Why does SARS-CoV-2 hit in different ways? Host genetic factors can influence the acquisition or the course of COVID-19

Abstract

The identification of high-risk factors for the infection by SARS-CoV-2 and the negative outcome of COVID-19 is crucial. The genetic background of the host might account for individual responses to SARS-CoV-2 infection besides age and comorbidities. A list of candidate polymorphisms is needed to drive targeted screens, given the existence of frequent polymorphisms in the general population. We carried out text mining in the scientific literature to draw up a list of genes referable to the term "SARS-CoV*". We looked for frequent mutations that are likely to affect protein function in these genes. Ten genes, mostly involved in innate immunity, and thirteen common variants were identified, for some of these the involvement in COVID-19 is supported by publicly available epidemiological data. We looked for available data on the population distribution of these variants and we demonstrated that the prevalence of five of them, Arg52Cys (rs5030737), Gly54Asp (rs1800450) and Gly57Glu (rs1800451) in MBL2, Ala59Thr (rs25680) in CD27, and Val197Met (rs12329760) in TMPRSS2, correlates with the number of cases and/or deaths of COVID-19 observed in different countries. The association of the TMPRSS2 variant provides epidemiological evidence of the usefulness of transmembrane protease serine 2 inhibitors for the cure of COVID-19. The identified genetic variants represent a basis for the design of a cost-effective assay for population screening of genetic risk factors in the COVID-19 pandemic.

Keywords: CD27; COVID19; Data mining; MBL2; SNPs; TMPRSS2.

Fig. 1

Procedure to identify candidate variantsThe Exome Aggregation Consortium (ExAC) provided a list of human genes and annotated them with a value that measures the probability of being loss-of-function intolerant (pLI). The names of the genes were associated to the names of the corresponding proteins. Such list was used as the input to look within the Scopus database for papers containing any of the protein names and the term “SARS-CoV*” in the article's title, abstract or keywords. “SARS-CoV referable genes”, encompassing frequent variants, missense or loss of function, were identified setting the threshold to allele frequency to 1%, and excluding variants that did not pass quality control or occur in non-canonical transcripts. The genes covered by the GWAS analysis (Group, 2020) were excluded. Among these genes, those encompassing frequent deleterious variants were identified.

Fig. 2

MBL2 allele frequencies per population correlation with COVID-19 cases and deaths. rs5030737 (Arg52Cys) allele frequencies in different countries from Verdu et al. were correlated with the respective number of cases (panel A; p-value 0.0394) and deaths (panel B; n. s.) of COVID-19. rs1800450 (Gly54Asp) allele frequencies were correlated both with cases (panel C; p-value 0.0018) and deaths (panel D; p-value 0.0199) of COVID-19 per country. Different symbols relate to different populations associated with the same country [Purple circle: GBP; purple square: GBB.] Details about populations and their association to countries are provided in supplementary file S1. . (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

MBL2 allele frequencies per population correlation with COVID-19 cases and deaths. rs1800451 (Gly57Glu) allele frequencies were correlated with cases (panel A; p-value 0.0122) and deaths (panel B; p-value 0.0241) of COVID-19 per country. The outsider Iran was excluded to observe a significant correlation with the number of deaths. Different symbols relate to different populations associated with the same country [Purple circle: GBP; purple square: GBB.] Details about populations and their association to countries are provided in supplementary file S1. . (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

TMPRSS2 allele frequencies per population correlation with COVID-19 cases and deaths. rs12329760 (Val197Met) allele frequencies in different countries from ALFRED were correlated with the respective number of cases (panel A; p-value 0.0012) and deaths (panel B; p-value 0.0057) of COVID-19. Different symbols relate to different populations associated with the same country [Blu circle: SA004049R; blu square: SA004617S. Light brown circle: SA001504K; light brown square: SA001503J. Red circle: SA004057Q; red square: SA002255O; red triangle SA001505L. Light green circle: SA004050J; light green square: SA001508O.] Details about populations and their association to countries are provided in supplementary file S1. . (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5

CD27 allele frequencies per population correlation with COVID-19 cases and deaths. rs25680 (Ala59Thr ) allele frequencies in different countries from ALFRED were correlated with the respective number of cases (panel A; n. s.) and deaths (panel B; p-value 0.0337) of COVID-19. Different colors relate to different countries while different symbols relate to different populations associated with the same country [Blue circle: SA004377V; blue square: SA004049R. Red circle: SA002255O; red square: SA004057Q; red triangle SA001505L.] Details about populations and their association to countries are provided in supplementary file S1. . (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)