Retinoic Acid-Mediated Inhibition of Mouse Coronavirus Replication Is Dependent on IRF3 and CaMKK

Abstract

The ongoing COVID-19 pandemic has revealed the shortfalls in our understanding of how to treat coronavirus infections. With almost 7 million case fatalities of COVID-19 globally, the catalog of FDA-approved antiviral therapeutics is limited compared to other medications, such as antibiotics. All-trans retinoic acid (RA), or activated vitamin A, has been studied as a potential therapeutic against coronavirus infection because of its antiviral properties. Due to its impact on different signaling pathways, RA's mechanism of action during coronavirus infection has not been thoroughly described. To determine RA's mechanism of action, we examined its effect against a mouse coronavirus, mouse hepatitis virus strain A59 (MHV). We demonstrated that RA significantly decreased viral titers in infected mouse L929 fibroblasts and RAW 264.7 macrophages. The reduced viral titers were associated with a corresponding decrease in MHV nucleocapsid protein expression. Using interferon regulatory factor 3 (IRF3) knockout RAW 264.7 cells, we demonstrated that RA-induced suppression of MHV required IRF3 activity. RNA-seq analysis of wildtype and IRF3 knockout RAW cells showed that RA upregulated calcium/calmodulin (CaM) signaling proteins, such as CaM kinase kinase 1 (CaMKK1). When treated with a CaMKK inhibitor, RA was unable to upregulate IRF activation during MHV infection. In conclusion, our results demonstrate that RA-induced protection against coronavirus infection depends on IRF3 and CaMKK.

Keywords: CaMKK; IRF3; MHV; coronavirus; retinoic acid.

Figure 1

A total of 100 µM RA upregulates antiviral mRNA expression. In L929 cells infected with MHV at MOI 0.1, RA increased (A) Ifit1, (B) Ifit2, (C), and IFNβ mRNA expression. When administered to cells infected with MHV at MOI 1, RA decreased (D) MHV-N mRNA and increased (E) Ifit1 and (F) Ifit3 expression. Data were taken from experiments performed in triplicate. Statistical significance was calculated by Dunnett’s One-Way ANOVA, with comparisons to DMSO labeled as “A” and comparisons to MHV labeled as “B”. * p < 0.05, ** p < 0.01.

Figure 2

RA reduces MHV replication in L929 cells. (A) RA reduced MHV titers at all concentrations (i.e., 1 µM, 10 µM, and 100 µM), except 0.1 µM. (B) Representative plaques demonstrate decreased plaque formation at RA concentrations above 0.1 µM. Data were taken from experiments performed in triplicate. Statistical significance was calculated by Dunnett’s One-Way ANOVA, with comparisons to MHV labeled as “B”. * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 3

A total of 1 µM and 10 µM RA upregulates antiviral mRNA expression. (A) RA downregulates MHV-N mRNA expression in WT cells but not in IRF3-KO cells. IRF3-KO cells exhibited increased expression of (B) Ifit1 and (C) Ifit3 when treated with RA. Data were taken from experiments performed in triplicate. Statistical analysis was calculated by Dunnett’s One-Way ANOVA, with comparisons to WT-DMSO labeled as “A”, comparisons to WT-MHV labeled as “B”, comparisons to IRF3-KO-DMSO labeled as “C”, and comparisons to IRF3-KO-MHV labeled as “D”. * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 4

RA-induced suppression of MHV is IRF3-dependent. (A–C) A total of 1 µM RA significantly reduced MHV-N protein expression in WT cells but not IRF3-KO cells. (D) RA decreased MHV titers in WT cells and exhibited no effect in IRF3-KO cells. (E,F) Representative plaques demonstrate that RA reduced MHV replication in infected WT cells. Data were taken from experiments performed in triplicate. Statistical analysis was calculated by Dunnett’s One-Way ANOVA, with comparisons to WT-MHV labeled as “A” and comparisons to IRF3-KO-MHV labeled as “B”. * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 5

RA-mediated MHV suppression is dependent on IRF3 and CaMKK. (A) A total of 10 µM RA induced IRF activation in infected WT cells but not IRF3-KO cells. In IRF3-KO cells, only 4000 IU/mL of IFNβ could induce IRF activation. STO-609 (STO) disrupted RA’s ability to induce (B) IRF activation and reduce (C) MHV titers in WT cells. (D) Representative plaques demonstrate that STO disrupts RA’s antiviral activity. Data were taken from experiments performed in triplicate and quadruplicate. Statistical significance was calculated by Dunnett’s One-Way ANOVA, with comparisons to WT-DMSO labeled as “A”, comparisons to WT-MHV labeled as “B”, comparisons to IRF3-KO-DMSO labeled as “C”, and comparisons to IRF3-KO-MHV labeled as “D”. * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 6

RA induces widespread gene expression in WT cells. (A) Heatmap illustrates differential gene expression between uninfected and infected WT cells. Volcano plots show that in both (B,C) uninfected and (D,E) infected WT cells, RA upregulates CaMKK1 expression. Data were taken from experiment performed in singlicate.

Figure 7

RA induces widespread gene expression in IRF3-KO cells. (A) Heatmap illustrates differential gene expression between uninfected and infected IRF3-KO cells. Volcano plots show that in both (B,C) uninfected and (D,E) infected IRF3-KO cells, RA upregulates CaMKK1 expression. Data were taken from experiment performed in singlicate.