and GAD65 could lead to the production of an immune cross-response to self-antigens, with self-tolerance breakdown, which could thus trigger DM1.