key: cord-0766215-d2yvf666
authors: Finkel, Yaara; Gluck, Avi; Winkler, Roni; Nachshon, Aharon; Mizrahi, Orel; Lubelsky, Yoav; Zuckerman, Binyamin; Slobodin, Boris; Yahalom-Ronen, Yfat; Tamir, Hadas; Ulitsky, Igor; Israely, Tomer; Paran, Nir; Schwartz, Michal; Stern-Ginossar, Noam
title: SARS-CoV-2 utilizes a multipronged strategy to suppress host protein synthesis
date: 2020-11-25
journal: bioRxiv
DOI: 10.1101/2020.11.25.398578
sha: a986d09aae68300370d5b3c6a4bbc6be864177d1
doc_id: 766215
cord_uid: d2yvf666

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the cause of the ongoing coronavirus disease 19 (COVID-19) pandemic. Despite the urgent need, we still do not fully understand the molecular basis of SARS-CoV-2 pathogenesis and its ability to antagonize innate immune responses. Here, we use RNA-sequencing and ribosome profiling along SARS-CoV-2 infection and comprehensively define the mechanisms that are utilized by SARS-CoV-2 to shutoff cellular protein synthesis. We show SARS-CoV-2 infection leads to a global reduction in translation but that viral transcripts are not preferentially translated. Instead, we reveal that infection leads to accelerated degradation of cytosolic cellular mRNAs which facilitates viral takeover of the mRNA pool in infected cells. Moreover, we show that the translation of transcripts whose expression is induced in response to infection, including innate immune genes, is impaired, implying infection prevents newly transcribed cellular mRNAs from accessing the ribosomes. Overall, our results uncover the multipronged strategy employed by SARS-CoV-2 to commandeer the translation machinery and to suppress host defenses.

infection and comprehensively define the mechanisms that are utilized by SARS-CoV-2 to 23 shutoff cellular protein synthesis. We show SARS-CoV-2 infection leads to a global reduction in 24 translation but that viral transcripts are not preferentially translated. Instead, we reveal that 25 infection leads to accelerated degradation of cytosolic cellular mRNAs which facilitates viral 26 takeover of the mRNA pool in infected cells. Moreover, we show that the translation of 27 transcripts whose expression is induced in response to infection, including innate immune genes, 28 is impaired, implying infection prevents newly transcribed cellular mRNAs from accessing the 29 ribosomes. Overall, our results uncover the multipronged strategy employed by SARS-CoV-2 to 30 commandeer the translation machinery and to suppress host defenses. Binding of IFN to its cognate receptor in autocrine and paracrine manners leads to the 54 propagation of the signal and to the transcription and translation of hundreds of interferon 55 stimulated genes that act to hamper viral replication at various stages of the viral life cycle 12 . In 56 the case of SARS-CoV-2, interferon response seems to play a critical role in pathogenesis [13] [14] [15] [16] [17] . 57

In addition, the extent to which SARS-CoV-2 suppresses the IFN response is a key characteristic 58 that distinguishes it from other respiratory viruses 18, 19 . 59

Viruses utilize various strategies to cause shutoff of host mRNA translation 10 , including 60 hampering mRNA processing steps and export, inducing degradation of mRNAs and inhibiting translation. Coronaviruses (CoVs) are known to cause host shutoff 10,20 and several strategies 62 have been proposed for how the beta CoVs may shut-off host protein synthesis and evade 63 immune detection. These include degradation of host mRNA in the nucleus or the cytosol and 64 inhibition of host translation 21, 22 . Nonetheless, the extent to which SARS-CoV-2 uses these or 65 other strategies remains unclear. 66 NSP1 is the best characterized and most prominent coronavirus host shutoff factor 23 . Several 67 recent studies showed SARS-CoV-2 NSP1 binds the 40s ribosome and inhibits translation [24] [25] [26] [27] [28] . 68

In addition, other SARS-CoV-2 proteins were shown to interfere with cellular gene expression. 69

For example, one of the SARS-CoV-2 accessory proteins, ORF6, was suggested to disrupt 70 nucleocytoplasmic transport leading to inhibition of gene expression 29 Here we employ RNA-sequencing and ribosome profiling along SARS-CoV-2 infection to 75 explore the mechanisms which the virus utilizes to interfere with host protein synthesis. We 76 reveal SARS-CoV-2 uses a multi-faceted approach to shutoff cellular protein production. SARS-77

CoV-2 infection induces global translation inhibition but surprisingly the translation of viral 78 transcripts is not preferred over their cellular counterparts. Instead we reveal that infection leads 79 to accelerated cellular mRNAs degradation, likely conferred by NSP1. Viral transcripts are 80 refractory to these effects, an evasion potentiated by their 5'UTRs, enabling viral dominance 81 over the mRNA pool in infected cell. Finally, we show that the translation of transcripts whose 82 expression is induced in response to infection, including innate immune genes, is severely 83 due to the massive production of viral transcripts but at the same time the relative fraction of 146 cellular mRNA is reduced by approximately 2-fold (Figure1F). This suggests that during 147 infection there is both massive production of viral transcripts and a concomitant substantial 148 reduction in the levels of cellular transcripts. 149

Next, we quantitatively assessed the expression pattern of cellular genes along SARS-CoV-2 150 infection. We clustered the mRNA levels of genes that showed the most significant changes 151 along infection using partitioning clustering, allowing grouping of cellular transcripts into three 152 distinct classes based on similarities in temporal expression profiles in the RNA-seq. Overall, we 153 found that changes in ribosome footprints tracked the changes in transcript abundance ( Figure  154 1G), with some exceptions that will be discussed in detail below. This shows that part of the 155 reduction in cellular protein synthesis is driven by the reduction in cellular RNA levels. 156

Interestingly, although the levels of the majority of host transcripts were reduced during SARS-157

CoV-2 infection, we identified numerous transcripts that were significantly elevated ( Figure 1G,  158 cluster 3). We carried out pathway enrichment analysis for each of these three clusters. As 159 expected, the group of upregulated mRNAs (cluster 3) was significantly enriched with genes 160 related to immune response, including Toll receptor Signaling , chemokine and cytokine 161 signaling ( Figure 1H and Table S1 ) and these upregulated genes include IL6 and IL8 which play 162 a significant role in the pathogenesis of SARS-CoV-2 35 and several IFN stimulated genes like, 163 IFIT1, 2 and 3, IRF1, ISG15 and TNF alpha induced proteins. These measurements and 164 analyses reveal that the shutoff in host protein synthesis is driven by several mechanisms 165 including; general reduction in the translation capacity of infected cells and reduction in the 166 levels of most cellular mRNAs. 167 168

We next analyzed viral gene expression dynamics along SARS-CoV-2 infection. Viral ORFs are 170 translated from the 30kb genomic RNA or from a nested series of subgenomic RNAs that 171 contain a common 5′ leader fused to different segments from the 3′ end of the viral genome 172 ( Figure 2A and V'kovski et al., 2020). Viral transcripts and translation levels significantly increased from 3 to 5 hpi but the relative abundance of some viral transcripts and their translation 174 rates were reduced from 5 to 8 hpi ( Figure 2B and 2C). This relative reduction in translation rates 175 of some viral transcripts prompted us to assess the relative translation efficiency of viral genes 176 along infection. Since, as indicated above, the translation efficiency of viral genes compared to 177 their cellular counterparts is relatively reduced along infection ( Figure 1D Our results indicate that the levels of the majority of cellular RNA are reduced during SARS-214

CoV-2 infection and this reduction contributes to the shutoff of cellular protein synthesis. 215

Reduction in cellular RNA levels could be due to interference with RNA production or 216 accelerated RNA degradation. To explore the molecular mechanism, we analyzed if the 217 reduction in cellular transcripts is associated with their subcellular localization. We used 218 measurements of the subcellular localization of transcripts by cytoplasmic and nuclear 219 fractionation 41 to assess the importance of subcellular localization. The levels of transcripts that 220 mostly localize to the cytoplasm were more reduced in infected cells compared to transcripts that 221 are mostly nuclear ( Figure 3A ) and there was a clear correlation between subcellular localization 222 and the extent of reduction in transcript levels following SARS-CoV-2 infection ( Figure S4A ). 223

Furthermore, compared to transcripts encoded in the nuclear genome, mitochondrial encoded 224 transcripts were refractory to the effects of SARS-CoV-2 infection ( Figure 3B ). The specific 225 sensitivity of cytosolic transcripts implies these transcripts may be specifically targeted during 226 SARS-CoV-2 infection. In CoVs, the most prominent and well characterized cellular shutoff 227 protein is NSP1 23 . So far, studies on SARS-CoV-2 NSP1 demonstrated it restricts translation by 228 directly binding to the ribosome 40S subunit 25-28 , thereby globally inhibiting translation 229 initiation. For SARS-CoV, on top of this translation effect, NSP1 interactions with the 40S was 230 also shown to induce cleavage of translated cellular mRNAs, thereby accelerating their turnover 231 their translation we compiled a list of 14 long non-coding RNAs (lncRNAs) that localize to the 233 cytoplasm and are well expressed in our data but, as expected, poorly translated ( Figure S4B ). 234

Relatively to cellular mRNAs, cytoplasmic lncRNAs were less affected by SARS-CoV-2 235 infection ( Figure 3C ), indicating accelerated turnover of cellular transcripts in infected cells may 236 be related to their translation. Recently, ribosome profiling and RNA-seq were conducted on 237 cells transfected with NSP1 42 . Analysis of the RNA expression from this data revealed that 238 ectopic NSP1 expression leads to weaker but similar signatures to the ones we identified in 239 infected cells; stronger reduction of cytosolic transcripts compared to nuclear transcripts, 240 stronger sensitivity of nuclear encoded transcripts, and stronger reduction of translated mRNA 241 compared to cytosolic lncRNAs ( Figure S5A -C). 242

We further noticed SARS-CoV-2 infection leads to increased levels of intronic reads in many 243 cellular transcripts ( Figures 3D and 3E ) indicating SARS-CoV-2 may interfere with cellular 244 mRNA splicing, as was recently suggested 24 . However, massive degradation of mature cytosolic 245 mRNAs may also generate a relative increase in intronic reads. We therefore analyzed the ratio 246 of intronic and exonic reads relative to rRNA. Whereas relative to rRNA levels, exonic reads 247 showed drastic reduction along SARS-CoV-2 infection, the intronic reads levels showed a more 248 subtle change ( Figure 3F ). Furthermore, the increase in the ratio of intronic to exonic reads was 249 greater in genes whose expression was reduced along infection compared to genes whose 250 expression was induced ( Figure 3G and Figure S6A ), illustrating the relative increase in intronic 251 reads is mostly independent of newly transcribed RNAs. Finally, we also detected more intronic 252 reads in cells that exogenously expressed NSP1 42 ( Figure S6B ). Together these results indicate 253 that the increase in intronic reads compared to exonic reads during SARS-CoV-2 infection is 254 largely driven by accelerated degradation of mature cellular transcripts that leads to relative 255 reduction in exonic reads. It is likely that SARS-CoV-2 also directly regulates splicing 256 efficiency, as was recently proposed 24 , but this effect seems more subtle. Together these 257 findings indicate SARS-CoV-2 infection leads to accelerated degradation of cytosolic cellular 258 mRNAs and that this degradation significantly affects the production of cellular proteins.

An important aspect of host shutoff during infection is the ability of the virus to hamper the 261 translation of cellular transcripts while recruiting the ribosome to its own transcripts. Although it 262 has been suggested that SARS-CoV-2 mRNAs are refractory to the translation inhibition induced 263 by NSP1 24, 43 our measurements indicate that SARS-CoV-2 transcripts are not preferentially 264 translated in infected cells ( Figure 1D and 9 ), that RNA degradation plays a prominent role in 265 remodeling the mRNA pool in infected cells and that SARS-CoV-2 dominates the mRNA pool. 266

All SARS-CoV-2-encoded subgenomic RNAs contain a common 5′ leader sequence that is 267 added during negative-strand synthesis 44 . We therefore explored whether the 5'UTR sequence 268 protects viral mRNAs from NSP1 induced degradation. We fused the viral 5'UTR sequence or a 269 short control 5'UTR to the 5′ end of a GFP reporter ( Figure 4A ) and transfected these constructs 270 together with expression vectors encoding NSP1 or NSP2 (that was used as a control) into 293T 271 cells. We found that NSP1 expression suppresses the production of the control-GFP but not of 272 the 5'UTR-containing GFP ( Figure 4B and 4C). We extracted RNA from these cells and 273 observed that the NSP1 induced reduction in control-GFP expression was associated with 15-274 fold reduction in the RNA GFP levels whereas the levels of the GFP RNA with the SARS-CoV-275 2 5'UTR were not reduced, and were even slightly increased, by NSP1 expression ( Figure 4D ). 276

The plasmid we used also contains an mCherry reporter expressed from an independent 277 promoter. Reassuringly NSP1 also induces a reduction in both mCherry protein ( Figure S7A and 278 S7B) and RNA levels ( Figure S7C ) when compared to NSP2. These results indicate that the 5′ 279 UTR of viral RNAs provides them protection from NSP1 induced degradation and that this 280 protection contributes to the ability of the virus to dominate the mRNA pool in infected cells. 281

Our results so far exemplify how SARS-CoV-2 remodels the transcript pool in infected cells. To 283 quantitatively evaluate the role of translational control along SARS-CoV-2 infection, we 284 calculated relative translation efficiency (TE, ratio of footprints to mRNAs for a given gene) 285 along infection. We then centered on genes that showed the strongest change in their relative TE 286 along infection. We clustered these genes into four clusters, based on similarity in their temporal 287 TE profiles, which largely reflect genes whose relative TE is reduced along infection and genes 288 whose relative TE is increased. The mRNA and footprint temporal profiles of these genes reveal 289 a clear signature; the genes whose relative TE along infection was reduced were genes whose 290 mRNA increased during infection but did not show a corresponding increase in footprints 291 ( Figure 5A and Figure S8 ). These genes were enriched in immune response genes (FDR < 10 -4 ) 292 like IRF1, IL-6 and CXCL3. Comparing changes in mRNA and TE levels of cellular genes along 293 infection demonstrates that generally transcripts which are transcriptionally induced following 294 infection show a reduction in their relative TE and vice versa ( Figure 5B ). This inability of RNA 295 that is elevated in response to infection to reach the ribosomes, may explain why infected cells 296 fail to launch a robust IFN response 13,14 . These data demonstrate that the overall capacity of 297 infected cells to generate new proteins is severely impaired. Disruption of cellular protein production using these three components may represent a multi-312 pronged mechanism that synergistically acts to suppress the host antiviral response ( Figure 5C ). we do not know the underlying molecular mechanism, one appealing hypothesis is viral 352 inhibition of cellular mRNA nuclear export. Since SARS-CoV-2 replicates in the cytoplasm, 353

inhibiting the nuclear export of mRNAs can provide unique advantages as it will specifically 354 inhibit cellular mRNA translation and more explicitly it will lead to suppression of the host's 355 antiviral response genes which are transcriptionally induced and therefore fully depend on 356 efficient nuclear export. ORF6 was recently shown to co-purify with host mRNA export factors 357 48 and by over expression it disrupts nucleocytoplasmic export 29 , providing a candidate that may 358 explain the phenotype we observe in infection. Another possibility is that NSP16, which was 359 recently shown to inhibit cellular RNA splicing 24 , is driving this phenotype as interference with 360 splicing will prevent efficient nuclear export. Indeed, we show here that infection leads to 361 increased levels of intronic reads in many cellular transcripts. Although our analysis reveals 362 some of this signature is attributed to accelerated degradation of mature cellular transcripts it is 363 likely there is also a processing and splicing defect that leads to more intronic reads and 364 eventually aberrant export. 365

Overall, our study provides an in-depth picture of how SARS-CoV-2 efficiently interferes with 366 cellular gene expression, leading to shutdown of host protein production using a multipronged 367 strategy. 368 For RNA-seq, cells were washed with PBS and then harvested with Tri-Reagent (Sigma-467 Aldrich), total RNA was extracted, and poly-A selection was performed using Dynabeads 468 mRNA DIRECT Purification Kit (Invitrogen) mRNA sample was subjected to DNAseI 469 treatment and 3' dephosphorylation using FastAP Thermosensitive Alkaline Phosphatase 470 (Thermo Scientific) and T4 PNK (NEB) followed by 3' adaptor ligation using T4 ligase (NEB). 471

The ligated products used for reverse transcription with SSIII (Invitrogen) for first strand cDNA 472 synthesis. The cDNA products were 3' ligated with a second adaptor using T4 ligase and 473 amplified for 8 cycles in a PCR for final library products of 200-300bp. For Ribo-seq libraries, 474 cells were treated with 100µg/mL CHX for 1 minute. Cells were then placed on ice, washed 475 twice with PBS containing 100µg/mL CHX, scraped from 10cm plates, pelleted and lysed with lysis buffer (1% triton in 20mM Tris 7.5, 150mM NaCl, 5mM MgCl2, 1mM dithiothreitol 477 supplemented with 10 U/ml Turbo DNase and 100µg/ml cycloheximide). After lysis samples 478 stood on ice for 2h and subsequent Ribo-seq library generation was performed as previously 479 described 50 . Briefly, cell lysate was treated with RNAseI for 45min at room temperature 480 followed by SUPERase-In quenching. Sample was loaded on sucrose solution (34% sucrose, 481 20mM Tris 7.5, 150mM NaCl, 5mM MgCl2, 1mM dithiothreitol and 100µg/ml cycloheximide) 482 and spun for 1h at 100K RPM using TLA-110 rotor (Beckman) at 4c. Pellet was harvested using 483 TRI reagent and the RNA was collected using chloroform phase separation. For size selection, 484 15uG of total RNA was loaded into 15% TBE-UREA gel for 65min, and 28-34 footprints were 485 excised using 28 and 34 flanking RNA oligos, followed by RNA extraction and ribo-seq reads the p-site of the ribosome was calculated according to read length using the off-set from 498 the 5' end of the reads that was calculated from canonical cellular ORFs. The offsets used are 499 +12 for reads that were 28-29 bp and +13 for reads that were 30-33 bp. Reads that were in 500 different length were discarded. In all figures presenting ribosome densities data, all footprint 501 lengths (28-33bp) are presented.

Junctions spanning reads were quantified using STAR 2.5.3a aligner 53 , with running flags as 503 suggested by 44 , to overcome filtering of non-canonical junctions. Reads aligned to multiple 504 locations were discarded 44 . 505

For the metagene analysis only genes with more than 10 reads were used. For each gene, 506 normalization was done to its maximum signal and each position was normalized to the number 507 of genes contributing to the position. 508

Filtering of genes, quantification and RPKM normalization 509 

A novel coronavirus from patients with pneumonia in China

A pneumonia outbreak associated with a new coronavirus of probable bat 598 origin

Continuous and Discontinuous RNA 600 Synthesis in Coronaviruses

SARS -beginning to understand a new virus

The Nonstructural Proteins Directing Coronavirus 604 RNA Synthesis and Processing

Coronavirus biology and 607 replication: implications for SARS-CoV-2

Proteomics of SARS-CoV-2-infected host cells reveals therapy targets

Characterisation of the transcriptome and proteome of SARS-CoV-612 2 reveals a cell passage induced in-frame deletion of the furin-like cleavage site from the 613 spike glycoprotein

The coding capacity of SARS-CoV-2

Translational control in 617 virus-infected cells

Stimulation of Innate Immunity by Host and Viral RNAs

Interferon-Stimulated Genes: What Do They All Do?

Interplay between SARS-CoV-2 and 623 the type I interferon response

Interferon deficiency can lead to severe COVID

Impaired type I interferon activity and inflammatory responses in severe 627 COVID-19 patients. Science (80-. )

Autoantibodies against type I IFNs in patients with life-threatening

Inborn errors of type I IFN immunity in patients with life-threatening

Imbalanced Host Response to SARS-CoV-2 Drives Development 633 of COVID-19

Bulk and single-cell gene expression profiling of SARS-CoV-2 infected 635 human cell lines identifies molecular targets for therapeutic intervention

Emerging roles for RNA degradation in viral replication 638 and antiviral defense

A two-640 pronged strategy to suppress host protein synthesis by SARS coronavirus Nsp1 protein

Middle East Respiratory Syndrome Coronavirus nsp1 Inhibits 643

Host Gene Expression by Selectively Targeting mRNAs Transcribed in the Nucleus while

Sparing mRNAs of Cytoplasmic Origin

Viral and Cellular mRNA Translation in

Advances in Virus Research

SARS-CoV-2 Disrupts Splicing, Translation, and Protein 649 Trafficking to Suppress Host Defenses

Structural basis for translational shutdown and immune evasion by the

Nsp1 protein of SARS-CoV-2. Science

Dynamic competition between SARS-CoV-2 NSP1 and mRNA on 653 the human ribosome inhibits translation initiation

SARS-CoV-2 Nsp1 binds the ribosomal mRNA channel to inhibit 656 translation

SARS-CoV-2 Nsp1 suppresses host but not viral translation through a 658 bipartite mechanism

SARS-CoV-2 ORF6 disrupts nucleocytoplasmic transport through 660 interactions with Running title: SARS-CoV-2 ORF6 disrupts mRNA nuclear export

Evasion of Type I Interferon by SARS-CoV-2

Ribosome profiling of mouse embryonic 664 stem cells reveals the complexity and dynamics of mammalian proteomes

Ribosome Profiling as a Tool to Decipher Viral

A molecular pore spans the double membrane of the coronavirus 669 replication organelle. Science (80-. )

Imaging protein synthesis in cells and tissues with 671 an alkyne analog of puromycin

An inflammatory cytokine signature predicts COVID-19 severity 673 and survival

High-Resolution Analysis of Coronavirus Gene Expression by RNA

A putative new SARS-CoV protein, 3c, encoded in an ORF overlapping ORF3a

Sarbecovirus comparative genomics elucidates gene 679 content of SARS-CoV-2 and functional impact of COVID-19 pandemic mutations

Coding potential and sequence 682 conservation of SARS-CoV-2 and related animal viruses

Molecular Mechanism of Scanning and Start Codon Selection in

Sequence Composition Underpin Selective Dependency of Nuclear Export of Long RNAs 687 on NXF1 and the TREX Complex

Genes with 5' terminal oligopyrimidine tracts preferentially escape global 689 suppression of translation by the SARS-CoV-2 NSP1 protein

The viral protein NSP1 acts as a ribosome gatekeeper for shutting down 692 host translation and fostering SARS-CoV-2 translation Runing title: SL1 controls NSP1 693 for SARS-CoV-2 viral translation

The architecture of SARS-CoV-2 transcriptome

Mechanisms and enzymes involved in SARS coronavirus genome 698 expression

Alphacoronavirus Transmissible Gastroenteritis Virus nsp1 Protein 700

Suppresses Protein Translation in Mammalian Cells and in Cell-Free HeLa Cell Extracts 701 but Not in Rabbit Reticulocyte Lysate

SARS coronavirus nsp1 protein induces template-dependent 703 endonucleolytic cleavage of mRNAs: Viral mRNAs are resistant to nsp1-induced RNA 704 cleavage

A SARS-CoV-2 protein interaction map reveals targets for drug 706 repurposing

A Regression-Based Analysis of Ribosome-Profiling Data Reveals a 708

Comprehensive annotations of human herpesvirus 6A and 6B genomes 710 reveal novel and conserved genomic features

The Transcription and Translation Landscapes during Human 712

Ultrafast and memory-efficient 715 alignment of short DNA sequences to the human genome

STAR: ultrafast universal RNA-seq aligner

PANTHER pathway: an ontology-based pathway database coupled 719 with data analysis tools

PANTHER version 14: 721 more genomes, a new PANTHER GO-slim and improvements in enrichment analysis 722 tools