Individual variation of the SARS-CoV-2 receptor ACE2 gene expression and regulation

Abstract

The COVID-19 coronavirus is now spreading worldwide. Its pathogen, SARS-CoV-2, has been shown to use angiotensin-converting enzyme 2 (ACE2) as its host cell receptor, same as the severe acute respiratory syndrome coronavirus (SARS-CoV) in 2003. Epidemiology studies found males although only slightly more likely to be infected than females account for the majority of the severely ill and fatality, which also bias for people older than 60 years or with metabolic and cardiovascular diseases. Here by analyzing GTEx and other public data in 30 tissues across thousands of individuals, we found a significantly higher level in Asian females, an age-dependent decrease in all ethnic groups, and a highly significant decrease in type II diabetic patients of ACE2 expression. Consistently, the most significant expression quantitative loci (eQTLs) contributing to high ACE2 expression are close to 100% in East Asians, >30% higher than other ethnic groups. A shockingly common enrichment of viral infection pathways was found among ACE2 anti-expressed genes, and multiple binding sites of virus infection related transcription factors and sex hormone receptors locate at ACE2 regulatory regions. Human and mice data analysis further revealed ACE2 expression is reduced in T2D patients and with inflammatory cytokine treatment and upregulated by estrogen and androgen (both decrease with age). Our findings revealed a negative correlation between ACE2 expression and COVID-19 fatality at both population and molecular levels. These results will be instrumental when designing potential prevention and treatment strategies for ACE2 binding coronaviruses in general.

Keywords: ACE2; COVID-19; SARS-CoV2.

Figure 1

ACE2 expression in different tissues in males, females of Asians, Africans, and Caucasians with age. Panels a–d and e–h are tissues where Asian females show significantly and moderately (marginally significant) higher expression than other groups, respectively. n, sample numbers in each group; k, p, and FDR are slope, p value and multiple testing corrected FDR of linear regression, respectively. The significance p values and FDR of the difference between Asians versus other ethnic groups are shown on top of the graphs. j, Summary statistics of association with each variable while controlling for other variables

Figure 2

Transcription regulation of human ACE2 gene. (a) Cis regulatory elements. eQTL to ACE2 gene expression in the ACE2 TSS‐10kb to TTS + 10 kb region are visualized together with chromatin modification ChIP‐seq signals, DNAse I sensitivity sites and TF and chromatin state regulators ChIP‐seq targets from ENCODE and ESR1, ESR2, and AR motifs identified by the FIMO function from MEME package. (b,c), The effect of estrogen (b) and testosterone (c) on ACE2 expression in mouse thymus and kidney, respectively

Figure 3

ACE2 expression is negatively with virus infection pathways and downregulation by inflammation cytokines and in T2D. (a) KEGG pathway enrichment of ACE2 correlated and anti‐correlated genes in each tissue were determined by GSEA (Methods). (b) Dosage and time‐dependent SARS‐CoV infection on ACE2 mRNA level. (c–d), Suppression of ACE2 expression by IL2 and IL‐7 treatment in mouse T cells. (e) Reduction of ACE2 expression in T2D human pancreas ductal cells compared to age‐matched controls as shown by scRNA‐seq data. f, Fraction of ACE2 + cells among ductal cells and the average ACE2 expression per ACE2 + ductal cell in healthy and T2D individuals. ACE2 + cells are defined by normalized expression level > 0.01. k and p, slope and p value of linear regression to age, respectively. The significance p values of the difference between healthy and T2D groups are shown on top of the graphs. (g) ACE2 expression patterns projected on the t‐SNE (t‐distributed stochastic neighbor embedding) plot from type 2 diabetes and healthy pancreas single‐cell RNA‐seq data. (h) Schematic illustrations of the negative correlation of high basal ACE2 level with CoVID‐19 severity/fatality at the population level (top, solid lines indicate our findings; dotted lines are knowledge derived from literature), and its anti‐correlation with virus infection pathway expression levels, upregulation by sex hormones, and suppression by inflammatory cytokine at the molecular level (bottom). (i) A putative model illustrating the potential regulations and functions of the ACE2 gene. *Function of immunosuppressive peptide and their proposed action on macrophage and monocyte are from (Gallaher & Gallaher, 2020; Sarzi‐Puttini et al., 2020)