Early endonuclease-mediated evasion of RNA sensing ensures efficient coronavirus replication

Abstract

Coronaviruses are of veterinary and medical importance and include highly pathogenic zoonotic viruses, such as SARS-CoV and MERS-CoV. They are known to efficiently evade early innate immune responses, manifesting in almost negligible expression of type-I interferons (IFN-I). This evasion strategy suggests an evolutionary conserved viral function that has evolved to prevent RNA-based sensing of infection in vertebrate hosts. Here we show that the coronavirus endonuclease (EndoU) activity is key to prevent early induction of double-stranded RNA (dsRNA) host cell responses. Replication of EndoU-deficient coronaviruses is greatly attenuated in vivo and severely restricted in primary cells even during the early phase of the infection. In macrophages we found immediate induction of IFN-I expression and RNase L-mediated breakdown of ribosomal RNA. Accordingly, EndoU-deficient viruses can retain replication only in cells that are deficient in IFN-I expression or sensing, and in cells lacking both RNase L and PKR. Collectively our results demonstrate that the coronavirus EndoU efficiently prevents simultaneous activation of host cell dsRNA sensors, such as Mda5, OAS and PKR. The localization of the EndoU activity at the site of viral RNA synthesis-within the replicase complex-suggests that coronaviruses have evolved a viral RNA decay pathway to evade early innate and intrinsic antiviral host cell responses.

Fig 1. The CoV endoribonuclease is essential for replication and spread in vivo.

(a) Genome organization of the EndoU-deficient murine hepatitis virus (MHV) with an active site His to Ala substitution (MHV_H277A) and a corresponding human coronavirus 229E mutant (HCoV-229E_H250A) in the non-structural protein 15. (b) Replication kinetics of MHV-A59 and MHV_H277A in murine L929 fibroblasts after infection at a MOI of 1 and 0.1, presented as viral titer in plaque forming units (pfu). Data represent two independent experiments, each performed in duplicates. Mean and SEM are depicted. The 95% confidence band is highlighted in grey. The differences in peak levels of viral titers were calculated by using the non-linear regression model described in Material and Methods (peak MHV-A59: 6.0, MHV_H277A: 5.6, p = 0.024, left panel; peak MHV-A59: 6.6, MHV_H277A: 6.0, p = 0.016, right panel) and significance is displayed as * p < 0.05. (c) Viral titers of MHV-A59 and MHV_H277A in liver and spleen of C57BL/6, IFNAR^-/-, Mda5^-/-, TLR7^-/-, and Mda5^-/-/TLR7^-/- mice at two days post intraperitoneal infection (500 pfu). Data represent three to four independent experiments, each based on two to three mice per strain and virus. Mean and SEM are depicted. Data points that show significant differences in a two-sided, unpaired Student’s t-test are displayed; * p < 0.05, ** < 0.01, *** < 0.001. ND, not detected.

Fig 2. EndoU-deficient coronaviruses are severely attenuated in primary macrophages and trigger an elevated IFN-I response.

(a) Replication kinetics of MHV-A59 and MHV_H277A in C57BL/6 mouse embryonic fibroblasts (MEFs) after infection at a MOI of 1, presented as viral titer in pfu. Data represent four independent experiments, each performed in two to four replicas. The difference in peak levels of viral titers (peak MHV-A59: 5.9, peak MHV_H277A: 4.8) was statistically significant (***, p<0.001). (b) Replication kinetics of MHV-A59 and MHV_H277A (left panel; titers in pfu) and cell-associated viral RNA (right panel; qRT-PCR) following infection of C57BL/6 bone marrow-derived macrophages (MOI = 1). Data represent eight (left panel) or five (right panel) independent experiments, each performed in two to three replicas. The difference in peak levels of viral titers (left panel: peak MHV-A59: 5.3, MHV_H277A: 3.3) and the difference in peak levels of RNA copies (right panel: peak MHV-A59: 9.5, MHV_H277A: 8.5) were statistically significant (p = 0.002, p = 0.018, respectively). (c) Expression of IFN-β mRNA (left panel; qRT-PCR) and protein (right panel; ELISA) in C57BL/6 macrophages following infection of MHV-A59 and MHV_H277A (MOI = 1). Expression of IFN-β mRNA was normalized to levels of the household genes GAPDH and Tbp and is displayed relative to mock as ΔΔC_T. The IFN-β ELISA detection limit is indicated with a dashed line. Data represent seven (left panel) or eight (right panel) independent experiments, each performed in two to three replicas. The difference in peak levels of IFN-β mRNA (MHV-A59: 13.4, MHV_H277A: 15.8) was statistically significant (p = 0.04). Significance of IFN-β protein at 9 h.p.i. was assessed by a two-sided, Wilcoxon matched-pairs test (p = 0.016). (a-c) Mean and SEM are depicted. The 95% confidence band is highlighted in grey. Statistically significant comparisons are displayed; * p < 0.05, ** < 0.01, *** < 0.001. (d) Titers (pfu) of HCoV-229E wild type and HCoV-229E_H250A (left panel) and expression of IFN-I (right panel; IFN-I bioassay) in human blood-derived macrophages, 24 hours after infection (MOI = 1). Data represent six (left panel) or seven (right panel) independent experiments, each performed in three to four replicas. Significance was assessed by a two-sided, unpaired Student’s t-test (left panel, p<0.001) and a Wilcoxon matched-pairs test (right panel; p = 0.016). (c-d) Mean and SEM are depicted. Statistically significant comparisons are displayed; * p < 0.05, *** < 0.001.

Fig 3. Replication of MHV_H277A in primary macrophages with deficiencies in the IFN-I induction pathway.

Kinetics of viral replication (left panels) and IFN-β mRNA (right panels) in bone marrow-derived macrophages deficient for Mda5 (a), MAVS (b) and in macrophages triple knockout for IRF3, IRF5 and IRF7 (c) following infection with MHV-A59 and MHV_H277A (MOI = 1). (d) IFN-β in the supernatant of infected macrophages was measured using an ELISA. All values were outside of the detection limit of 15.6 pg/ml (dashed line) and thus depicted as ND (not detected). (a-d) Data represent three independent experiments, each performed in two to three replicas. (a) The difference in peak levels of viral titers (MHV-A59: 4.9, MHV_H277A: 3.0) was statistically significant (p<0.001), the difference in peak levels of IFN-β expression (MHV-A59: 9.2, MHV_H277A: 7.8) was statistically not significant (p = 0.44). (b) The differences in peak levels of viral titers (MHV-A59: 5.5, MHV_H277A: 3.9) and IFN-β expression (MHV-A59: 9.1, MHV_H277A: 12.1) were statistically significant (p<0.001, p = 0.024, respectively). (c) The difference in peak levels of viral titers (MHV-A59: 5.5, MHV_H277A: 4.9) was statistically significant (p = 0.002). The difference in peak levels of IFN-β expression (MHV-A59: 6.5, MHV_H277A: 5.1) was statistically not significant (p = 0.368). Mean and SEM are depicted. The 95% confidence band is highlighted in grey. Statistically significant comparisons are displayed; * p < 0.05, ** < 0.01, *** < 0.001.

Fig 4. Replication of EndoU-deficient MHV is partially restored in IFNAR^-/- macrophages and EndoU mutants display a pronounced sensitivity to IFN-I treatment.

(a) Replication kinetics of MHV-A59 and MHV_H277A (left panel; titers in pfu) and cell-associated viral RNA (right panel; qRT-PCR) following infection of IFNAR^-/- bone marrow-derived macrophages (MOI = 1). Data represent four independent experiments, each performed in two to three replicas. Mean and SEM are depicted. The 95% confidence band is highlighted in grey. The differences in peak levels of viral titers (MHV-A59: 6.0, MHV_H277A: 5.2) and RNA copies (MHV-A59: 9.7, MHV_H277A: 9.3) were statistically significant (p<0.001, p = 0.032, respectively). (b) Expression of IFN-β mRNA (left panel; qRT-PCR) and protein (right panel; ELISA) in IFNAR^-/- macrophages following infection of MHV-A59 and MHV_H277A (MOI = 1). Data represent four (left panel) and three (right panel) independent experiments, each performed in two to three replicas. Median and the 1–99 percentiles are displayed. Dashed line depicts limit of detection (right panel). The difference in peak levels of IFN-β expression (MHV-A59: 9.4, MHV_H277A: 13.8) was statistically significant (p = 0.002). Significance of IFN-β expression was assesses by a Wilcoxon matched-pairs test, * p < 0.05. ND, not detected. (c) Sensitivity of wild type and EndoU-deficient MHV (left panel) and HCoV-229E (right panel) viruses to IFN-I pre-treatment (4 h) in L929 cells (left panel) and MRC-5 cells (right panel) with various dosages of IFN-I (MOI = 1). Virus replication was measured at 24 h.p.i. by plaque assay (MHV) and at 48 h.p.i. by qRT-PCR (HCoV-229E), respectively. Data represent three independent experiments, each performed in two to three replicas. Data are displayed as differences to untreated controls and statistical comparisons between wild type and EndoU-deficient viruses were performed for each concentration. Mean and SEM are displayed. Data points that show significant differences in a two-sided, unpaired Student’s t-test are depicted. * p < 0.05, ** p < 0.01 and *** p < 0001.

Fig 5. EndoU-deficient MHV induces activation of the OAS-RNase L pathway, resulting in early breakdown of ribosomal RNA.

(a) Analysis of rRNA integrity in bone marrow-derived macrophages derived from wild type C57BL/6, Mda5^-/-, RNase L^-/-, and IFNAR^-/- mice following infection with MHV-A59 and MHV_H277A (MOI = 1). Total RNA was isolated at indicated time points and degradation of ribosomal RNA as marker for RNase L activation was assessed with a Fragment Analyzer. One representative picture and migration of 18S and 28S ribosomal RNA is displayed. The RNA Quality Number (RQN) is indicated. (b) The integrity of rRNA from MHV-A59 and MHV_H277A infected (MOI = 1) L929 cells, with or without IFN-I pre-treatment (12.5 U of IFN-I 16h prior to infection). Analysis was performed as in panel (a) and one representative image out of five is displayed.

Fig 6. Involvement of RNase L and PKR in restricting replication of EndoU-deficient MHV.

(a) Replication kinetics of MHV-A59 and MHV_H277A following infection (MOI = 1) of bone marrow-derived RNase L^-/- macrophages. Mean and SEM are shown. The 95% confidence bands are highlighted in grey. Data represent four independent experiments, each performed in two to three replicas. The difference in peak levels of viral titers (MHV-A59: 5.5, MHV_H277A: 3.5) was statistically significant (***, p<0.001). (b) Intracellular staining of p-eif2α (left panel) and puromycin (right panel) and FACS analysis of MHV-A59 and MHV_H277A infected (MOI = 1) C57BL/6 macrophages. One representative histogram out of three is shown. Phosphorylation of eif2α was determined using an antibody directed against p-eif2α. Cells without virus infection (mock) were used as controls (left panel). To label active translation (right panel), puromycin was added to the cells 15min prior to harvesting. Cells without puromycin-treatment (no puromycin) as well as cells without virus infection (mock) were used as controls. (c, d) Replication kinetics of MHV-A59 and MHV_H277A following infection (MOI = 1) of bone marrow-derived PKR^-/- (c) and RNase L^-/-/PKR^-/- (d) macrophages. Mean and SEM are shown. The 95% confidence bands are highlighted in grey. Data in (c) represent three independent experiments, each performed in two to three replicas. The difference in peak levels of viral titers (MHV-A59: 4.6, MHV_H277A: 2.7) was statistically significant (**, p = 0.004). Data in (d) represent three independent experiments, each performed in two to three replicas. The difference in peak levels of viral titers (MHV-A59: 6.1, MHV_H277A: 5.4) was statistically significant (*, p = 0.036). (e) Comparison of differences in peak titers calculated by using the non-linear regression model. Mean and 95% confidence intervals of calculated peak titers of MHV-A59 and MHV_H277A following infection (MOI = 1) of bone marrow-derived C57BL/6 (data correspond to Fig 2B), IFNAR^-/- (data correspond to Fig 4A) and RNase L^-/-/PKR^-/- (data correspond to Fig 6C) macrophages are displayed. Statistical analysis was performed to compare differences of calculated MHV-A59 and MHV_H277A peak titers between C57BL/6 and IFNAR^-/- (**, p = 0.008), between C57BL/6 and RNase L^-/-/PKR^-/- (**, p = 0.004) and between IFNAR^-/- and RNase L^-/-/PKR^-/- (p = 0.612; ns) macrophages following MHV-A59 and MHV_H277A infection.

Fig 7. Infection with EndoU-deficient MHV results in increased cytosolic dsRNA.

(a-b) Intracellular staining of dsRNA and FACS analysis of MHV-A59 and MHV_H277A infected (MOI = 1) C57BL/6 (a) and IFNAR^-/- (b) macrophages at 4, 6, 9 and 12 h.p.i.. One representative histogram out of two (a) and three (b) is shown for each time point. Cells without virus infection (mock) were used as controls. (c-d) The left panels show cells that were co-stained for MHV-nsp2/3 to control for MHV-A59 and MHV_H277A infection. The right panels display the median fluorescent intensity (MFI) of dsRNA peaks detected in (a-b). The left panels show data from two (c) and three (d) independent experiments. Cells without virus infection (mock) were used as controls. Mean and SEM are depicted. The 95% confidence band is highlighted in grey. Statistically significant comparisons are displayed (**, p< 0.01).

Fig 8. Coronavirus EndoU-mediated innate immune evasion.

Following coronavirus infection, the EndoU activity residing in the coronavirus replication complex prevents simultaneous activation of dsRNA sensors Mda5, OAS, and PKR. This strategy allows coronaviruses to efficiently evade antiviral innate host responses such as induction of IFN-I expression, RNase L-mediated RNA degradation, and inhibition of host cell translation.