A bioinformatics approach to investigating the structural and functional consequences of SNPs in TMPRSS2 for COVID‐19 infection

Abstract

SARS‐CoV‐2 is a highly infectious virus that is responsible for the COVID‐19 global pandemic that swept the world in 2020. Disease outcomes range from asymptomatic to fatal. The virus initiates entry into host cells by the binding of its spike protein to the ACE2 receptor. Entry is finalized by the activation of spike glycoprotein by proteases including transmembrane protease, serine 2 (TMPRSS2) and FURIN which cleave the spike protein of the virus. Single nucleotide polymorphisms (SNPs) in TMPRSS2 may lead to functional changes which could underlie differences in disease severity. TMPRSS2 is also known to activate different respiratory illnesses including coronaviruses and influenza A (Shen et al., 2020). Previous studies have shown that knockout TMPRSS2 mice appeared healthy, experienced a decrease in viral spread within the respiratory system, and had a less severe immune response when infected with SARS‐CoV and MERS‐CoV (Baughn et al., 2020). Thus, we asked whether genetic variations in TMPRSS2 in humans lead to differences in infection rates or severity of disease symptoms of SARS‐CoV‐2. We examined the NCBI dbSNP database to identify SNPs in the TMPRSS2 gene. As of 10 December 2020, we found there were 11,023 intron variants, 393 missense variants, 186 synonymous variants, 3 in‐frame insertion variants, 2 in‐frame deletion variants, and 1 initiator codon variant reported. To narrow these down to 23 SNPs of interest, we first searched the ClinVar database to identify SNPs with general clinical significance, followed by searching the literature to determine SNPs specifically related to SARS‐CoV‐2 severity. One missense variant, rs12329760, results in an amino acid substitution, V160M, which has been predicted to alter TMPRSS2 function. A subset of these SNPs show differences in frequency in world populations, and we wondered if these SNPs had structural and functional consequences for the protein. A crystal structure of TMPRSS2 is not currently available. To visualize the structural consequences of amino acid substitutions, we performed homology modeling on TMPRSS2 (UniProt O15393) using the structure prediction software HHPred, RaptorX, and SwissModel based on the ~30% similarity to hepsin. The predicted structures of TMPRSS2 with various amino acid substitutions were then docked to the SARS‐CoV‐2 spike protein using I‐TASSER and Haddock 2.4 to observe differences in binding interactions and therefore determine which sequence changes are predicted to alter binding interactions, potentially contributing to the wide variation of symptoms caused by COVID‐19.Baughn, L. B., Sharma, N., Elhaik, E., Sekulic, A., Bryce, A. H., & Fonseca, R. (2020). Targeting TMPRSS2 in SARS‐CoV‐2 Infection. Mayo Clinic proceedings, 95(9), 1989–1999. https://doi.org/10.1016/j.mayocp.2020.06.018Shen, L.W.; Mao, H.J.;, Wu, Y.L.; Tanaka,Y.; Zhang,W. (2017) TMPRSS2: A potential target for treatment of influenza virus and coronavirus infections, Biochimie, 142, 1‐10. https://doi.org/10.1016/j.biochi.2017.07.016