ACE2 expression in various human tissues
Among the 31 GTEx human tissues, small intestine, testis, kidney, heart, thyroid, and adipose tissue had the highest ACE2 expression levels, while blood, spleen, bone marrow, brain, blood vessel, and muscle had the lowest ACE2 expression levels (Fig. 1A). In lungs, colon, liver, bladder, and adrenal gland, ACE2 showed the medium expression levels (Fig. 1A). These results suggest that ACE2 are expressed in a wide variety of human tissues in addition to lungs. Furthermore, the Human Protein Atlas (HPA) database (http://www.proteinatlas.org/) shows that the ACE2 protein has relatively high expression levels in duodenum, small intestine, gallbladder, kidney, testis, seminal vesicle, colon, rectum, and adrenal gland. The HPA database also shows that the gastrointestinal tract (duodenum, small intestine, colon, and rectum), kidney, gallbladder, and male tissues (testis and seminal vesicle) have high expression levels of both ACE2 gene and protein. Taken together, these data indicate that: i) 2019-nCoV may infect other human tissues in addition to lungs; ii) Males may be more susceptible to 2019-nCoV infection than females. These indications have been confirmed by recent publications [2-4]. For example, Holshue et al. uncovered that stool from a patient with 2019-nCoV infection was positive for 2019-nCoV, suggesting that 2019-nCoV may infect the gastrointestinal tract [5]. Huang et al. reported the virus-related cardiac injury in five patients with 2019-nCoV infection [2]. A study of 99 cases of 2019-nCoV pneumonia revealed the increased susceptibility of males to virus infection [3].
We further compared ACE2 expression levels between males and females in 22 individual human tissues using the GTEx datasets and found no statistically significant difference between males and females in any tissue with a threshold of adjusted p value < 0.05 and fold change > 1.5 (Fig. 1B). These results may indicate that like SARS-CoV [11], 2019-nCoV can affect males and females equally, although males are likely to have higher mortality risk than females with coronavirus infections [3, 12]. We also compared ACE2 expression levels between younger (ages £ 49 years) and older (ages > 49 years) groups and did not observe statistically significant difference between both groups in any tissue with a threshold of adjusted p value < 0.05 and fold change > 1.5 (Fig. 1B). These results indicate that like SARS-CoV [11], 2019-nCoV can attack young and old persons equally, although old age is associated with higher risk of death for patients with coronavirus infections [3, 13]. Likewise, in the TCGA datasets, ACE2 was not differentially expressed between males and females and between younger and older groups in any tissue (Fig. 1C), consistent with the results in GTEx. In TCGA, we also compared ACE2 expression levels between Asia and other races in five normal tissues (stomach, thyroid, breast, liver, and pancreas), and did not found significant difference between them in any tissue (Fig. 1C).
Association of ACE2 expression with immune signatures in various human tissues
Adaptive and innate immune responses play an important role in fighting off invading coronavirus, while they may induce cytokine storm to cause the immunopathological damage of patients with coronavirus infections [14]. We analyzed the correlations between ACE2 expression levels and immune signature (CD8+ T cells, interferon response, B cells, and NK cells) enrichment levels in various human tissues for males and females, respectively. In skin, digestive system (esophagus, stomach, colon, and pancreas), brain, and blood vessel, we observed significant positive correlations between ACE2 expression levels and CD8+ T cell enrichment levels in both males and females (Pearson’s correlation test, adjusted p < 0.05, 0.27 £ r £ 0.78) (Fig. 2A). In addition, in thyroid, lungs, adrenal gland, liver, and kidney, ACE2 expression levels showed significant positive correlations with CD8+ T cell enrichment levels solely in males (0.20 < r < 0.68). However, in thyroid and lungs, ACE2 expression levels were negatively correlated with CD8+ T cell enrichment levels in females (r = -0.36). In heart, ACE2 expression levels had a negative and a positive correlation with CD8+ T cell enrichment levels in males (r = -0.23) and females (r = 0.32), respectively. Likewise, the interferon response signature had significant positive correlations with ACE2 expression levels in skin, digestive system (esophagus, stomach, liver, and pancreas), brain, and blood tissue in both males and females (0.14 £ r £ 0.75) (Fig. 2A). In thyroid, lungs, kidney, adrenal gland, colon, and bladder, ACE2 expression levels had significant positive correlations with the interferon response signature solely in males (0.32 £ r £ 0.82). In contrast, in thyroid, lungs, and colon, ACE2 expression levels were negatively correlated with the interferon response signature in females (-0.26 < r < -0.20). In heart, ACE2 expression levels had a negative and a positive correlation with the interferon response signature in males (r = -0.18) and females (r = 0.54), respectively. Similar results were observed for B cells and NK cells (Fig. 2A). Collectively, these results demonstrate the commonality and distinction in the association of ACE2 expression with immune signatures between males and females. Interestingly, the inflammation of the lungs (pneumonia) is the most common disease caused by coronavirus, while the correlations between its receptor expression and immune signatures in lungs are adverse between males and females. This indicates that the host immune response to coronavirus infections could be different between males and females, partially explaining why males and females have markedly distinct clinical outcomes of coronavirus infections [3, 12].
We further analyzed the correlations between ACE2 expression and immune signatures in the younger and older groups, respectively. In most individual tissues, the correlations between ACE2 expression levels and immune signature enrichment levels displayed consistent trends between the younger and older groups. However, in lungs they had a positive and a negative correlation in the older and younger groups (Fig. 2B). Again, these results suggest the potential difference in the host immune response to coronavirus infections between young and old persons.