EZH2-mediated H3K27me3 inhibits ACE2 expression

Abstract

The outbreak of corona virus disease 2019 (COVID-19) caused by SARS-CoV-2 infection is spreading globally and quickly, leading to emerging health issues. SARS-CoV-2 enters into and infects host cells through its spike glycoprotein recognizing the cell receptor Angiotensin-converting enzyme II (ACE2). Here, we noticed that ACE2 was further enhanced by SARS-CoV-2 infection. Human germ cells and early embryos express high level of ACE2. Notably, RNA-seq result showed that reduction of H3K27me3, but not H3K4/9/36me3, led to upregulation of Ace2 expression in mouse germ cell line GC-2. In agreement with this result, we found in human embryonic stem cells that ACE2 expression was significantly increased in absence of EZH2, the major enzyme catalyzing H3K27me3. ChIP-seq analysis further confirmed decrease of H3K27me3 signal and increase of H3K27ac signal at ACE2 promoter upon EZH2 knockout. Therefore, we propose that EZH2-mediated H3K27me3 at ACE2 promoter region inhibits ACE2 expression in mammalian cells. This regulatory pattern may also exist in other human cells and tissues. Our discovery provides clues for pathogenesis and targeted drug therapy towards ACE2 expression for prevention and adjuvant therapy of COVID-19.

Keywords: ACE2; Coronavirus; EZH2; H3.3; SARS-CoV-2.

Fig. 1

ACE2 expression in human tissues and coronavirus-infected human cells. (a) RNA-seq data of tissue samples from 14 different human tissues were analyzed to demonstrate average expression level of ACE2 gene. (b) Average expression level of ACE2 was upregulated by SARS-CoV-2 infection in primary human bronchial epithelial cells, while ACE2 was not upregulated by MERS-CoV infection in human lung adenocarcinoma Calu-3 cells. Both coronaviruses were incubated with cells at 2 MOI for 24 h.

Fig. 2

ACE2 expression was upregulated by inhibiting H3K27me3 modification or depleting EZH2. (a) Levels of mouse Ace2 transcript from RNA-seq data reveal change of Ace2 expression in mouse GC-2 cell lines after stably expression of H3.3 K4/9/27/36 M (H3.3 wildtype as control). Data show mean ± SD; n = 2 biological replicates. ∗ indicates P < 0.05, ns indicates P > 0.05. (b) Transcript levels of Gapdh from RNA-seq data revealing consistent expression levels of Gapdh in mouse GC-2 cell lines stably expressing H3.3 (control group) and H3.3 K4/9/27/36 M. Data show mean ± SD; n = 2 biological replicates. (c–e) The average expression level of ACE2 (c), ZSCAN4 (d) and POU5F1 (e) in human early embryos (mature oocytes and preimplantation embryos at zygote, 2-cell, 4-cell, 8-cell, morula and late blastocyst stage) by RNA-seq. (f–h) The level of ACE2 (f), cleavage embryonic gene (g) and pluripotent gene (h) expression by RNA-seq in three groups (WT, EZH2-deficient and EZH2 add-back human ESCs). Data show mean ± SD; n = 3 biological replicates. ∗∗ indicates P < 0.01, ns indicates P > 0.05.

Fig. 3

EZH2 knockout in human ESCs resulted in H3K27me3 decrease and H3K27ac increase. (a–d) Normalized reads of H3K27me3 (a), H3K27ac (b), H3K4me1 (c) and H3K4me3 (d) at the promoters of ACE2 and Nanog loci in wildtype, EZH2-deficient and EZH2 add-back human ESCs.

Fig. 4

Scheme of transcriptional regulation of ACE2 gene. Expression of ACE2 gene can be modulated by cell type, SARS-CoV-2 infection, epigenetic factors such as EZH2, etc.