Computational Modeling of T Cell Hypersensitivity during Coronavirus Infections Leading to Autoimmunity and Lethality

Abstract

The human angiotensin-converting enzyme 2 (hACE2) receptor is the primary receptor for SARS-CoV-2 infection. However, the presence of alternative receptors such as the transmembrane glycoprotein CD147 has been proposed as a potential route for SARS-CoV-2 infection. The outcomes of SARS-CoV-2 spike protein binding to receptors have been shown to vary among individuals. Additionally, some patients infected with SARS-CoV-2 develop autoimmune responses. Given that CD147 is involved in the hyperactivation of memory T cells resulting in autoimmunity, we investigated the interaction of the SARS-CoV-2 viral spike protein with CD147 receptor and retinal specific CD147 Ig0 domain in silico using molecular docking and molecular dynamics (MD) simulations. The results indicated that binding involves two critical residues Lys63 and Asp65 in a ubiquitous CD147 isoform, potentially leading to the hyperactivation of T cells for only SARS-CoV-2, but not for SARS-CoV or MERS-CoV. Overall binding was confirmed by docking simulations. Next, MD analyses were completed to verify the docking poses. Polar interactions suggested that the interaction via Lys63 and Asp65 might be one of the determinants associated with severe COVID-19 outcomes. Neither did SARS-CoV nor MERS-CoV bind to these two critical residues when molecular docking analyses were performed. Interestingly, SARS-CoV was able to bind to CD147 with a lower affinity (-4.5 kcal/mol) than SARS-CoV-2 (-5.6 kcal/mol). Furthermore, Delta and Omicron variants of SARS-CoV-2 did not affect the polar interactions with Lys63 and Asp65 in CD147. This study further strengthens the link between SARS-CoV-2 infection and autoimmune responses and provides novel insights for prudent antiviral drug designs for COVID-19 treatment that have implications in the prevention of T cell hyperactivation.