key: cord-0909148-igwdmrzp
authors: Sampaio, Natalia G.; Chauveau, Lise; Hertzog, Jonny; Bridgeman, Anne; Fowler, Gerissa; Moonen, Jurgen P.; Dupont, Maeva; Russell, Rebecca A.; Noerenberg, Marko; Rehwinkel, Jan
title: The RNA sensor MDA5 detects SARS-CoV-2 infection
date: 2021-03-27
journal: bioRxiv
DOI: 10.1101/2021.03.26.437180
sha: 1ee6568fec7ff8999924e7b3b21f2ae4ec2acb7b
doc_id: 909148
cord_uid: igwdmrzp

Human cells respond to infection by SARS-CoV-2, the virus that causes COVID-19, by producing cytokines including type I and III interferons (IFNs) and proinflammatory factors such as IL6 and TNF. IFNs can limit SARS-CoV-2 replication but cytokine imbalance contributes to severe COVID-19. We studied how cells detect SARS-CoV-2 infection. We report that the cytosolic RNA sensor MDA5 was required for type I and III IFN induction in the lung cancer cell line Calu-3 upon SARS-CoV-2 infection. Type I and III IFN induction further required MAVS and IRF3. In contrast, induction of IL6 and TNF was independent of the MDA5-MAVS-IRF3 axis in this setting. We further found that SARS-CoV-2 infection inhibited the ability of cells to respond to IFNs. In sum, we identified MDA5 as a cellular sensor for SARS-CoV-2 infection that induced type I and III IFNs.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) emerged at the end of 2019 31 and is causing an ongoing global pandemic. As of March 26 th , 2021, it has infected 32 124,535,520 and killed 2,738,876 patients worldwide (https://covid19.who.int/). SARS-CoV-2 33 causes Coronavirus Disease 2019 (COVID-19), a respiratory disease that in some patients 34 results in severe pneumonia and acute respiratory distress syndrome leading to death. Severe 35 disease is linked to exacerbated inflammation with increased production of pro-inflammatory 36 cytokines such as TNF and IL-6, and delayed type I and type III interferon (IFN) responses 1-4 . 37

Although serum levels of type I and III IFNs are low or undetectable in many patients, the 38 increased expression of genes known to be induced by IFN (called interferon stimulated 39 genes; ISGs) suggests that production of IFN occurs. The importance of the IFN system in 40 controlling disease was confirmed by the discovery of an association between severe disease 41 and inborn errors in IFN immunity as well as autoantibodies against type I IFNs 2,5 . Moreover, 42 recent studies show that intranasal, but not intravenous, administration of type I IFN Its genome is a ~30kb positive-sense single-stranded RNA that shares ~80% sequence identity 54 with SARS-CoV and ~50% sequence identity with MERS-CoV 11 . Nucleic acid sensors mediate 55 the early detection and host response to virus infections. These sensors recognise either viral 56 nucleic acids or 'unusual' cellular nucleic acids present upon infection 12 . Cytosolic nucleic acid 57 sensors from the RIG-I-Like Receptor (RLR) family have been identified as important PRRs that 58 sense coronaviruses 9,13 . The two signalling receptors in this family are retinoic acid-inducible 59 gene I (RIG-I) and melanoma differentiation-associated protein 5 (MDA5), which detect RNAs 60 with specific structures such as 5'-triphosphate or 5'-diphosphate ends 14,15 . Once activated, 6 Calu-3 cells with lentiviruses expressing Cas9, an sgRNA and the puromycin resistance gene. 111

Using puromycin selection, we generated polyclonal cell lines. We included sgRNAs targeting 112 MAVS, MDA5 and RIG-I. To determine if the cytosolic DNA sensing pathway was activated 113 during SARS-CoV-2 infection, we also targeted STING. IRF3 is activated downstream of both 114 STING and MAVS; therefore, we included a IRF3 sgRNA. Protein levels of MDA5, RIG-I, STING 115 and IRF3 were notably reduced by this approach, whereas the targeting of MAVS was less 116 efficient ( Fig 3A) . Next, we infected these cells with SARS-CoV-2 (MOI=0.1) and analysed their 117 response by RT-qPCR after 48 hours. Cells lacking MDA5 or IRF3 showed minimal induction of 118 type I IFN mRNAs and ISGs in response to SARS-CoV-2 infection (Fig 3B) . This effect was 119 uncoupled from the pro-inflammatory response, as lack of MDA5 and IRF3 did not affect the 120 expression of TNF, and minimally affected expression of IL6, in response to infection ( Fig 3B) . 121

Lack of RIG-I or STING had no effect on the cellular response to SARS-CoV-2, suggesting that 122 neither RIG-I nor the cGAS-STING pathway were involved in sensing SARS-CoV-2 in this 123 setting. It likely that partial depletion of MAVS ( Fig 3A) explains why cells transduced with 124 MAVS sgRNA responded comparably to control cells. 125

To confirm these observations with a different readout, we used intracellular staining and 126 flow cytometry to measure protein levels of MxA, which is encoded by the ISG MX1, and SARS-127

CoV-2 N ( Fig 3C) . We analysed cells infected for 48 hours with SARS-CoV-2 (MOI=0.1). In this 128 setting, ~40-60% of cells stained positive for N and there was little difference in infection 129 levels between the knockout cells ( Fig 3D) . Consistent with our RT-qPCR data, targeting of 130 MDA5 and IRF3 largely prevented upregulation of MxA in response to SARS-CoV-2 infection 131 infected N+ cells (Fig 3F) . This shows that SARS-CoV-2 infection inhibited the autocrine 136 response of the same cell to released IFNs, whereas uninfected bystander cells were able to 137 respond to IFNs more strongly, implicating viral antagonism of IFNAR and/or IFNLR signalling. 138

In this study, consistent with several recent reports 19-21 , we identified Calu-3 cells as a suitable We further show, using both shRNA-mediated knockdown and CRISPR/Cas9 genetic ablation, 161 that the type I and III IFN response to SARS-CoV-2 was dependent on the RNA sensor MDA5, 162 its downstream adapter MAVS and the transcription factor IRF3. In Calu-3 cells, RIG-I was 163 largely dispensable for the antiviral cytokine response to infection. While our study was in 164 preparation, other manuscripts reported MDA5 as the cellular sensor that recognises SARS-165

CoV-2 infection 19,20 . These reports included both siRNA-mediated knockdown and genetic 166 ablation in Calu-3 cells. In contrast to these two studies and to our work, another study found 167 that both MDA5 and RIG-I sense SARS-CoV-2 infection in Calu-3 cells, and that upregulation 168 of the pro-inflammatory cytokine IL-6 was uncoupled from MDA5, but dependent on RIG-I 169 and MAVS 21 . This study utilised an siRNA knockdown approach; whether this or other technical differences explain the disparity between these findings and what we and others 171 report remains to be determined. Another research group reported that total RNA extracted 172 from SARS-CoV-2-infected Vero E6 cells activated MDA5, but not RIG-I, after transfection into 173 human lung fibroblasts 32 . Interestingly, a similar experimental approach in a different study 174 been proposed that the STING pathway mediates NF-κB activation and TNF induction in 190 response to SARS-CoV-2 37 . In our work, loss of STING reduced but did not abolish the TNF 191 response to SARS-CoV-2 ( Fig 3B) . However, it is important to note that some residual STING 192 protein was still present in the targeted cells (Fig 3A) , so it is possible that cytosolic DNA Cells were lysed with RIPA buffer (10 mM TRIS-HCl pH 8, 140 mM NaCl, 1% Triton-X100, 0.1% 288 SDS, 0.1% sodium deoxycholate, 1 mM EDTA, 1 mM EGTA) and protein was quantified by BCA 289 assay (Pierce). NuPAGE LDS sample loading buffer (Life Technologies) and 10% 2-290 mercaptoethanol were added to samples, which were then denatured by incubation at 95°C for 5 minutes. Samples were resolved by electrophoresis on 4-12% Bis-Tris gels with MOPS 292

Running Buffer (Life Technologies NuPAGE system) and transferred to nitrocellulose 293 membrane by electrophoresis at 100V for 1 hour in cold transfer buffer (Life Technologies) 294 with 20% methanol. Membranes were blocked with 0.05% NP-40 (IGEPAL) in Tris-buffered 295 saline (TBS-N; 50 mM NaCl, 50 mM Tris-HCl, pH 7.6) containing 5% non-fat milk (5% milk TBS-296 N) for 1 hour and probed with primary and HRP-conjugated secondary antibodies (GE 297

Healthcare, 1:3000) diluted in 5% milk TBS-N for 1 hour at room temperature or overnight at 298 Fisher Scientific) and data were analysed using FlowJo software (BD (B) Cells from (A) were infected and analysed as described in Figure 2C . 537

Dengue virus NS2B protein targets cGAS for degradation and prevents 379 mitochondrial DNA sensing during infection

MDA5 Governs the Innate Immune Response to SARS-CoV-2 in Lung 382

SARS-CoV-2 triggers an MDA-5-dependent interferon response 384 which is unable to control replication in lung epithelial cells

SARS-CoV-2 sensing by RIG-I and MDA5 links epithelial infection to 387 macrophage inflammation

Enhanced isolation of SARS-CoV-2 by TMPRSS2-expressing cells

Isolation and characterization of SARS-CoV-2 from the first US 392 COVID-19 patient

Angiotensin-converting enzyme 2 (ACE2), but not ACE, is 394 preferentially localized to the apical surface of polarized kidney cells

SARS-CoV-2 and the safety margins of cell-based biological medicinal 397 products

Characterization of spike glycoprotein of SARS-CoV-2 on virus entry and 399 its immune cross-reactivity with SARS-CoV

A Novel Coronavirus from Patients with Pneumonia in China

Severe Acute Respiratory Syndrome Coronavirus 2 from Patient with 404 Coronavirus Disease, United States

Morphological Cell Profiling of SARS-CoV-2 Infection Identifies Drug 407 Repurposing Candidates for COVID-19

A pneumonia outbreak associated with a new coronavirus of probable 410 bat origin

Type-I interferon signatures in SARS-CoV-2 infected Huh7 cells

ISG15-dependent Activation of the RNA Sensor MDA5 and its Antagonism 414 by the SARS-CoV-2 papain-like protease

SARS-CoV-2 ORF9b inhibits RIG-I-MAVS antiviral signaling by interrupting 417 K63-linked ubiquitination of NEMO

Identification and characterization of MAVS, 420 a mitochondrial antiviral signaling protein that activates NF-kappaB and IRF 3

Toll-like receptor signaling pathways

SARS-CoV-2 infection induces a pro-inflammatory cytokine 427 response through cGAS-STING and NF-κB. bioRxiv

LGP2 binds to PACT to regulate RIG-I-and MDA5-mediated 430 antiviral responses

NOD1 Promotes Antiviral Signaling by Binding Viral RNA and 432 Regulating the Interaction of MDA5 and MAVS

Complex Immune Dysregulation in COVID-19

Patients with Severe Respiratory Failure

Improved vectors and genome-wide libraries for 474 CRISPR screening

Infection with a Brazilian isolate of Zika virus generates RIG-I 476 stimulatory RNA and the viral NS5 protein blocks type I IFN induction and signaling

Breadth and function of antibody response to acute SARS-CoV-2 479 infection in humans. bioRxiv

C-F) Cells from (A) were infected as in Figure 2C, stained for live cells

Live cells were assessed for SARS-CoV-2 N protein 539 expression (D) and MxA induction, shown as mean fluorescence intensity (MFI; E). (F) Cells in 540 the SARS-CoV-2-infected samples were further subdivided into SARS-CoV-2 N positive

SARS-CoV-2 N negative (N-) cells, and MxA MFI was determined within these 542 subpopulations

Data in (A, C) are representative of two independent biological repeats. Data in (B and D-F) 544

The authors thank William James for providing access to the BSL3 facility. The authors thank 484 Arthur Huang, Pramila Rijal, Lisa Schimanski, Tiong Kit Tan and Alain Townsend for providing  485