key: cord-0685875-sbr5ug3u authors: Li, Taiwei; Kenney, Adam D.; Liu, Helu; Fiches, Guillaume N.; Zhou, Dawei; Biswas, Ayan; Que, Jianwen; Santoso, Netty; Yount, Jacob S.; Zhu, Jian title: SARS-CoV-2 Nsp14 activates NF-κB signaling and induces IL-8 upregulation date: 2021-05-26 journal: bioRxiv DOI: 10.1101/2021.05.26.445787 sha: 59009ea871b5efc7a35883645cd5cd93f4dc982d doc_id: 685875 cord_uid: sbr5ug3u Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection leads to NF-κB activation and induction of pro-inflammatory cytokines, though the underlying mechanism for this activation is not fully understood. Our results reveal that the SARS-CoV-2 Nsp14 protein contributes to the viral activation of NF-κB signaling. Nsp14 caused the nuclear translocation of NF-κB p65. Nsp14 induced the upregulation of IL-6 and IL-8, which also occurred in SARS-CoV-2 infected cells. IL-8 upregulation was further confirmed in lung tissue samples from COVID-19 patients. A previous proteomic screen identified the putative interaction of Nsp14 with host Inosine-5’-monophosphate dehydrogenase 2 (IMPDH2) protein, which is known to regulate NF-κB signaling. We confirmed the Nsp14-IMPDH2 protein interaction and found that IMPDH2 knockdown or chemical inhibition using ribavirin (RIB) and mycophenolic acid (MPA) abolishes Nsp14-mediated NF-κB activation and cytokine induction. Furthermore, IMDPH2 inhibitors (RIB, MPA) efficiently blocked SARS-CoV-2 infection, indicating that IMDPH2, and possibly NF-κB signaling, is beneficial to viral replication. Overall, our results identify a novel role of SARS-CoV-2 Nsp14 in causing the activation of NF-κB. SARS-CoV-2 is a beta-coronavirus that causes the current, severe COVID-19 pandemic globally. The 40 viral genome of SARS-CoV-2 is a ~30 kb polycistronic, positive-strand RNA that encodes multiple 41 structural and nonstructural proteins (1, 2). SARS-CoV-2 nonstructural proteins (Nsp1-16) play 6 with TNF-α (Fig 3C) . However, overexpression of IMPDH2 had no significant effect on NF-κB activation 115 by Nsp14 in HEK293T cells with or without TNF-α (Fig S4) . Since IMPDH2 is required for IL-8 induction by Nsp14, we expected that its inhibition would reduce 118 Nsp14-mediated NF-κB activation and IL-8 induction. We tested two reported IMPDH2 inhibitors, 119 ribavirin (RIB) and mycophenolic acid (MPA) (23, 31) . RIB is a synthetic nucleoside that occupies the 120 IMPDH2 catalytic site to inhibit IMP conversion to xanthosine 5'-phosphate (XMP) during the guanine 121 nucleotide (GTP) biosynthesis (31) (32) (33) . MPA shares similar features with the IMPDH2 cofactor, 122 nicotinamide adenine dinucleotide (NAD + ). MPA stacks and traps the XMP intermediate at the catalytic 123 site to inhibit IMPDH2 enzyme activity (31, 34) . We confirmed that NF-κB activation by Nsp14 124 significantly decreases in HEK293T cells treated with RIB (Fig 4A) or MPA (Fig 4B) at multiple doses 125 in the absence or presence of TNF-α using the NF-κB luciferase reporter assays. Likewise, treatment of 126 HEK293T cells with RIB (Fig 4C) or MPA (Fig 4D) also caused the reduction of IL-8 induction by Nsp14. 127 We next tested whether IMPDH2 inhibitors (RIB, MPA) also repress SARS-CoV-2 infection in vitro, 128 considering that virus-mediated NF-κB activation would likely benefit its replication (35) (36) (37) (38) . Indeed, we 129 showed that the infection rate of SARS-CoV-2 decreases in both A549-ACE2 and HEK293T-ACE2 cells 130 treated with RIB or MPA through quantification of cells expressing N protein by immunofluorescence 131 staining assays (Fig 4E-F, S5A-B) or sgRNA level by RT-qPCR (Fig 4G, S5C) . Consistently, we also 132 identified that treatment of RIB or MPA leads to a significant reduction of IL-8 expression (Fig 4H) . translocation of IRF3 to facilitate viral invasion of the host's antiviral immune response (3, 4) . Our results 138 showed that Nsp14, which is expressed at the early stage of primary infection (7), also affects other cell 139 signaling pathways, such as NF-κB signaling (Fig 1) , likely to support viral replication. Activation of NF-140 κB may further trigger the production of downstream pro-inflammatory cytokines to initiate the cytokine 141 storm and contribute to ARDS. In this study, we identified that Nsp14 increases nuclear translocation of 142 p65 and induces expression of NF-κB's downstream cytokines, such as IL-6 and IL-8, which have also 143 been detected in lung tissues of 29) and animal models of SARS-CoV-2 infection 144 (7). These cytokines are reported to play a critical role in regulating the recruitment and infiltration of 145 immune cells (macrophages, neutrophils) during viral infection (39, 40) . Infiltrating immune cells may 146 further escalate inflammatory responses leading to lung damage. Indeed, we showed that IL-8 expression 147 is much higher in lung tissue samples of COVID-19 patients than in uninfected controls (Fig 2E, F) . 148 Another key finding is that IMPDH2 is a host mediator of Nsp14 involved in NF-κB activation, 149 verified by both genetic knockdown (Fig 3) and chemical inhibition (Fig 4) . We confirmed the protein 150 interaction of Nsp14 with IMPDH2, which was initially reported in earlier proteomic studies (23, 29 promote nuclear translocation and phosphorylation of p65 (24). In addition, we also noticed that Nsp14 162 partially localizes in the nuclei of cells (Fig 1C, D) , similar to findings from other groups (51, 52) . Thus, 163 Nsp14 may associate with and modify the host cellular RNAs via its exonuclease and methyltransferase 164 activities. Nsp14 may also affect the transcriptional activity of nuclear p65 and the expression of its gene 165 targets. Future studies will be needed for further understanding how Nsp14 and IMPDH2 cooperate to 166 activate NF-κB. Our study has the translational significance since we showed that IMPDH2 inhibitors, RIB and MPA, 168 effectively reduce viral replication of SARS-CoV-2 and expression of NF-κB's downstream cytokines 169 (IL-6 and IL-8) induced by SARS-CoV-2 (Fig 4E-H) . It has been reported that IL-8 increases the 170 replication of human immunodeficiency virus-1 (HIV-1), HCV, and cytomegalovirus (CMV) (53-56). Lopinavir-Ritonavir therapy is currently in clinical trials for treating SARS-CoV-2 infection (58), which 177 has been shown to significantly alleviate the COVID-19 symptoms and suppress IL-6 levels in serum. In 178 another preclinical study, MPA was reported to inhibit SARS-CoV-2 replication (59) and viral entry (60). Our study delineated a potentially new mode of action (MOA) for these IMPDH2 inhibitors, which may 180 disrupt the Nsp14-IMPDH2 axis that plays a crucial role in regulating activation of NF-κB signaling and 181 induction of its downstream cytokines (Fig 5) . Millipore) containing protease inhibitor cocktail (Cat. # A32965, Thermo Scientific) on ice, followed by 227 brief sonication to prepare cell lysate. The BCA assay kit (Cat. #23225, Thermo Scientific) was used to 228 quantify the total protein amount in cell lysate, which was boiled in the SDS loading buffer with 5% β-229 11 mercaptoethanol (Cat. #60-24-2, Acros Organics). The denatured protein samples were separated by 230 Novex™ WedgeWell™ 4-20% SDS-PAGE Tris-Glycine gel and transferred to PVDF membrane (iBlot™ 231 2 Transfer Stacks, Invitrogen) using iBlot 2 Dry Blotting System (Cat. # IB21001, Thermo Scientific). The membranes were blocked by 5% milk in PBST and probed by the specific primary antibodies at 4°C Anti-GAPDH and anti-histone H3 immunoblotting were used as internal controls to determine the 251 cytoplasmic and nuclear fractions. Paraffin-embedded lung tissue blocks were baked on the hotplate at 75 °C for 20 min and 300 deparaffinized in xylene. The slides were rehydrated from 100%, 90%, to 70% alcohol, and then to PBS. 301 We performed the antigen unmasking using the retriever (Cat. # 62700-10, Electron Microscopy Sciences) 302 with R-Buffer A pH 6.0 (Cat. # 62706-10, Electron Microscopy Sciences) for 2 h to complete the cycle Sheng-Ce Tao (Shanghai Jiao Tong University) for providing Nsp14 cloning plasmid. We also thank Dr. Cells were harvested for RNA extraction, and N protein sgRNA was analyzed and normalized to the mock treatment (C) . The results were calculated from 3 technical repeats and presented as mean +/-standard deviation (SD). (** p <0.01; **** p <0.001 by one-way ANOVA and Tukey's multiple comparison test) A) HEK293T cells transfected with V5-FLAG-Nsp14 or empty vector along with NF-κB-driven firefly luciferase and TK-driven renilla luciferase reporter vectors were treated with ribavirin (RIB) at the basal or TNFα-stimulated condition. Luciferase activity (firefly/renilla) in these cells was measured and normalized to that of un-treated, empty vector-transfected cells. (B) Mycophenolic acid (MPA) was tested similarly as in (A) Tukey's multiple comparison test). (C) HEK293T cells transfected with V5-FLAG-Nsp14 or empty vector were treated with RIB at the basal or TNF-α-stimulated condition Results were calculated from 3 independent experiments and presented as mean +/-standard deviation (SD). (* p <0.05 by unpaired Student's t-test). (E-H) A549-ACE2 cells were treated with RIB (500 µM), MPA (100 µM), or mock, and infected with SARS-Cov-2 viruses for 24 h. The SARS-CoV-2 infection was detected by intracellular staining of SARS-CoV-2 N protein (E) 001 by two-way ANOVA and Tukey's multiple comparison test) A) HEK293T cells were transiently transfected with the vector expressing V5-FLAG-Nsp10, 14, 16, or empty vector. Protein expression was analyzed by protein immunoblotting. (B) HEK293T cells transfected with the vector expressing V5-FLAG-Nsp10, 14, 16, or empty vector along with ISRE-driven firefly luciferase and TK-driven renilla luciferase reporter vectors were un or empty vector along with NF-κB-driven firefly luciferase and TK-driven renilla luciferase reporter vectors were untreated or treated with TNF-α. Luciferase activity (firefly/renilla) in these cells was measured and normalized to the empty vector. Results were calculated from 3 technical repeats and presented as mean +/-standard deviation (SD) 0001 by unpaired Student's t-test) .