CNBP restricts SARS-CoV2 by regulating IFN and disrupting RNA-protein condensates

Summary: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) evades antiviral immunity through the expression of viral proteins that block detection, signaling, interferon (IFN) induction, and IFN-stimulated gene (ISG) expression1, 2. Weak induction of type I IFNs is associated with a hyperinflammatory response in patients that develop severe COVID-193, 4, 5. Here we uncover a role for cellular nucleic acid-binding protein (CNBP) in restricting SARS-CoV-2. Typically, CNBP resides in the cytosol and, in response to RNA sensing pathways, undergoes phosphorylation, nuclear translocation, and IFNβ enhancer DNA binding to turn on IFNβ gene transcription. In SARS-CoV-2-infected cells CNBP coordinates IFNβ gene transcription. In addition, CNBP binds SARS-CoV-2 viral RNA directly. CNBP competes with the nucleocapsid (N) protein and prevents viral RNA and nucleocapsid protein from undergoing liquid-liquid phase separation (LLPS) forming condensates critical for viral replication. Consequently, cells and animals lacking CNBP have higher viral loads and CNBP-deficient mice succumb rapidly to infection. Altogether, these findings identify CNBP as a key antiviral factor for SARS-CoV-2, functioning both as a regulator of antiviral IFN gene expression and a cell intrinsic restriction factor that disrupts LLPS to limit viral replication and spread.

showed that, although SARS-CoV-2 infection leads to nuclear translocation of CNBP and reduced 53 induction of type I IFNs, CNBP also binds SARS-CoV-2 viral RNA directly, interfering with a key 54 step in the viral life cycle-blocking viral replication. In both cells and animals this leads to a 55 reduction in viral loads with a profound influence on susceptibility to infection.

58
Given our previous studies linking CNBP to antiviral immunity to other RNA viruses, we examined 59 its role in controlling SARS-CoV-2 infection. A549-ACE2 expressing cells which are permissive 60 to SARS-CoV-2 infection were transfected with CNBP or a vector control. We monitored the 61 accumulation of double-stranded RNA using J2 antibody staining by immunofluorescence as a 62 readout of virus infection and found the levels of J2 staining were reduced in cells overexpressing 63 CNBP (Fig. 1A). Cells expressing CNBP also had reduced levels of viral N and NSP14 RNA and 64 lower viral titers as measured by plaque assay relative to vector control cells ( Fig. 1B-D). We also 65 generated CNBP-deficient A549-ACE2 cells and after infection the levels of SARS-CoV-2 protein 66 assessed using anti-NP antibodies was also higher in CNBP-deficient cells (Fig. 1E). Similarly, 67 these cells had higher levels of J2 staining, N and NSP14 RNA levels, and had increased viral 68 titers relative to wild-type (WT) cells ( Fig. 1F-I). We observed similar effects with HCoV-OC43 69 infection, a related betacoronavirus (Extended Data Fig. 1A-D). Together, these data indicate that 70 CNBP plays a role in limiting the replication of SARS-CoV-2 and related coronaviruses.

73
Infection of A549-ACE2 cells with SARS-CoV-2 leads to a delayed IFNb response that is weak 74 relative to that seen with either influenza or Sendai viruses ( Fig. 2A). Treating SARS-CoV-2-75 infected cells with recombinant IFNa led to a marked decrease in viral RNA levels, indicating that 76 SARS-CoV-2 is sensitive to type I IFN treatment (Extended Data Fig. 2A-B). The levels of IFNb,

79
Endogenous CNBP is predominantly localized in the cytoplasm at steady state and

107
We next wanted to understand how CNBP curbs SARS-CoV-2 infection through IFN-independent 108 mechanisms. Two independent groups reported an unbiased analysis of host proteins that bind 109 to SARS-CoV-2 viral RNA. CNBP was the top SARS-CoV-2 genomic RNA-host binding protein 110 identified in these studies 19,20 . We therefore considered the possibility that CNBP bound viral RNA 111 directly. We confirmed that CNBP directly binds SARS-CoV-2 viral RNA by performing RNA 112 immunoprecipitation (RIP) followed by qPCR to quantify viral RNA levels (N and NSP14 RNAs).

113
SARS-CoV-2 viral RNA was enriched in the CNBP pulldowns (Fig. 3A). CNBP could also bind 114 RNA from HCoV-OC43 but not respiratory syncytial virus (Extended Data Fig. 3A and B). We next 115 mapped the region(s) of SARS-CoV-2 genomic RNA that was bound by CNBP. We generated 116 biotin-labeled RNAs corresponding to the 5′ UTR, 3′ UTR and three internal regions by in vitro 117 transcription (IVT) and used these in pulldown experiments. CNBP was enriched in the 118 streptavidin pulldowns using both the 5′ UTR and 3′ UTR RNA fragments but not by the RNA 119 fragments corresponding to internal regions of the genomic RNA (Fig. 3B). We also performed 120 the anti-CNBP RIP qPCR experiments in infected cells and showed that endogenous CNBP 121 binding to SARS-CoV-2 genomic RNA was reduced by incubating these pulldown reactions with

122
IVT RNAs corresponding to the 5′ UTR and 3′ UTR but not by IVT RNAs from other regions of the 123 genomic RNA (Fig. 3C).

124
The SARS-CoV-2 nucleocapsid protein is an RNA-binding protein that plays a critical role 125 in viral genome packaging and virion assembly. We speculated that CNBP might compete with 126 the N protein for viral RNA. We confirmed viral RNA binding to the N protein by RIP-qPCR. Anti-127 NP pulldowns demonstrated that NP bound viral RNA in infected cells and NP binding to RNA 128 was elevated in cells lacking CNBP (Fig. 3D). Further, overexpression of CNBP or the CNBP 129 T173/177A mutant blocked the binding of the N protein to viral RNA in a dose-dependent manner 130 ( Fig. 3E). We could also detect N protein associated with CNBP during SARS-CoV-2 infection; however, the interaction between CNBP and SARS-CoV-2 N was sensitive to RNase digestion, 132 suggesting that CNBP and SARS-CoV-2 N form a complex in the presence of viral RNA (Fig. 3F).

133
Recently, several independent groups have reported that NP can undergo liquid-liquid 134 phase separation (LLPS) in the presence of viral genomic RNA, and the formation of these RNA-

293
The RNA was extracted using TRIzol reagent before real-time PCR analysis for SARS-CoV-2 or 294 OC43 RNA.

296
Immunofluorescence 297 Cells were fixed using 4% PFA for 30 min. After two PBS washes, cells were permeabilized with 298 0.2% Triton X-100/PBS before incubation with primary antibodies for 2 h at room temperature.

299
Cells were washed in PBS, followed by incubation with secondary antibodies. Nuclei were stained 300 with DAPI.

302
In vitro transcription RNA assay.

303
Full RNA genome of SARS-CoV-2 was purified from supernatant of Vero E6 cells infected with 304 SARS-CoV-2 by TRIzol (Thermo Fisher), 1 µg of RNA was reverse transcribed using the iScript 305 cDNA synthesis kit (Bio-Rad). cDNA of the RNA genome of SARS-CoV-2 was used and amplified 306 by PCR through primers with the T7 promoter sequence in the 5′ end for PCR to prepare 307 templates of the in vitro transcription of the 5′ UTR, 3′ UTR and three other RNA fragments. The 308 purified PCR products were used for genomic RNA fragment synthesis using a HiScribe T7 high 309 yield RNA synthesis kit (NEB) according to the manufacturer's instructions. The synthesized 310 genomic RNA fragments were purified and labeled with biotin using the Label IT Biotin Labeling 311 Kit (Mirus) for RNA pull-down assay and RIP assay with RNA competition. The sequences of 312 primers with the T7 promoter sequence used in this study are listed in Table S1.