For the in vitro characteristic study of human ACE2, human cell lines BEAS-2B cells and HMC3 cells were stimulated with IFNβ, viral RNA mimic poly(I:C) or viral DNA mimic poly(dA:dT) as indicated. We found that treatments of IFNβ, viral RNA mimic poly(I:C) and viral DNA mimic poly(dA:dT) signifcantly increased ACE2 mRNA levels in BEAS-2B cells (Figure 1A), along with significant enhancement of IFNβ, TNFα and IL6 levels (Figures 1B–D). Consistently, treatments of IFNβ, viral RNA mimic poly(I:C) and viral DNA mimic poly(dA:dT) in HMC3 cells also signifcantly increased the mRNA levels of ACE2, IFNβ, TNFα and IL6 (Figures 1E–H). Furthermore, we found that the protein levels of ACE2 were also increased upon IFNβ, viral RNA mimic poly(I:C) and viral DNA mimic poly(dA:dT) treatment (Figure 1I). Together, these results suggested that human ACE2 was an ISG in human cell lines.

Luciferase reporter system was constructed using human ACE2 promoter (−1119 to 103) and pGL3-luciferase reporter vector (Figure 2A). We found that overexpression of MAVS caused significant elevation in ACE2 luciferase activities (Figure 2B), suggesting the promotive effect of MAVS signaling in ACE2 expression. To further confirm the role of RLR signaling on ACE2 expression, key components MAVS signaling: retinoic acid inducible gene-I (RIG-I) and melanoma differentiation associated gene 5 (MDA-5) were overexpressed. In the results, we found that both RIG-I and MDA-5 significantly increased the level of ACE2 (Figure 2C).

In our previous study, through ISG library screening, we found TNFR1 positively regulated RLR signaling (Cheng et al., 2016). However, it was still unclear how ISGs regulate ACE2 levels and NF-κB activities. To further study the effect of ISGs on ACE2, IFNβ and NF-κB activities, an ectopic expression assay of a library containing 117 human ISGs was performed (Figure 2D). Each ISG was co-transfected into HEK293T cells with ACE2 luciferase reporter, IFNβ luciferase reporter or NF-κB luciferase reporter. In the results, we found that there were 13 ISGs (ISG10, ISG52, ISG62, ISG63, ISG71, ISG88, ISG90, ISG92, ISG95, ISG96, ISG103, ISG104 and ISG109) significantly promoted the expression of ACE2 (>2-fold). Notably, the highest ACE2 level was induced by ISG95 expression (Figure 2E). Furthermore, we found that there were 19 ISGs (ISG6, ISG10, ISG32, ISG41, ISG53, ISG54, ISG55, ISG57, ISG59, ISG81, ISG82, ISG86, ISG90, ISG92, ISG94, ISG95, ISG96, ISG9103 and ISG104) significantly increased NF-κB activities (>2-fold) (Figure 2F). Meanwhile, as shown in Figure 2G, only ISG32, ISG41, ISG43, ISG95 and ISG120 were found to increase IFNβ levels (>2-fold) (Cheng et al., 2016). There results suggested that the expression of ACE2, IFNβ and NF-κB were regulated by different ISGs.

For further investigation of the effects of ISGs on ACE2, IFNβ and NF-κB levels, corrlation analysis was performed. Interestingly, we found that ACE2 levels were highly correlated with NF-κB activities (R value of 0.8169 and p value <0.0001) (Figure 3A), but not with IFNβ levels (R value of 0.1588 and p value of 0.09) (Figure 3B). Meanwhile, IFNβ levels were correlated with NF-κB activities (R value of 0.2960 and p value <0.01) (Figure 3C).

Among the ACE2-regulatory ISGs, ISG10 (CASP10), ISG52 (XRN1), ISG71 (FTSJD2), ISG88 (SEMA4G) and ISG95 (TNFR1) were further analyzed for their effects on ACE2, NF-κB and IFNβ levels (Figure 3D). In the results, we found that ISG10 (CASP10), ISG52 (XRN1), ISG71 (FTSJD2), ISG88 (SEMA4G) and ISG95 (TNFR1) all significantly increased ACE2 levels (Figure 3E). Meanwhile, ISG10 (CASP10), ISG52 (XRN1), ISG71 (FTSJD2), and ISG95 (TNFR1) caused significant increase of NF-κB level (Figure 3F). Meanwhile, only ISG71 (FTSJD2) and ISG95 (TNFR1) significantly increased the level of IFNβ (Figure 3G). Moreover, as overexpression of ISG95 increased the highest luciferase activities of ACE2, NF-kB and IFNβ, we examined the effects of overexpression of ISG95 on ACE2, IL6 and IFNβ levels in BEAS-2B cells. As shown in Figure 3H, overexpression of ISG95 significantly increased the mRNA levels of ACE2, IL6 and IFNβ in BEAS-2B cells. Together, these results suggested that ISGs-induced ACE2 levels were significantly correlated with NF-κB level, but not IFNβ level.

Multiple evidences showed that increased ACE2 levels and NF-κB downstream cytokine levels were presented in patients with severe COVID-19 symptom (Pinto et al., 2020; Soy et al., 2020; Zhuang et al., 2020; Rabaan et al., 2021). Together with our findings that ACE2 levels were highly correlated with NF-κB levels. Therefore, we hypothesized that treatment targeting NF-κB signaling potentially also produced inhibitory effects on ACE2 level, which was important for the prevention and clinical treatment of COVID-19 infection and the infection-caused inflammatory cytokine storm. 12 clinical approved drugs targeting NF-κB signaling were investigated for their effects on the levels of ACE2, IFNβ and NF-κB downstream cytokines (Figure 4A). Among these 12 approved clinical drugs, we found that Cepharanthine and Glucosamine significantly inhibited viral RNA mimic poly(I:C)-induced upregulation of ACE2, IFNβ, and NF-κB signaling downstream TNFα and IL6 levels (Figures 4B–E). The inhibitory effects of Cepharanthine and Glucosamine were validated and shown in Figure 4F. Notably, solo use of Cepharanthine or Glucosamine inhibited viral RNA mimic poly(I:C)-induced upregulation of ACE2, however dual treatment of both drugs failed to further downregulate the level of ACE2. Consistently, similar results were observed in the regulation of IFNβ, TNFα and IL6 levels (Figures 4G–I). Furthermore, we found that Cepharanthine and Glucosamine treatment not only inhibited p-p65 and p-IκB levels, but also decreased ACE2 proteins levels (Figures 4J,K). These results suggested that clinical drugs Cepharanthine and Glucosamine not only inhibited ACE2 levels, but also suppressed the expression of IFNβ and NF-κB signaling.

In summary, through ISGs library screening, we found ACE2 levels were positively regulated by multiple ISGs. ACE2 levels were high correlated with ISGs-induced NF-κB levels. Furthermore, through clinical approved drugs screening, we found that clinical drugs Cepharanthine and Glucosamine significantly inhibited the expression of ACE2, IFNβ, TNFα and IL6, suggesting the possible protective role of those drugs in the prevention and clinical treatment of coronaviruses infection.

BEAS-2B cell line, HMC3 cell line and HEK 293T cell line were obtained from the American Type Culture Collection and cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco) containing 10% fetal bovine serum and 1% penicillin/streptomycin at 37°C in a humidified atmosphere with 5% CO₂.

Poly (I:C) and Poly (dA:dT) were purchased from InvivoGen. IFNβ was purchased from Biolegend. The approved drug library was purchased from Topscience Co., Ltd.

ISG expression library was provided by Dr. Guangxia Gao (Institute of Biophysics, Chinese Academy of Science, China) and used in our previous study (Cheng et al., 2016).

Briefly, promoter regions from −1119 to 103 of human ACE2 was cloned into a pGL3-luciferase reporter vector. The sequence of plasmids was validated by sequencing. NF-κB luciferase reporter plasmid and IFNβ reporter luciferase plasmid were previously constructed and used in our previous studies (Cheng et al., 2016; Li et al., 2020). Luciferase activity were measured using Dual luciferase reporter assay system according to the manufacture’s protocol (Promega).

Total RNA was extracted from indicated cell samples using Trizol reagent (Invitrogen, cat#15596018). 1 µg of RNA was used for the synthesize of cDNA using a one-step first strand cDNA synthesis kit (Transgen Biotech, cat#AT341). Quantitative real-time PCR was performed using 2 × SYBR Green PCR master mix (Transgen Biotech, cat#AQ131) and Agilent Mx3005P RT-PCR system. The expressions of tested genes were normalized to the expression of GAPDH, and the 2^−ΔΔCT method was used to analyze the relative changes in gene expression.

The primers for human ACE2, IFNβ, TNFα, IL6 and GAPDH were listed below:

Human ACE2: Forward: 5′-CGA​AGC​CGA​AGA​CCT​GTT​CTA-3′; Reverse: 5′-GGG​CAA​GTG​TGG​ACT​GTT​CC-3′;

Human IFN-β: Forward: 5′-ATG​ACC​AAC​AAG​TGT​CTC​CTC​C-3′; Reverse: 5′-GGA​ATC​CAA​GCA​AGT​TGT​AGC​TC-3′;

Human TNFα: Forward: 5′-CCT​CTC​TCT​AAT​CAG​CCC​TCT​G′; Reverse: 5′-GAG​GAC​CTG​GGA​GTA​GAT​GAG-3′;

Human IL6: Forward: 5′-ACT​CAC​CTC​TTC​AGA​ACG​AAT​TG′; Reverse: 5′-CCA​TCT​TTG​GAA​GGT​TCA​GGT​TG-3′;

Human GAPDH: Forward: 5′-GGA​GCG​AGA​TCC​CTC​CAA​AAT-3′; Reverse: 5′-GGC​TGT​TGT​CAT​ACT​TCT​CAT​GG3′;

Extracted proteins were separated by polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes (GE Amersham, catalog no. 10600002), and incubated with the primary antibody for overnight at 4°C. Immunoblots were probed with the first antibody with anti-phospho-IκBα (Ser32) (#2859, Cell Signaling Technology), anti-IκBα (44D4) (#4814, Cell Signaling Technology), anti-phospho-p65 (Ser536) (#3033, Cell Signaling Technology), anti-p65 (D14E12) (#8242, Cell Signaling Technology), anti-ACE2 (#ab108209, Abcam) and β-actin (#CW0096M, CWBiotech). And ECL luminescent solution was used for detection.

Student’s t-test was used for comparisons between 2 groups. One-way analysis of variance (ANOVA) was used for multiple groups and correlation analysis. All statistical analysis were performed using GraphPad (Prism GraphPad Software). All values were expressed as the mean ± SEM. p value <0.05 were considered as significant.