key: cord-0847626-7xt0cis8
authors: Lan, Shuiyun; Tedbury, Philip R.; Ong, Yee Tsuey; Shah, Raven; Slack, Ryan L.; Cilento, Maria E.; Zhang, Huanchun; Du, Haijuan; Lulkin, Nicole; Le, Uyen; Kirby, Karen A.; Melcak, Ivo; Cantara, William A.; Boggs, Emerson A.; Sarafianos, Stefan G.
title: Subgenomic SARS-CoV-2 replicon and reporter replicon cell lines enable ultrahigh throughput antiviral screening and mechanistic studies with antivirals, viral mutations or host factors that affect COVID-19 replication
date: 2021-12-30
journal: bioRxiv
DOI: 10.1101/2021.12.29.474471
sha: fd5bbacc06e5abe96e64f93efe48c2e2fc29715e
doc_id: 847626
cord_uid: 7xt0cis8

Replicon-based technologies were used to develop reagents and assays for advanced drug discovery efforts against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and for examining all facets of the SARS-CoV-2 replication cycle at reduced biocontainment level. Specifically: a) 21 replicons were cloned in bacterial artificial chromosomes (BACs) and delivered as transfectable plasmid DNA or transcribed RNA in various cell types. Replicons carrying mutations that affect the activity or antiviral susceptibility of SARS-CoV-2 enzymes were used to establish utility for mechanistic studies while reducing the community risks associated with gain-of-function studies in fully infectious virus. b) A BHK-21 stable cell line harboring SARS-CoV-2 replicon was generated and characterized in robust high/ultra-high throughput assays of antiviral efficacy with orthogonal SARS-CoV-2 replication reporter genes (Nano luciferase and enhanced green fluorescent protein-eGFP); the estimated antiviral potencies in the fully infectious SARS-CoV-2 system and in the transient or stable replicon systems were similar. HEK293 and Calu1 stable cell lines expressing SARS-CoV-2 replicon have also been prepared. Finally, c) we generated trans-encapsidated replicons by co-expression with SARS-CoV-2 structural proteins, thus producing single-round infectious SARS-CoV-2 virus-like particles that are able to transduce susceptible cell types and have expanded utility to enable study of virion assembly and entry into target cells. Hence, these SARS-CoV-2 replicon-based reagents include a novel approach to replicon-harboring cell line generation and are valuable tools that can be used at lower biosafety level (BSL2) for drug discovery efforts, characterization of SARS-CoV-2 and variant evolution in the COVID-19 pandemic, mechanisms of inhibition and resistance, and studies on the role of SARS-CoV-2 genes and host dependency factors.

As of the end of 2021, we are at the peak of the global COVID-19 pandemic, a respiratory disease with more than 280 million confirmed cases (record-high single-day infections on 12/23/2021 of 982,000 cases) and over 5.4 million fatalities around the world (1). The causative agent of COVID-19 (2, 3) is Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), an enveloped, positive-sense, singlestranded RNA betacoronavirus of the order Nidovirales, family Coronaviridae. Prior to SARS-CoV-2, illness caused by endemic coronaviruses (4) , and the outbreaks of the original SARS-CoV in 2002 (5) (6) (7) and Middle-East Respiratory Syndrome CoV (MERS-CoV) in 2012 (8) (9) , highlight the threat coronaviruses pose to human health.

The vaccines currently in use for the prevention of SARS-CoV-2 infection (10, 11) have proven to be effective at reducing the hospitilization and mortality from COVID-19, and are likely to continue to be a major part of the response to this and future outbreaks. However, vaccination strategies face multiple challenges, including the limited global vaccine supply, the difficulty in achieving the theoretical herdimmunity threshold at current vaccination rates, the reported waning of immunity against the virus a few months after vaccination, and the emergence of SARS-CoV-2 variants that spread very efficiently and are extensively mutated at the Spike surface glycoprotein, which has been the main target for vaccine development (12, 13) (14) (15) (16) . Therefore, effective antivirals that target viral proteins less likely to mutate will be an critical component of the global response to coronaviral outbreaks.

The nucleoside analogue remdesivir (RDV) was the first SARS-CoV-2-targeting antiviral approved by FDA for treatment of COVID-19 patients requiring hospitalization (reviewed in (17)). However, it can only be injected at a hospital setting and its efficacy has been questioned (18, 19) . Recently, another nucleoside analog was approved: molnupiravir (EIDD-2801 or MK-4482), which is the orally available pro-drug of the β-d-N4-hydroxycytidine (20) . Molnupiravir reduces hospitalization or death by 30% (21, 22) . Viral proteases have also been excellent targets for the development of antiviral drugs (23, 24) .

Several have been reported to have activity in vitro or are in clinical trials (25) (26) (27) (28) , and one formulation, Paxlovid (nirmatrelvir with ritonavir) (29, 30) , was recently approved for use in the USA (31) . Paxlovid was reported to reduce the risk of hospitalization or death by 89% in non-hospitalized high-risk adults (clinical trial NCT04960202) (32) . These major targets, and several other enzymatic activities are essential to replicon activity, allowing replicons to be used for initial screening for drug discovery, and characterization of mechanism of action and potential mechanisms drug of resistance, following mutation of the viral genome.

There is clearly a need for discovery and development of novel drugs that target SARS-CoV-2 replication. The virus is readily cultured and infectious clones have been produced with reporter viruses to facilitate such studies (33, 34) . However, a challenge for such studies is the requirement of handling infectious SARS-CoV-2 in biosafety level-3 (BSL3) facilities, which limits the number of academic and pharmaceutical laboratories able to contribute to the search for effective therapeutics. Subgenomic replicon systems provide biologically safe models that recapitulate a large part of the viral replication cycle and can play a major role in the discovery and development of antivirals, in addition to providing an invaluable tool for basic research into viral replication (35) (36) (37) .

The general strategy of producing subgenomic replicons involves removing the structural protein coding sequences and replacing them with reporters and/or selectable markers, while retaining all proteins and any nucleic acid sequences required for genome replication. The genomic RNA (gRNA) of coronaviruses is flanked by 5' and 3' untranslated regions (UTR); the first two thirds of the genome codes for the non-structural proteins (nsp) in the form of ORF1a and ORF1b polyproteins (reviewed in (38, 39) ), the final third codes for the structural proteins (spike S, Envelope E, Membrane M, and Nucleocapsid N), as well as several putative ORFs for accessory factors (9) (Figure 1A) . The genome has a 5' cap and 3' polyA tail and serves initially as the template for translation of ORF1a and ORF1b, into the pp1a and pp1b polyprotein precursors that are processed by the viral proteases nsp3 and nsp5 to release 16 mature proteins (nsp1-16, key functions listed in Figure 1A ) including all predicted components of the replication-transcription complex (RTC). The RTC produces negative strand (-) genomic and subgenomic (sg) RNAs from the genomic RNA. The generation of (-) sgRNAs is regulated by the transcriptional regulatory sequences (TRS). During the synthesis of the (-) RNA, the body TRSs (TRS-B) that precede each ORF ( Figure 1A ) act as potential dissociation signals for the RTC leading to a strand transfer to the TRS leader region of the 5'-UTR (TRS-L), to resume the synthesis of (-) sgRNA. The (-) RNAs are used as templates to synthesize full-length genomic RNAs and a nested set of sgRNAs. The frequency with which each TRS-B triggers dissociation during (-) RNA synthesis determines the relative abundance of each sgRNA and genomic RNA, such that the genomic RNA is relatively scarce while some sgRNAs, and consequently the proteins they code, are highly abundant (39) .

A variety of SARS-CoV-2 replicon systems have been published to date, featuring a range of advantages and disadvantages. Most involve the production of an RNA carrying one or more reporters and can be introduced into permissive cells by electroporation (24, (40) (41) (42) (43) (44) (45) . Some additionally have demonstrated the capability to be trans-encapsidated by co-expression with the deleted structural elements to produce single round infectious particles (41, 43, 45, 46) .

Here, we report the generation of more than 20 SARS-CoV-2 replicons, including constructs containing orthogonal reporter proteins and selectable markers. The replicons are based on the "parental" Washington isolate (WA) as well as several Variants of Concern (VOC) and Variants Being Monitored (VBM). We validate the replication dependence of reporter gene expression by use of replication-inactive mutants and well characterized inhibitors of viral enzymes, and thus the utility of these replicons for characterization of antiviral agents. We show that these subgenomic replicons can be delivered as plasmid DNA or transcribed RNA; we also show that the replicons can be transencapsidated by co-expression with SARS-CoV-2 structural proteins to produce virus-like particles (VLPs) capable of transducing angiotensin converting enzyme 2 (ACE2)-expressing cells (when bearing the SARS-CoV-2 S protein) or non-ACE2-expressing cells (when bearing the vesicular stomatitis virus glycoprotein).

Importantly, we were able to generate stable cell lines harboring the SARS-CoV-2 replicon, a potentially valuable resource to facilitate identification and characterization of novel antivirals.

To facilitate the cloning of such large constructs, replicons were engineered in a bacterial artificial chromosomes (BAC) through stepwise cloning of DNA fragments (synthetic or prepared by RT-PCR from the SARS-CoV-2 Washington isolate, WA) into the BAC, exploiting unique restriction sites ( Figure 1B ).

Many replicons have employed T7 promoters to permit in vitro transcription of replicon RNA, to be introduced into cells by electroporation (24, (40) (41) (42) (43) (44) (45) . To remove the necessity of generating RNA, an alternative approach employs a 5' cytomegalovirus (CMV) promoter to drive initial transcription of replicon RNA (46) (47) (48) (49) (50) . We built a replicon with a T7 promoter and a 3' hepatitis delta virus ribozyme to ensure generation of the correct 3' terminus. The base design incorporated the viral sequences essential for genome replication: the 5' and 3' UTRs, ORF1a and ORF1b, and N with its TRS-B. Reporters [Nano luciferase (NLuc) or NLuc with eGFP] and selectable markers (puromycin-N-acetyltransferase -Puro R or neomycin phosphotransferase -Neo R ) were inserted in place of M, E, S and most of the accessory ORFs, following the TRS-B sequences for M or S (TRS-M or -S) to permit transcription of the sgRNA. Placement of the reporter cassette in a sgRNA ensures that these genes will be expressed at a high level, but only following transcription by the SARS-CoV-2 RTC. This replicon produced a high luciferase signal that could be inhibited by RDV, indicating that the vast majority of reporter gene expression is generated by SARS-CoV-2 replication ( Figure S1 ). However, to reduce labor and potentially increase reproducibility between constructs, we generated a second generation of replicon, replacing the T7 promoter with CMV promoter. After transfection of the BAC into target cells, transcription of full-length RNA is driven by the CMV promoter, allowing translation of ORF1a and ORF1b and formation of SARS-COV-2 RTC (Figure 1) .

The RTC then produces new gRNA and sgRNAs, via (-) RNA intermediates, leading to expression of the reporter genes and selectable markers. The BAC platform provides a stable platform for the generation of these large constructs; the introduction of unique restriction sites allows for modification of the reporter gene cassette or introduction of mutations into the replicon sequence through replacement of individual fragments. A partial list of replicons generated is shown in Supplemental Table I .

A. Genome structure of SARS-CoV-2 and open reading frames (ORF). L, is the leader sequence. Some non-structural proteins, nsp, and their functions are: nsp3, contains PLPpro, papain-like protease; nsp5, 3CLpro protease; nsp12, RNA-dependent RNA polymerase or RdRp; nsp13, helicase; nsp14, ExoN, exonuclease; nsp15, EndoU, endonuclease; nsp16, MTase, methyltransferase. PLPpro-, and 3CLpro-cleavage sites are depicted as black or white triangles, respectively. TRSs are transcription-regulatory sequences located immediately adjacent to ORFs; TRSs contain a conserved 6-7 nt core sequence (CS) surrounded by variable sequences. During negativestrand synthesis, nsp12 pauses when it crosses a TRS in the body (TRS-B) and switches the template to the TRS in the leader (TRS-L); UTR, untranslated regions. Select mutations in various replicons are shown: F480L/V557, blue-RDV-resistance mutations; D618A/D760A, red-nsp12 active site; K345A/K347A, brown-nsp13 nucleic acid binding; D374A/E375A, nsp13 active site; D90A/E92A light brown nsp14 active site; H234A, purple-nsp15 active site; E203A green-nsp16 active site. B. General strategy for the construction of SARS-CoV-2 replicons. Multi-step cloning strategy was used, whereby the fragments named SARS-CoV-2/F1 to SARS-CoV-2/F9 were sequentially cloned into pBeloBAC11 vector to generate pBAC-SARS-CoV-2-REP. The genetic structure of the replicon and the position of relevant restriction enzyme sites are shown at the ends of each fragment. F9 contains the reporter cassette. Early SARS-2R versions included T7 promoter (downstream from the CMV promoter) and were used for in vitro transcription of replicon RNA.

To determine the optimal conditions for using nanoluciferase-producing SARS-CoV-2 replicon (SARS-2R_NL) we measured secreted NLuc over time in a range of cell lines (Figure 2) . To differentiate between NLuc produced from the CMV promoter and replication-dependent NLuc expression, we used SARS-2R_NL with inactive nsp12 polymerase (nsp12 D618A/D760 ). In all cases, nsp12 D618A/D760 samples exhibited >99% decrease in NLuc activity, demonstrating that reporter gene expression is overwhelmingly replication dependent. We found maximum NLuc signal in 293T, BHK-21 and CHO-K1 cells, while the greatest signal-to-noise (WT ≥10,000-fold greater than nsp12 D618A/D760 ) was achieved in BHK-21 and

Caco2. As expected, mock controls (no SARS-2R_NL) produced no NLuc activity, while a robust signal for SARS-2R_NL WT over nsp12 D618A/D760 was observed in all cell lines (Figure 2) . In most cell lines, NLuc activity is measurable by 8 hours post transfection (hpt), with the signal typically plateauing 48 hpt. The differences in NL activity following SARS-2R_NL transfection among the various cell lines are likely dependent on a combination of transfection efficiency and differences in replication permissivity. These data demonstrate that protein expression from sgRNAs is replication dependent, and the reporter gene assays can be performed with high signal-to-noise in a range of convenient and/or physiologically relevant cell types, making this a suitable system for the study of SARS-COV-2 replication or screening for replication inhibitors. 

To determine whether SARS-2R replicons can be used to study the role of individual viral proteins and viral mutations in the replication mechanism of SARS-CoV-2 we constructed SARS-2R_NL_Puro R replicons with mutations at various nsp active sites. Mutations at the predicted active sites of nsp12 (polymerase), nsp13 (helicase), and nsp15 (endonuclease), significantly reduce the activity of replicon activity in BHK-21 cells (Figure 3A ). In particular, the nsp12 D618A/D760A and nsp13 D374A/E375A mutations entirely suppressed replication (>99.9% loss of activity). Similarly, the nsp15 H234A mutation resulted in >95% loss of activity. Mutations that were predicted to affect the nucleic acid binding function of SARS-CoV-2 helicase (51) (nsp13 K345A/K347A ) were also detrimental to replicon activity (>80% loss of activity).

Interestingly, mutations at the exonuclease active site of nsp14 (nsp14 D90A/E92A ) and 2'-Omethyltransferase active site of nsp16 (nsp16 E203A ) only induced partial loss of replication (approximately 50% in each case), indicating that these enzymes/functions are less critical to replication, or that under these conditions the mutations may not cause a complete loss of activity. ExoN(-) mutant SARS-CoV-2 than we initially observed (54), we repeated our analysis of these mutants in a range of cells and consistently observed a moderate reduction in replicon activity (typically 50-75%);

Calu3 cells exhibited a more profound reduction (~98%) in activity with the ExoN(-) mutant replicon, however, replicon activity was lower generally in this cell line, potentially due to reduced transfection efficiency ( Figure 3B ). These data provide further validation of the replication-dependence of reporter gene expression, and indicate that this system would be particularly useful for studies of the polymerase and helicase, including screening for and characterization of small molecule inhibitors.

One of the major concerns during the COVID-19 pandemic has been the emergence of variants that exhibit properties such as faster spread, immune escape, and potentially increased pathogenicity.

Although many studies have focused on the role of the S protein, there is increasing evidence that other mutations throughout the genome contribute to the phenotypic differences between variants.

Additionally, mutations in the RTC enzymes have the potential to affect susceptibility to current and future therapeutics.

We generated SARS-2R_NG_Neo R _NL constructs derived from parental (WA), the alpha VBM B.1.1.7 (also known as "UK strain"), the beta VBM B.1.351 (also known as "South African or SA"), and the delta VOC B.1.617.2 (delta) viruses. Of note, the classification of SARS-CoV-2 variants as "of concern", "of interest", or "being monitored" has evolved during the course of the pandemic but the above variants have been thus classified by CDC as of 12/20/21. The VOC replicons differ from WA in ORF1a, ORF1b, N, and in non-coding regions (Supplemental Table I ). To gain insight into whether the amino acid differences among variants in genes other than the S affect the replication efficiency of the B.1.1.7, B.1.351, and B.1.617.2 strains, we assessed replication in 293T cells ( Table I) . All replicons evinced clear replication above the polymerase-defective mutant WA replicon. However, the B.1.351-and B.1.1.7-derived replicons showed measurably reduced replication relative to the WA-derived replicon, suggesting that the mutations in these strains may affect fitness in the 293T cell culture model. We next compared the antiviral susceptibilities of the VOC replicons in 293T cells. We found no statistically significant difference between the potency of RDV, NHC, EIDD-2801, and GC-376 among the four replicons (Table I) . To verify that this lack of change represented the biology of the virus, we compared the replicon values for WA and B.1.1.7 to the values obtained with same compounds in fully infectious WA and B.1.1.7 viruses; indeed, we did not find significant differences in antiviral potency for any of the compounds against the two strains (Supplemental Table II ). These data demonstrate the similarity between virus and replicon in terms of drug resistance, and that these constructs represent a valuable resource for comparative studies of drug susceptibility of the major variants. 

Initial experiments with SARS-CoV-2 replicons were conducted in 24-well plates. To provide evidence for the utility of SARS-2R as a tool for drug discovery efforts we explored conditions for its use in high throughput multi-well plate formats. We showed that reproducible measurements of RDV inhibition can be carried out in a 384-well format, at a range of cell seeding densities using 293T cells, without significant differences in the estimated EC 50 and EC 90 s (Table II and Figure S2 ). Of note, these data demonstrate that the assay is reliable even at 1,000 cells/well, which is well within the range of an ultrahigh throughput format in 1536-well plates. Assay miniaturization is underway. These values are comparable to published data using infectious virus (20) . The variation in RDV potency between cell types has been observed previously and likely reflects efficiency of conversion of RDV nucleotide to RDVtriphosphate (55, 56) . Finally, to quantify the robustness of the assay, we calculated the Z', comparing DMSO treatment to RDV; the average Z' and 95 % CI was 0.70 ± 0.032. These data support the robustness and reproducibility of the replicon-based assay, and its suitability to high-throughput applications. We next used high throughput conditions with SARS-2R_NL_Puro R to determine the potency of several antivirals with well-characterized activity against SARS-CoV-2 in 293T cells (Table III) (Table III) analogous to the modest resistance previously reported in MHV (nsp12 F476L/V553L ) and SARS-CoV (nsp12 F480L/V557L ) (up to 5.6-fold) (57) . A decrease in RDV susceptibility was observed when these replicons were tested in BHK-21 cells, however, it was not found to be statistically significant under the testing conditions (not shown). The RDV-resistant replicon appears to retain susceptibility to NHC, EIDD-2801, and GC-376.

It is also of interest to understand whether inhibition by various antivirals that target nsp12 and are incorporated in the elongating RNA chain can be affected by the ExoN activity of nsp14. In MHV, the ExoN(-) mutant had increased sensitivity to RDV (57) . Hence, we also tested the potency of such antivirals using the SARS-2R_NL_Puro R nsp14 D90A/E92A replicon that lacks ExoN activity. We found that genetic suppression of this activity did not significantly affect the EC 50 of NHC or EIDD-2801. Moreover, we did not observe any significant effect of the ExoN(-) mutations on RDV susceptibility (Table III) ; previously, Agostini et al. reported a 4.5-fold increase in sensitivity of the corresponding mutant MHV compared to the WT virus (57) . Surprisingly, we did find a statistically significant increase in GC-376 susceptibility of SARS-2R_NL_Puro R nsp14 D90A/E92A compared to WT (two-way ANOVA p-value 0.003). We are currently investigating whether (and how) the mutations that suppress the ExoN function of nsp14 affect the function of nsp5 and its interactions with GC-376.

In addition to resistance to an antiviral agent, the potential risk associated with drug-resistance mutations depends on their fitness. We used the replicon system to examine the effects of RDVresistance associated mutations on fitness. Consistent with the reported modest decrease in production of infectious virus (57) our data show that the nsp12 F480L/V557L mutations reduce the replicon activity by ~20 % (Figure S3) . Interestingly, the nsp12 F480L/V557L mutations in the background of ExoN(-) nsp14

(nsp12 F480L / V557L , nsp14 D90A/E92A ) seemed to significantly suppress replicon activity by 98%, far more than either pair of mutations alone. Collectively, these data illustrate the value of replicons for exploring the potential impacts of drug-resistance associated mutations on SARS-CoV-2 replication.

While the use of replicon DNA to transfect BHK-21 cells streamlines the workflow relative to the use of transcribed RNA, the workflow can be simplified further through the use stable replicon harboring cell lines; similar tools have proven extremely beneficial in drug development for HCV (58, 59) . Surprisingly, while a stable replicon harboring cell line was generated for SARS-CoV (60) Figure 4) . Thus, this cell line can be used for multiplex assays that concomitantly follow NLuc and fluorescent signals, allowing for following highly sensitive and live replicon signal. There was no observable cytotoxicity due to the replicon and the cell morphology was typical of BHK-21 cells.

The generation of a replicon stable cell line in which on a small percentage of cells exhibit evidence of active replicon gene expression is atypical and prompted further characterization. We performed qPCR and RT-qPCR to determine the levels of replicon RNA and plasmid DNA present in the cell line. After several weeks of passaging following transfection, we expected to find no replicon plasmid DNA, however, we found an estimated 1.2 (±0.9) replicon DNA copies per cell. We also found 35 (±26) replicon genomic RNA copies per cell. These data suggest that most/all of the cells carry an integrated copy of the plasmid, but at any given time only a small percentage of those cells contain active replicons. To determine whether replicon activity was a stable phenotype, we separated the population of 

A major application for such a stable cell line is to facilitate compound screening and characterization of hit compounds. To validate the use of the BHK-SARS-2R_GFP_Neo R _NL cell line, we determined the EC 50 for RDV using both the NLuc and GFP markers (Table V and Figure S7 ). The EC 50 values determined using either reporter were generally comparable, both to one another and to values reported in the literature (2, 20, 27, 55). Of note, some differences in RDV potency have also been reported when tested in different cell lines, due to changes in its intracellular metabolism. 

The replicon system provides an invaluable resource for the study of processes related to replication of RNA and protein expression, however, the absence of essential structural proteins makes it impossible to study the processes of particle assembly and release, and virus entry. To expand the utility of the replicon system to address these functions, while retaining the advantages of a BSL2 compatible system, we supplied the M, E and S in trans by co-transfecting structural protein expression vectors with the replicon plasmid SARS-2R_GFP_Neo R _NL into 293T cells. After 48 h, media were collected and filtered, and used to transduce SARS-CoV-2 susceptible cell lines, either 293T-hAT or Huh7.5-hAT. After a further 48 h, reporter gene expression was measured as GFP-positive cells and NLuc activity in the media.

We were able to transduce both of the ACE2-expressing cell lines with VLPs bearing the SARS-CoV-2 spike protein (VLP_SARS2-S) or the SARS-CoV spike protein (VLP_SARS-S), but not with VLPs lacking any envelope protein (VLP_ Mock) (Figure 6 ). Comparing 293T cells (which do not endogenously express ACE2) to ACE2-and TMPRSS2-expressing 293T-hAT cells, we confirmed that both VLP_SARS2-S and VLP_SARS-S require ACE2 for successful transduction (Figure 6A) . Similarly, comparing Huh7.5 cells to Huh7.5-hAT (Huh7.5 cells express ACE2; Huh7.5-hAT express additional ACE2 and TMPRSS2) revealed enhanced transduction with VLP_SARS2-S and VLP_SARS-S; by contrast VLP_ VSV-G particles were able to transduce the target cells; similarly, transduction of ACE2-expressing cells by VLP_SARS-2R_GFP_Neo R _NL_S was inhibited by a SARS-CoV-2 neutralizing monoclonal antibody. As a final control, we confirmed that reporter gene expression was dependent on replicon replication, following transduction, as all VLPs could be inhibited by treatment with RDV. The replicon comprised all replication-essential proteins and sequences but omitted the structural proteins S, E, and M; additionally, reporter genes were included as a separate ORF using the TRS-B of M to drive production of the sgRNA expressing the reporter cassette. Placing the reporter cassette in a sgRNA ensured that formation of a functional RTC was essential for reporter gene expression. We validated the replication dependence using both genetic ablation of essential viral enzymes and the use of well characterized inhibitors of SARS-CoV-2 replication. Combining these controls with a screen of suitable cell lines, we were able to demonstrate replication-dependent reporter expression with signalto-noise ratios between 1:100 and 1:>10,000; this wide dynamic range combined with an excellent z' score indicate that these replicons will be well suited to small molecule screening and should be able to resolve a wide range of magnitude of phenotypes associated with modulation of viral or cellular factors. (61) . They also showed that the delta P681R mutation on S enhanced the cleavage of the full-length spike to S1 and S2, leading to increased infection via cell surface entry. The measurably increased replication of the delta replicon suggests that the increased transmissibility of the delta VOC may also be due in part to mutations at regions outside the

In addition to the dynamic range of the assay, a major concern when screening for small molecules inhibitors of viral replication is the potential of scoring inhibitors of the reporter gene as antiviral hits. To mitigate that risk, we have generated replicons with orthogonal reporter genes (NLuc and GFP), an approach successfully employed by others (40, 46, 50) , and shown that measurements of antiviral potency are similar, whichever reporter is used.

In a final step to enhance the utility of these replicons for small molecule screening, we generated is necessary to give the cells the opportunity to produce neomycin phosphotransferase, the more rapid toxicity of puromycin may kill cells when the replicon is inactive.

If the replicon-harboring cells are only transiently exhibiting RNA replication, this system may be unsuited for evolution of resistance studies, as the replicons will not have the opportunity to acquire mutations and persist over time under selection; nevertheless, our approach provides a powerful, convenient and reproducible system for library screening and studies of putative antivirals. Moreover, reproducibility will be enhanced by continuous re-initiation of replicon expression from the DNA, such that the replicon will not deviate from the cloned sequence over time; in autonomously replicating RNA cell lines, replicons are often seen to undergo culture adaptation (for example (64, 65) ). The stability of these DNA-based replicon-harboring cell lines could be of particular value when it is necessary to be able to compare potencies of compounds over large numbers of experiments.

Finally, to expand the utility of the replicons in the study of SARS-CoV-2 replication, we complemented replicon expression with expression of structural proteins, as has been employed in other RNA virus replicon systems (66) including SARS-CoV-2 (41, 43, 45, 46) , and were able to demonstrate transmission of SARS-CoV-2 particles bearing S (of either SARS-CoV or SARS-CoV-2) or VSV-G. Reporter gene expression was shown to be dependent on receptor binding and replication, making this a viable approach to study effects on particle assembly and release, and binding and fusion to target cells.

Introduction of these replicons with cells expressing structural proteins in trans may permit viral spread and thus passaging, as seen in other SARS-CoV-2 replicon systems (41, 45) .

In conclusion, the replicon-based systems described here provide tools to examine all facets of the SARS-CoV-2 replication cycle, from cell entry, through protein expression and genome replication, to particle assembly, and can be applied to the diverse range of existing and emerging virus variants. In addition to the numerous applications to basic research, the production of genetically stable, cDNAbased, replicon-harboring cell lines has great potential value for the identification and evaluation of antivirals. The use of integrated cDNA-based replicons may also represent a viable strategy for the production of stable cell lines based on RNA virus replicons that do not readily establish persistent RNAbased replicon-harboring cell lines.

VeroE6 (ATCC) are derived from African green monkey kidney cells and lack the genes encoding type I interferons (67, 68) . Huh-7.5 (provided by Charles Rice) are derived from human hepatoma cells. Huh7.5- The sequences of the SARS-2R replicons were confirmed by both Sanger and deep sequencing (initial WT constructs), and subsequent mutations were confirmed by Sanger sequencing the relevant regions.

The DNA template for T7 transcription was prepared by linearizing the T7 replicon plasmid (#453, Supplemental Nano luciferase assays were performed 48 h post electroporation.

Cells were seeded in 24-well plates and incubated until approximately 80 percent confluent. Then cells were transfected with either mock (no replicon) or 0.25 µg replicon plasmid and 0.025 µg N, ORF3b, ORF6 expression plasmids with jetPRIME transfection reagent (Polyplus). In all cases, the amount of SARS-2R nucleic acid transfected was quantified by both agarose gel electrophoresis with ethidium bromide staining and quantification of the absorbance at 260 nm and 280 nm using a NanoDrop spectrophotometer (ThermoFisher Scientific). Nano luciferase activity was assayed using the Nano-Glo

Luciferase Assay System (Promega) following manufacturer's instructions. The luciferase signal was measured using a GloMax Navigator Microplate Luminometer (Promega).

BHK21 cells seeded in 6-well plate were transfected with SARS-2R using jetPRIME transfection reagent (Polyplus transfection). At 16 h post transfection, cells were trypsinized then seeded into 96-or 384-well plate and treated with serial diluted compounds. Stable cells were seeded directly and treated with serial diluted compounds. Nano luciferase assays were performed 48 hours post treatment. Dose response curves were calculated with Prism software.

BHK-21 cells were seeded into a 6-well plate, then transfected with replicon SARS-2R_NG_Neo R _NL using jetPRIME transfection reagent following the manufacturer's instructions. The following day, cells were trypsinized and transferred into a 10 cm dish. After a further 24 h incubation, G418 was added at 1 mg/ml and cells were maintained until distinct colonies formed. Individual colonies were picked and transferred to a 24-well plate. These clones were expanded until near confluency, then NLuc activity was assayed to confirm the presence of active replicon replication. Clones that exhibited a robust NLuc signal were trypsinized then seeded into two separate wells, one treated with 1 μM RDV to inhibit replicon replication. Clones that exhibited a robust NLuc signal that could be inhibited with RDV were expanded for future studies.

BHK-SARS-2R_GFP_NeoR_NL cells were trypsinized and washed twice with DPBS (Sigma-Aldrich) and

once with sort buffer (PBS supplemented with 0.3 % BSA, 2 mM EDTA, 25 mM HEPES pH 7.0). Cells were adjusted to 1-2 x 10 7 cells/ml in sort buffer. GFP+ and GFP-cell sorting was carried out using a BD FACS Aria II SORP Cell Sorter (Pediatric/Winship Flow Cytometry Core, Emory University). After sorting, GFP+ and GFP-cells were seeded into 96-wells at 2500-5000 cells/well. Cells were stained with Hoechst 33342 and counted for GFP+ and total cells with Cytation 5 high-content live-cell imaging system (Biotek) using a 4x objective.

Cells were seeded into 96-well plates at 10,000 cells per well. After 24 h, culture media were removed and replaced with media supplemented with ruxolitinib suppress IFN signaling, sodium butyrate to enhance transcription, and/or RDV to suppress SARS-CoV-2 replication. GFP-positive cells were imaged on a Cytation 5 high-content live-cell imaging system (Biotek) using a 4x objective at 37 o C in 5% CO 2 over 48 to 72 h. Images were captured as a 5x4 matrix and GFP positive cells were enumerated using Gen5 software (Biotek).

RNA and DNA was harvested from the BHK-SARS-2R_GFP_Neo R _NL cell line using the RNeasy mini kit (Qiagen) and the QIAamp genomic DNA mini kit (Qiagen), respectively. RNA was treated with DNase I (Qiagen) before reverse transcription. Reverse transcription was performed by using the SuperScript® III First-Strand Synthesis System (ThermoFisher Scientific) with a gene specific primer 5'tgaagtctgtgaattgcaaag-3'. qPCR was performed with QuantStudio 3 and ABsolute qPCR SYBR Green Mix (Thermo Fisher Scientific) using primers SEM397 5'-cttatgattgaacggttcgtgtc-3' and SEM398 5'cagaatacatgtctaacatgtgtcc-3'. SARS-CoV-2 nucleic acid copy numbers were derived from a calibration curve comprising six 10-fold dilutions of plasmid DNA for SARS-2R_GFP_Neo R _NL. Final values were averaged from two independent experiments and shown with 95% confidence intervals.

SARS-CoV-2 stocks were produced by infection of VeroE6 with SARS-CoV strain 2019-nCoV/USA_WA1/2020 or SARS2 hCoV-19/England/204820464/2020 (B.1.1.7). Media were harvested after 48 hours or when cytopathic effect was observed. To titer viral stocks, VeroE6 cells were seeded in a 96-well plate at 20,000 cells per well. After 24 h, infections were performed with serially diluted viral stocks. Cells were fixed 6 hours post infection (i.e., long enough for expression of viral proteins but short enough that no spread will have occurred) and stained for SARS-CoV-2 N protein to identify infected cells. Infected and total cells were determined by high content microscopy, virus titer was calculated as the number of infected cells per volume of virus stock added to cells.

Calu3 cells were seeded in a 96-well plate at 30,000 cells per well and infected when at least 80% confluent. Prior to infection, medium was replaced with fresh medium and supplemented with serially 

Cells were seeded in a 6-well plate then transfected with SARS-2R and treated with DMSO or RDV. After 48 h, the NLuc activity in the media was measured for 3 replicate samples per condition. The mean (µ) and standard deviation (σ) were calculated for each treatment condition. Z' was calculated as 1-[3σ DMSO + 3σ RDV]/[µ DMSO -µ RDV]; an excellent assay should have Z' > 0.5, indicating a wide range between positive and negative controls, combined with small standard deviations. A theoretically perfect assay would have Z' = 1.

Data were tabulated and graphs plotted using Excel (Microsoft) or Prism (Graphpad). Statistical tests were performed using Prism.

COVID-19 CORONAVIRUS PANDEMIC

Both Boceprevir and GC376 efficaciously inhibit SARS-CoV-2 by targeting its main protease

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

Hosts and sources of endemic human coronaviruses

A novel coronavirus associated with severe acute respiratory syndrome

Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome

Summary of probable SARS cases with onset of illness from 1

Middle East respiratory syndrome coronavirus (MERS-CoV): announcement of the Coronavirus Study Group

A new coronavirus associated with human respiratory disease in China

COVID-19 vaccine tracker

SARS-CoV-2 vaccines

Covid-19: global vaccine production is a mess and shortages are down to more than just hoarding

Five reasons why COVID herd immunity is probably impossible

Waning immunity after the BNT162b2 vaccine in Israel

The emergence, genomic diversity and global spread of SARS-CoV-2. Nature:1-11

Efficacy and safety of remdesivir for the treatment of severe acute respiratory syndrome due to coronavirus 19: systematic review and meta-analysis

Assessing the evidence on remdesivir

Remdesivir plus standard of care versus standard of care alone for the treatment of patients admitted to hospital with COVID-19 (DisCoVeRy): a phase 3, randomised, controlled, open-label trial

An orally bioavailable broad-spectrum antiviral inhibits SARS-CoV-2 in human airway epithelial cell cultures and multiple coronaviruses in mice

Molnupiravir for oral treatment of Covid-19 in nonhospitalized patients

Merck and Ridgeback Biotherapeutics Provide Update on Results from MOVe-OUT Study of Molnupiravir, an Investigational Oral Antiviral Medicine

Proteases and protease inhibitors in infectious diseases

A bacterial artificial chromosome (BAC)-vectored noninfectious replicon of SARS-CoV-2

Overview of antiviral drug candidates targeting coronaviral 3C-like main proteases

Potential of coronavirus 3C-like protease inhibitors for the development of new anti-SARS-CoV-2 drugs: Insights from structures of protease and inhibitors

Boceprevir, GC-376, and calpain inhibitors II, XII inhibit SARS-CoV-2 viral replication by targeting the viral main protease

3C-like protease inhibitors block coronavirus replication in vitro and improve survival in MERS-CoV-infected mice

Preclinical characterization of an intravenous coronavirus 3CL protease inhibitor for the potential treatment of COVID19

An oral SARS-CoV-2 Mpro inhibitor clinical candidate for the treatment of COVID-19

Coronavirus (COVID-19) Update: FDA Authorizes First Oral Antiviral for Treatment of COVID-19

Pfizer's Novel COVID-19 Oral Antiviral Treatment Candidate Reduced Risk of Hospitalization or Death by 89% in Interim Analysis of Phase 2/3 EPIC-HR Study

SARS-CoV-2 reverse genetics reveals a variable infection gradient in the respiratory tract

An infectious cDNA clone of SARS-CoV-2

Viral replicons as valuable tools for drug discovery

The hepatitis C virus replicon system: from basic research to clinical application

Rapid identification of coronavirus replicase inhibitors using a selectable replicon RNA

Nidovirales: evolving the largest RNA virus genome

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

Generation of SARS-CoV-2 reporter replicon for high-throughput antiviral screening and testing

A novel cell culture system modeling the SARS-CoV-2 life cycle

A PCR amplicon-based SARS-CoV-2 replicon for antiviral evaluation

Replication and single-cycle delivery of SARS-CoV-2 replicons

SARS-CoV-2 replicon for high-throughput antiviral screening

A trans-complementation system for SARS-CoV-2 recapitulates authentic viral replication without virulence

Development of a Single-cycle Infectious SARS-CoV-2 Virus Replicon Particle System for use in BSL2 Laboratories

Engineering a reliable and convenient SARS-CoV-2 replicon system for analysis of viral RNA synthesis and screening of antiviral inhibitors

Construction of a Noninfectious SARS-CoV-2 Replicon for Antiviral-Drug Testing and Gene Function Studies

DNAlaunched RNA replicon vaccines induce potent anti-SARS-CoV-2 immune responses in mice

Construction of Non-infectious SARS-CoV-2 Replicons and Their Application in Drug Evaluation

Delicate structural coordination of the Severe Acute Respiratory Syndrome coronavirus Nsp13 upon ATP hydrolysis

High fidelity of murine hepatitis virus replication is decreased in nsp14 exoribonuclease mutants

Infidelity of SARS-CoV Nsp14-exonuclease mutant virus replication is revealed by complete genome sequencing

The Enzymatic Activity of the nsp14 Exoribonuclease Is Critical for Replication of MERS-CoV and SARS-CoV-2

Remdesivir Inhibits SARS-CoV-2 in Human Lung Cells and Chimeric SARS-CoV Expressing the SARS-CoV-2 RNA Polymerase in Mice

Comparison of anti-SARS-CoV-2 activity and intracellular metabolism of remdesivir and its parent nucleoside

Coronavirus susceptibility to the antiviral remdesivir (GS-5734) is mediated by the viral polymerase and the proofreading exoribonuclease

Hepatitis C virus cell culture models: an encomium on basic research paving the road to therapy development

HCV replicon systems: workhorses of drug discovery and resistance

Derivation of a novel SARS-coronavirus replicon cell line and its application for anti-SARS drug screening

Delta spike P681R mutation enhances SARS-CoV-2 fitness over Alpha variant

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

Nonstructural protein 1 of SARS-CoV-2 is a potent pathogenicity factor redirecting host protein synthesis machinery toward viral RNA

Stable Cell Clones Harboring Self-Replicating SARS-CoV-2 RNAs for Drug Screen

Enhancement of hepatitis C virus RNA replication by cell culture-adaptive mutations

Replicon RNA viral vectors as vaccines

Regulation of the interferon system: evidence that Vero cells have a genetic defect in interferon production

Transcriptional and posttranscriptional regulation of exogenous human beta interferon gene in simian cells defective in interferon synthesis

Studies on avian infectious bronchitis virus (IBV)

T7 RNA polymerase-dependent and-independent systems for cDNA-based rescue of Rift Valley fever virus

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

Characterization of severe acute respiratory syndrome-associated coronavirus (SARS-CoV) spike glycoproteinmediated viral entry

Construction of a severe acute respiratory syndrome coronavirus infectious cDNA clone and a replicon to study coronavirus RNA synthesis

This research was supported in part by the Nahmias-Schinazi Chair fund at Emory University (SGS). The