SARS-CoV-2 nsp13 Restricts Episomal DNA Transcription without Affecting Chromosomal DNA

ABSTRACT

Nonstructural protein 13 (nsp13), the helicase of SARS-CoV-2, has been shown to possess multiple functions that are essential for viral replication, and is considered an attractive target for the development of novel antivirals. We were initially interested in the interplay between nsp13 and interferon (IFN) signaling, and found that nsp13 inhibited reporter signal in an IFN-β promoter assay. Surprisingly, the ectopic expression of different components of the RIG-I/MDA5 pathway, which were used to stimulate IFN-β promoter, was also mitigated by nsp13. However, endogenous expression of these genes was not affected by nsp13. Interestingly, nsp13 restricted the expression of foreign genes originating from plasmid transfection, but failed to inhibit them after chromosome integration. These data, together with results from a runoff transcription assay and RNA sequencing, suggested a specific inhibition of episomal but not chromosomal gene transcription by nsp13. By using different truncated and mutant forms of nsp13, we demonstrated that its NTPase and helicase activities contributed to the inhibition of episomal DNA transcription, and that this restriction required direct interaction with episomal DNA. Based on these findings, we developed an economical and convenient high-throughput drug screening method targeting nsp13. We evaluated the inhibitory effects of various compounds on nsp13 by the expression of reporter gene plasmid after co-transfection with nsp13. In conclusion, we found that nsp13 can specifically inhibit episomal DNA transcription and developed a high-throughput drug screening method targeting nsp13 to facilitate the development of new antiviral drugs.

IMPORTANCE To combat COVID-19, we need to understand SARS-CoV-2 and develop effective antiviral drugs. In our study, we serendipitously found that SARS-CoV-2 nsp13 could suppress episomal DNA transcription without affecting chromosomal DNA. Detailed characterization revealed that nsp13 suppresses episomal gene expression through its NTPase and helicase functions following DNA binding. Furthermore, we developed a high-throughput drug screening system targeting SARS-CoV-2 nsp13. Compared to traditional SARS-CoV-2 drug screening methods, our system is more economical and convenient, facilitating the development of more potent and selective nsp13 inhibitors and enabling the discovery of new antiviral therapies.

INTRODUCTION

Nonstructural protein 13 (nsp 13) is the helicase of SARS-CoV-2 and plays essential roles during viral replication (1, 2). Besides its helicase activity, nsp13 has also been reported to function as nucleoside triphosphatase (NTPase) in an NTP-dependent manner when facilitating the unwinding of DNA or RNA with a 5′ to 3′ polarity (3, 4). During this process, nsp13 can bind to the viral polymerase nsp12, possibly via mechano-regulation, to markedly enhance its own helicase activity (1, 2, 5). Because the polarity of the nsp13 helicase 5′ to 3′ translocation and polymerase are in opposition to each other, nsp13 was also reasoned to play a role in back-tracking, template switching, or disruption of downstream secondary structures (1, 6). In addition, nsp13 possesses RNA 5′-triphosphatase activity and is essential in the formation of the viral 5′ mRNA cap (7). Additionally, nsp13 has been shown to antagonize host innate immunity by repressing interferon β (IFN-β) production to facilitate viral replication (8–10).

In line with its essential roles in the viral life cycle, nsp13 is known to be the most conserved nonstructural protein (11). It contains five structural domains: an N-terminal zinc binding domain (ZBD), a helical “stalk” domain, a β-barrel 1B domain, and two “RecA-like” helicase subdomains, 1A and 2A. The ZBD domain is responsible for transmitting the unwinding signal to the nsp13 helicase core domains. The 1A and 2A domains can bind and hydrolyze nucleoside triphosphates, providing the resulting energy to the unwinding process. The five structural domains coordinate with each other to complete the final DNA or RNA unwinding (11, 12).

Given its conserved and indispensable role in coronavirus replication, nsp13 has been considered a promising target for the development of antiviral drugs (11). Despite decades of efforts to find chemical inhibitors of nsp13, to date, no nsp13 inhibitors have been approved for SARS-CoV-2 (4, 11, 13–17), probably due to their limited efficacy. Therefore, screening of effective antiviral drugs is urgently needed. In this study, we demonstrated that SARS-CoV-2 nsp13 can repress episomal but not chromosomal DNA transcription. Mechanically, the NTPase and helicase activities of nsp13 were exerted following direct binding of nsp13 with target episomal DNA. Given that the helicase function of nsp13 is correlated with its inhibitory effect on episomal DNA, we additionally developed a high-throughput drug screening method which utilizes the inhibitory effect of nsp13 on luciferase reporter plasmid expression after compound treatment to reflect the inhibitory effect of the compound on nsp13 helicase activity. We also validated the operability of this method by using bismuth salt, a known nsp13 inhibitor.

RESULTS

SARS-CoV-2 nsp13 suppressed episomal DNA expression.

Given that our initial goal was to study the interaction of SARS-CoV-2 and interferon signaling, we screened different SARS-CoV-2 proteins with an IFN-β promoter assay as previously reported (18). Similar to what has been described (8–10), we observed that while the luciferase expression driven by IFN-β promoter was induced by Sendai virus (SeV), the magnitude of this induction was remarkably reduced when additional nsp13 was added (Fig. 1A). To explore which step of the IFN induction pathway was affected by nsp13, we used different components of the RIG-I/MDA5 pathway to stimulate IFN-β promoter luciferase reporter, as has been previously described (9). As shown in Fig. 1B, overexpression of nsp13 suppressed IFN-β promoter-driven luciferase signals triggered by RIG-IN (the 2CARD domains of RIG-I) (9), MDA5, VISA, TBK1, and IKKε. Surprisingly, in addition to the negative impact of nsp13 on IFN-β reporter signals, nsp13 also downregulated the ectopic expression of Flag-RIG-IN, Flag-MDA5, Flag-VISA, Flag-TBK1, and Flag-IKKε, which were driven by cytomegalovirus (CMV) promoter (Fig. 1B). Moreover, the expression of IFN-β promoter-driven firefly luciferase (F-Luc), herpes simplex virus thymidine kinase (HSV-TK) promoter-driven Renilla luciferase, hPGK promoter-driven puromycin (Puro), and SV40 promoter-driven neomycin (Neo) were also inhibited by nsp13 (Fig. 1C and D). To further demonstrate this effect at the single-cell level, green fluorescent protein (GFP) and nsp13-expressing plasmids were co-transfected into cells for fluorescence microscopy. In keeping with the aforementioned data, imaging and quantification clearly showed that GFP intensity was attenuated by nsp13 (Fig. 1E). Furthermore, this suppression was not artifact of tagging, because non-tagged nsp13 exhibited similar effects (Fig. 1F). These results together suggested that nsp13 directly inhibited gene expression from transfected plasmids in general.

FIG 1.

The effect of SARS-CoV-2 nsp13 on episomal DNA and IFN-β. (A) HEK293T cells were co-transfected with IFN-β promoter firefly luciferase reporter plasmid (pIFN-β-luc), Renilla luciferase control plasmid (pRL-TK, normalizing transfection efficiency) and pHA-nsp13, pHA-IAV-NS1 (positive control) (32), or empty vector. 24 h post-transfection, cells were infected with SeV for 10 h, and luciferase activity was measured. The data were analyzed by normalizing the firefly luciferase (F-Luc) activity to the Renilla luciferase (R-Luc) activity. The values from cells transfected with empty vector together with SeV infection were set as 100%. HA-nsp13, HA-IAV NS1, and GAPDH (glyceraldehyde-3-phosphate dehydrogenase) levels were determined by Western blotting. (B) HEK293T cells were co-transfected with pIFN-β-luc, pRL-TK and pHA-nsp13, or empty vector, together with stimulator plasmid pFlag-RIG-IN, pFlag-MDA5, pFlag-VISA, pFlag-TBK1, or pFlag-IKKε for 24 h. 24 h posttransfection, luciferase activity was measured. Data were analyzed by normalizing the F-Luc activity to the R-Luc activity. The values from cells transfected with empty vector together with the relevant constructs were set as 100%. Flag-RIG-IN, Flag-MDA5, Flag-VISA, Flag-TBK1, Flag-IKKε, HA-nsp13, and GAPDH expression were determined by Western blotting. (C) HEK293T cells were co-transfected with pIFN-β-luc, pRL-TK, pHA-nsp13, or empty vector plasmid, and pHA-IAV-NS1 positive control. At 24 h posttransfection, cells were infected with SeV for 10 h, and F-Luc and R-Luc activity were determined. (D) Left: HEK293T cells were co-transfected with pLKO.1 and pHA-nsp13 or empty vector for 24 h. Puro mRNA was determined by RT-qPCR. Right: Huh-7 cells were co-transfected with pcDNA3.1 and pHA-nsp13 or empty vector for 24 h. Neo mRNA was detected by RT-qPCR. HA-nsp13 and GAPDH were detected by Western blotting. (E) HEK293T cells were co-transfected with pCMV-GFP and pHA-nsp13 or empty vector for 24 h. GFP fluorescence intensity was analyzed using Leica Application Suite X software (n = 10). Scale bar = 25 μm. (F) Left: HEK293T cells were transfected with nsp13 plasmid for 12, 24, or 36 h. Nsp13 mRNA was determined by RT-qPCR. Right: HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13, nsp13 plasmid, or empty vector for 24 h. F-Luc activity was determined. F-Luc, HA-nsp13, and GAPDH expression were determined by Western blotting. Data in all panels represent at least two independent experiments with similar results (mean ± standard error of the mean [SEM] of n = 3 biological replicates). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

SARS-CoV-2 nsp13 did not affect chromosomal gene expression.

One question raised was whether nsp13 could inhibit the expression of the same gene when it was chromosomally located. To this end, we transfected cells with HA-nsp13 plasmid or empty vector and stimulated the cells with or without SeV. As shown in Fig. 2A (left), SeV infection induced endogenous expression of RIG-I, which is known to be an interferon-stimulated gene (19). However, the levels of SeV-induced RIG-I were similar between cells with and without nsp13 overexpression (Fig. 2A, left). In contrast, when Flag-RIG-I was ectopically expressed by plasmid transfection, its expression was suppressed by nsp13 (Fig. 2A, right). These results suggested that nsp13 may specifically affect episomal gene expression. To further confirm this hypothesis, we generated the stable cell line HEK293T-Luc, in which the F-Luc gene driven by EF1α promoter was integrated into cell chromosomes by lentiviral vector transduction. As expected, nsp13 had no effect on the level of F-Luc expression in HEK293T-Luc stable cells, but inhibited F-Luc expression in cells co-transfected transiently with pEF1α-firefly luciferase plasmid (pEF1α-Luc) (Fig. 2B). To obtain an overview of the impact of nsp13 on cellular gene expression, we performed transcriptome sequencing (RNA-seq) to compare the transcriptomes of cells with and without nsp13 expression. We found that nsp13 had a limited effect on host gene expression. In particular, no host genes were found to be downregulated (Fig. 2C).

FIG 2.

The effect of SARS-CoV-2 nsp13 on chromosomal DNA. (A) Left: HEK293T cells were transfected with pHA-nsp13 or empty vector for 24 h and then infected with SeV for 8 h or not. RIG-I, HA-nsp13, and GAPDH were determined by Western blotting. Right: HEK293T cells were co-transfected with pFlag-RIG-I and pHA-nsp13 or empty vector for 24 h. Flag-RIG-I, HA-nsp13, and GAPDH expression were determined by Western blotting. (B) Left: F-Luc stable-expressing HEK293T cells were generated, in which the EF1α promoter driven F-Luc was randomly integrated into cell chromosomes by a lentiviral system. The cells were transfected with pHA-nsp13 or empty vector for 24 h. Firefly luciferase, HA-nsp13, and GAPDH expression were determined by Western blotting. Right: HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 or empty vector for 24 h. F-Luc, HA-nsp13, and GAPDH expression were determined by Western blotting. (C) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 or empty vector for 24 h. Total RNA was extracted and subjected to RNA-seq. Volcano plot shows that a total of 85 differentially expressed genes were identified by DESeq2. Data represent at least two independent experiments with similar results (mean ± SEM of n = 3 biological replicates in panel C).

Our results suggested that nsp13 can specifically repress episomal but not chromosomal gene expression, and this episomal suppression was not promoter sequence-specific.

SARS-CoV-2 nsp13 inhibited episomal DNA transcription.

Next, we wanted to gain more insight into the molecular mechanisms involved in nsp13 inhibition of episomal DNA expression. Because plasmid DNA commonly enters the nucleus for gene expression, we first examined whether nsp13 could affect the stability or mobility of episomal DNA in cells. A quantitative PCR (qPCR) assay and a Southern blot analysis (Fig. 3A and B) revealed that there were no differences in F-Luc plasmid levels in cells with or without nsp13. Next, we performed cytoplasm/nucleus fractionation to determine the location of the episomal DNA. The qPCR results showed that most of the transferred F-Luc plasmids reached the nucleus within 24 h and that this process was not mitigated by nsp13 (Fig. 3C).

FIG 3.

SARS-CoV-2 nsp13 restricts episomal DNA transcription. (A and B) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 or empty vector for 24 h. F-Luc plasmid level was determined by qPCR (A) and Southern blotting (B). (C) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 or empty vector for 24 h. F-Luc plasmid levels in the nuclear fractions or cytoplasmic fractions were determined by qPCR. (D and E) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 or empty vector for 24 h. F-Luc mRNA was quantified by RT-qPCR (D) and Northern blotting (E). (F) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 or empty vector for 20 h and then treated with Act D (10 μg/mL) for 2 or 4 h. F-Luc mRNA was quantified by RT-qPCR. (G) HEK293T cells were transfected with pHA-nsp13 or empty vector for 24 h, then transfected with GFP mRNA for 9 h. GFP mRNA was determined RT-qPCR. (H) HEK293T cells were transfected with pHA-nsp13 or empty vector for 24 h, then transfected with GFP mRNA. Left: After transfection for 9 h, GFP, HA-nsp13, and GAPDH expression was determined by Western blotting. Right: After transfection for 3, 6, or 9 h, GFP fluorescence intensity was determined. Scale bar = 100 μm. (I) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 or empty vector for 24 h, then treated with DMSO (dimethyl sulfoxide), MG132 (10 μM), or Bafilomycin A1 (100 nM) for 12 h. F-Luc, HA-nsp13, and GAPDH were detected by Western blotting. F-Luc expression was quantified by ImageJ software. The values from cells transfected with empty vector were set as 1. (J) A run-off transcription assay was used to detect the effect of HA-nsp13 on transcriptional elongation of RNA polymerase II on pEF1α-Luc expression plasmid. (K) Summary of SARS-CoV-2 nsp13 inhibition of episomal DNA transcription. Data represent at least two independent experiments with similar results (mean ± SEM of n = 3 biological replicates in panels A, C, D, F, G, and J). *, P < 0.05; **, P < 0.01; ns, not significant.

Following nucleus trafficking, the transferred DNA hijacks host cell machinery for transcription. We observed that in the presence of nsp13, the levels of F-Luc transcripts were markedly reduced compared to vector control. This was supported by both reverse transcription-quantitative PCR (RT-qPCR) and Northern blotting (Fig. 3D and E). Since decreased mRNA levels can result from blockage of de novo mRNA synthesis or degradation of existing mRNA, we sought to determine how nsp13 affected mRNA levels. Actinomycin D (Act D), which is known to inhibit RNA synthesis, was used to treat cells in the presence or absence of nsp13. Similar turnover curves were observed between these two groups, indicating that nsp13 did not affect the stability of mRNA transcribed from episomal DNA (Fig. 3F). In addition, when in vitro-transcribed GFP mRNA was transfected into nsp13-expressing cells, it maintained levels comparable to those in cells without nsp13 expression (Fig. 3G), further excluding the possibility that nsp13 might act on existing mRNA stability.

We then studied whether nsp13 inhibited episomal DNA expression by affecting translation. Accordingly, GFP mRNA was transfected into cells with or without nsp13 expression and translated GFP levels were examined by both Western blotting and immunostaining. Similar levels between different experimental groups were observed (Fig. 3H). To further discern whether nsp13 could promote protein degradation, we treated cells with MG132 or Bafilomycin A1 to block intracellular protein degradation or protein trafficking. These treatments did not change the level of inhibition of episomal F-Luc expression by nsp13 (Fig. 3I). These data indicated that nsp13 suppressed episomal DNA transcription. To further consolidate the role of nsp13 in episomal DNA transcription, we performed a runoff transcription assay and observed that nsp13 reduced F-Luc transcript levels in vitro (Fig. 3J). Taken together, these data proved that nsp13 acted specifically on the transcription step to reduce episomal DNA expression (Fig. 3K).

NTPase and helicase activities contributed to SARS-CoV-2 nsp13-mediated inhibition of episomal DNA transcription.

It has been reported that nsp13 contains five domains: ZBD, stalk, 1B, 1A, and 2A (2). To investigate which domains were involved in the repression of episomal DNA expression, we generated six plasmids expressing truncated nsp13, as illustrated in Fig. 4A. Cells were co-transfected with F-Luc plasmid and truncated nsp13-expressing plasmids. Firefly luciferase assays revealed that HA-nsp13 1–100 amino acids (aa), HA-nsp13 1–234 aa, and HA-nsp13 1–439 aa failed to inhibit luciferase activity. However, HA-nsp13 101–601 aa, HA-nsp13 235–601 aa, and HA-nsp13 440–601 aa closely resembled wild-type (wt) nsp13 and significantly inhibited luciferase activity (Fig. 4A). A Western blot analysis confirmed the expression of each nsp13 mutant as well as F-Luc expression in the same cells, which is consistent with the luciferase readout (Fig. 4B). Among all of these, HA-nsp13 235–601 aa exhibited the strongest inhibitory effect (Fig. 4A and B). These data collectively underlined the notion that the 1A/2A domains play essential roles in nsp13-mediated suppression of episomal DNA.

FIG 4.

NTPase and helicase activities contribute to SARS-CoV-2 nsp13 restriction of episomal DNA transcription. (A) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13, pHA-nsp13 1–100 aa, pHA-nsp13 1–234 aa, pHA-nsp13 1–439 aa, pHA-nsp13 101–601 aa, pHA-nsp13 235–601 aa, pHA-nsp13 440–601 aa, or empty vector for 24 h. F-Luc activity was determined. The values from cells co-transfected with pEF1α-Luc and empty vector were set as 100%. (B) Top: HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 1–100 aa, pHA-nsp13 1–234aa, pHA-nsp13 1–439 aa, pHA-nsp13, or empty vector for 24 h. F-Luc, HA-nsp13, HA-nsp13 truncated mutants, and GAPDH were subjected to Western blotting. Bottom: HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 101 to 601–aa, pHA-nsp13 235–601 aa, pHA-nsp13 440–601 aa, or empty vector for 24 h. F-Luc, HA-nsp13, HA-nsp13 truncated mutants, and GAPDH were subjected to Western blotting. F-Luc expression was quantified by ImageJ software. The values from cells co-transfected with pEF1α-Luc and empty vector were set as 1. (C) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13, pHA-nsp13 K288A, pHA-nsp13 S289A, pHA-nsp13 D374A, pHA-nsp13 E375A, pHA-nsp13 Q404A, pHA-nsp13 R567A, or empty vector for 24 h. F-Luc activity was determined. F-Luc, HA-nsp13, HA-nsp13 site mutants, and GAPDH were determined by Western blotting. F-Luc expression was quantified by ImageJ software. The values from cells co-transfected with pEF1α-Luc and empty vector was set as 1. (D) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13, pHA-nsp13 D374A, pHA-nsp13 Q404A, or empty vector for 24 h. F-Luc mRNA was quantified by RT-qPCR. (E) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13, pHA-nsp13 235–601 aa, pHA-nsp13 D374A, or empty vector for 24 h. F-Luc mRNA expression was determined by Northern blotting. Data represent at least two independent experiments with similar results (mean ± SEM of n = 3 biological replicates in panels A, C, and D). ****, P < 0.0001; ns, not significant.

The 1A/2A domains are “RecA-like” helicase domains which have NTPase and helicase activities (11). Our next step was to explore whether the observed transcriptional inhibitory effect relied on the known enzymatic activities of 1A/2A. It has been reported that six residues, including K288, S289, D374, E375, Q404, and R567, are crucial for nsp13 to perform its NTP hydrolysis and helicase activities (12). We then constructed nsp13 mutants with the corresponding amino acid mutations and tested their abilities to inhibit episomal gene expression. As shown by both Western blotting and luciferase measurement, all of those mutants exhibited impaired inhibition of F-Luc plasmid expression compared with wt nsp13 (Fig. 4C). Among these, the D374 and Q404 mutants completely lost their activities, as evidenced by similar levels of mRNA expression in the cells with no nsp13 expression (Fig. 4D). In addition, Northern blotting confirmed that F-Luc mRNA accumulation was significantly reduced by wt nsp13 and 1A/2A domains, but not by D374A mutant nsp13 (Fig. 4E). Taken together, our results revealed that the NTPase and helicase activities of SARS-CoV-2 nsp13 were indispensable for restricting episomal DNA transcription.

DNA interaction was required for nsp13 to suppress episomal DNA.

Given that its helicase activity is essential for nsp13-mediated blockage of episomal gene expression, we reasoned that nsp13 must bind with the target DNA. It has been reported that six nsp13 double mutants, N179A/R212A, R337A/R339A, K345A/K347A, R507A/K508A, K524A/Q531A, and S539A/Y541A, displayed different degrees of nucleic acid-binding deficiency. Among these, the mutants K345A/K347A exhibited the lowest DNA-binding capacity (11). To investigate whether nsp13 requires direct interaction with episomal DNA to function, we examined cells transfected with each of these six nsp13 double mutants. A luciferase assay showed that among the tested mutants, K345A/K347A clearly exhibited defective suppression of F-Luc plasmid expression compared to wt nsp13 (Fig. 5A, left panel). Western blot analysis further revealed that while most of the tested nsp13 mutants exerted similar levels of suppression of luciferase expression compared to wt nsp13, the nsp13 mutants had much higher expression levels than wt nsp13 (Fig. 5A, right panel), further suggesting that these DNA binding-defective mutants were less effective at abrogating episomal gene expression. Additionally, a chromatin immunoprecipitation (ChIP) assay using cells transiently transfected with F-Luc plasmid revealed that wt nsp13 was enriched on F-Luc plasmid, but K345A/K347A binding was not detectable. RNA Pol II was used as a positive control (Fig. 5B). In contrast, when cells constructed with F-Luc integrated into the chromosome were used in the ChIP assay, nsp13 binding was not observed. These results confirmed the specificity of nsp13 for inhibiting episomal DNA transcription (Fig. 5C). In conclusion, our results demonstrated that the binding of SARS-CoV-2 nsp13 to episomal DNA was a prerequisite for inhibiting gene transcription.

FIG 5.

Restriction activity of SARS-CoV-2 nsp13 requires interaction with episomal DNA. (A) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13, pHA-nsp13 N179A/R212A, pHA-nsp13 R337A/R339A, pHA-nsp13 K345A/K347A, pHA-nsp13 R507A/K508A, pHA-nsp13 K524A/Q531A, pHA-nsp13 S539A/Y541A, or empty vector for 24 h. F-Luc activity was determined. F-Luc, HA-nsp13, HA-nsp13 site mutants, and GAPDH were subjected to Western blotting. F-Luc expression was quantified by ImageJ software. The values from cells co-transfected with pEF1α-Luc and empty vector were set as 1. (B) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13, pHA-nsp13 K345A/K347A, or empty vector for 36 h, and then a ChIP assay was performed with anti-RNA polymerase II-specific antibody or anti-HA antibody. (C) F-Luc stable-expressing HEK293T cells were transfected with pHA-nsp13, then a ChIP assay with anti-HA antibody was performed. Data represent at least two independent experiments with similar results (mean ± SEM of n = 3 biological replicates in each panel). ***, P < 0.001; ****, P < 0.0001; ns, not significant.

Development of a high-throughput drug-screening system targeting nsp13.

Nsp13 plays a key role in SARS-CoV-2 replication, making it a potential target for antiviral drug development. Based on the helicase function of nsp13 is correlated with its inhibitory effect on episomal DNA, we developed a new high-throughput drug-screening method. As shown in Fig. 6A, cells were co-transfected with SAR-CoV-2 nsp13 and luciferase reporter plasmids and seeded in a 96-well plate. Each well was treated with a certain compound. Potential nsp13-inhibiting compounds can result in increased luciferase activity. To test the reliability of this system, we treated transfected cells with bismuth salt, which is known to inhibit the NTPase and helicase activities of nsp13 (3). As expected, the luciferase assays showed that bismuth salt successfully rescued F-Luc expression in the presence of nsp13 in a dose-dependent manner (Fig. 6B), which was further confirmed by Western blotting (Fig. 6C).

FIG 6.

Development of a high-throughput drug-screening method targeting nsp13. (A) Schematic of the drug-screening platform. (B) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 or empty vector for 8 h, then treated with bismuth subcitrate (BAC) (50, 100, and 500 μM) for 24 h. F-Luc activity was determined. Data represent at least two independent experiments with similar results. (C) HEK293T cells were co-transfected with pEF1α-Luc and pHA-nsp13 or empty vector for 8 h, then treated with BAC (50, 100, and 500 μM) for 24 h. F-Luc, HA-nsp13, and GAPDH were subjected to Western blotting. (D) Graphical summary of the study.

Collectively, we showed that nsp13 could specifically suppress episomal DNA transcription. The inhibition of episomal DNA required the NTPase and helicase activities of nsp13 as well as direct binding to episomal DNA. Based on these findings, we developed a high-throughput drug-screening system targeting nsp13. This approach will facilitate the development of novel antiviral drugs against SARS-CoV-2 (Fig. 6D).

DISCUSSION

Previous studies reported that nsp13 possessed NTPase (3), DNA/RNA helicase (4, 20), and RNA 5′-triphosphatase activities (7). Nsp13 has also been shown to be able to interfere with the host innate immune response (9, 21). In the present study, we found a novel function of nsp13 in repressing episomal DNA expression. Although the plasmids used in this study harbored various promoters and coding cassettes, they were universally inhibited by nsp13, which precludes sequence-specific recognition by nsp13. We tried to confirm the effect of nsp13 on episomal DNA in a live virus infection. However, SARS-CoV-2 infection caused cytopathic effects in cell cultures, leading to a loss of host cell viability as previously reported (22), which perturbed our analysis. Currently, the ectopic expression of nsp13 alone is a better way to demonstrate its role in episomal DNA transcription.

How nsp13 selectively affects the episomal gene but ignores the chromosomal gene of the same sequence is an interesting question. Through literature studies, we found that the chromosome maintenance complex SMC5/6 suppresses the transcription of episomal DNA, including hepatitis B virus (HBV) episomal and reporter plasmids (23, 24). Fabien et al. further explored the molecular mechanism of SMC5/6 inhibition and demonstrated that the specific recognition of episomal DNA by SMC5/6 depends on its ability to bind episomal DNA, but not chromosomal DNA; they also found that Nse4a has essential roles in the early steps before SMC5/6 traps episomal DNA (25). In the present study, we demonstrated that nsp13 could bind to episomal DNA, but not to a chromosomally integrated foreign DNA. Therefore, we hypothesized that nsp13 binds episomal DNA with the involvement of host factors such as Nse4a, then specifically recognizes episomal DNA.

In the present study, many efforts were made to explore the molecular mechanisms involved in the transcriptional repression ability of nsp13. By using different truncated and mutant forms of nsp13, we demonstrated that its NTPase and helicase activities contributed to the inhibition of episomal DNA transcription. We also investigated whether inhibition by nsp13 requires binding to episomal DNA by mutating nsp13 at different DNA-binding sites. K345/K347 are nsp13 double-stranded DNA (dsDNA)-binding sites located in the helicase domain (residues 235 to 601). Compared to other dsDNA-binding site mutants, K345A/K347A exhibited the lowest dsDNA-binding capacity and helicase activity, but it was still able to weakly bind and unwind dsDNA (12), and this weak binding might not be detectable by ChIP-qPCR. Therefore, K345A/K347A only weakened suppression of episomal DNA expression compared to wt nsp13, but it did not lose the activity to inhibit episomal DNA expression. We also found that the nsp13 DNA-binding site mutant S539/Y541 inhibited episomal DNA more strongly than wt nsp13, possibly because the S539/Y541 mutation could affect the structure of nsp13. Several studies have demonstrated that nsp13 can bind and unwind dsDNA (4) and plays a role in RNA polymerase back-tracking, RNA polymerase template switching, and disruption of downstream secondary structures (1, 6). In conclusion, we speculate that nsp13 inhibits the expression of episomal DNA by binding and unwinding episomal DNA and further affecting the movement of RNA polymerase II.

Nsp13 is essential for viral replication and highly conserved across related viruses, making it an attractive antiviral target. In recent years, researchers have tried to find inhibitors of nsp13 helicase (3, 4, 11, 13, 15–17). They have examined the inhibitory effects of compounds against nsp13 activity by using a fluorescence resonance energy transfer (FRET)-based dsDNA unwinding assay or a colorimetry-based ATP hydrolysis assay. These screening methods require purification of the nsp13 protein for experiments that are relatively time-consuming and labor-intensive. Additionally, these screens are not performed in cells and do not truly reflect nsp13 helicase activity. In comparison, our high-throughput screening method utilized compounds to treat cells expressing SAR-CoV-2 nsp13 and luciferase reporter plasmid. The upregulation of luciferase activity resulting from compound treatment could be a good proxy for the compound’s inhibitory effect on nsp13 activity. To make our high-throughput screening method more convenient, we tried to construct nsp13 stably expressed cell lines with a lentiviral system, but nsp13 could inhibit the expression of lentiviral helper plasmids, making lentiviral vector production insufficient. Alternative methods to establish nsp13-expressing cell lines should be considered.

MATERIALS AND METHODS

Cell lines and cultures.

The HEK293T (CRL-3216) cell line was obtained from the ATCC. HEK293T cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; cat no. C11995500BT, Gibco, USA) supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin sulfate and cultured at 37°C in 5% CO₂.

Plasmids, mRNA, and chemicals.

The sequence of SARS-CoV-2 nsp13 was amplified from a pET32a(+)-nsp13 plasmid, kindly provided by Yan Li (Huazhong University of Science and Technology), and cloned into mammalian expression vectors pXJ40-HA (N-terminal HA-tag, CMV promoter) and pXJ40 (no tag, CMV promoter). The mutant SARS-CoV-2 nsp13 sequence was inserted into pXJ40-HA (N-terminal HA-tag). The IAV NS1 plasmid was kindly provided by Ke Xu (Wuhan University) (26). The IFN-β promoter firefly luciferase reporter plasmid (pGL3-IFN-β-Luc), HSV-TK promoter Renilla luciferase reporter plasmid (pRL-TK), pcDNA3.1-Flag-RIG-IN, pcDNA3.1-Flag-MDA5, pcDNA3.1-Flag-VISA, pcDNA3.1-Flag-TBK1, and pcDNA3.1-Flag-IKKε have been described elsewhere (18). pEGFP-C1 (CMV promoter) plasmid and pEF1α-Luc (EF1α promoter) were stored in our lab. The PCR primers used in this study are listed in Table 1. The GFP mRNA was generated by the T7 mScript Standard mRNA Production System (cat no. C-MSC100625, CELLSCRIPT, Madison, WI, USA). The chemicals used in this study are listed in Table 2.

TABLE 1.

Primer sequences for PCR

HA-nsp13 oligonucleotide	Primer sequence
	Forward	Reverse
HA-nsp13	CCGCTCGAGATGGCGGTGGGTGCGTGCGTTCTGTG	ATAGTTTAGCGGCCGCTCACTGCAGGGTCGCAACG
1–100 aa	CCGCTCGAGATGGCGGTGGGTGCGTGCGTTCTGTG	TTTTCCTTTTGCGGCCGCCTAGCTACCCACGCAGG
1–234 aa	CCGCTCGAGATGGCGGTGGGTGCGTGCGTTCTGTG	TTTTCCTTTTGCGGCCGCCTACGGCATCACGGTGT
1–439 aa	CCGCTCGAGATGGCGGTGGGTGCGTGCGTTCTGTG	TTTTCCTTTTGCGGCCGCCTAACCCAGAAACATAT
101–601 aa	CCGCTCGAGAACGTTACCGATTTTAACGCGATTGC	ATAGTTTAGCGGCCGCTCACTGCAGGGTCGCAACG
235–601 aa	CCGCTCGAGAGCGCGCCGACCCTGGTTCCGCAGGA	ATAGTTTAGCGGCCGCTCACTGCAGGGTCGCAACG
440–601 aa	CCGCTCGAGTGCCGTCGTTGCCCGGCGGAAATTGT	ATAGTTTAGCGGCCGCTCACTGCAGGGTCGCAACG
K288A	GGTCCGCCGGGTACCGGCGCGAGCCACTTT	GCGCCGGTACCCGGCGGACCTTGCAGGGTG
S289A	CCGCCGGGTACCGGCAAGGCCCACTTTGCG	GCCTTGCCGGTACCCGGCGGACCTTGCAGG
D374A	GCGGACATCGTGGTTTTTGCTGAGATTAGC	GCAAAAACCACGATGTCCGCGGTGGTTTCC
E375A	GACATCGTGGTTTTTGATGCGATTAGCATG	GCATCAAAAACCACGATGTCCGCGGTGGTT
Q404A	TATATTGGTGACCCGGCGGCACTGCCGGCG	GCCGCCGGGTCACCAATATAAACGTAGTGC
R567A	TTTAACGTTGCGATTACCGCTGCGAAGGTT	GCGGTAATCGCAACGTTAAAACGGTTCACG
N179A	CCGCCGCTGAACCGTGCCTACGTGTTCACC	GGCACGGTTCAGCGGCGGACGCGGCTTGCC
N179A/R212A	GATGCGGTGGTTTATGCTGGTACCACCACC	AGCATAAACCACCGCATCGCCGTAGTCACC
R337A/R339A	ATCATTCCGGCGGCTGCGGCTGTTGAATGCTTCGA	AGCCGCAGCCGCCGGAATGATACGGCTGCATTTGT
K345A/K347A	GAATGCTTCGACGCGTTTGCAGTGAACAGCACCCT	TGCAAACGCGTCGAAGCATTCAACACGCGCACGCG
R507A/K508A	CGTAACCCGGCGTGGGCTGCGGCGGTTTTTATCAG	CGCAGCCCACGCCGGGTTACGGGTCAGGAACTCAC
K524A	CAGAACGCGGTGGCGAGCGCAATTCTGGGT	TGCGCTCGCCACCGCGTTCTGGCTGTTATA
Q531A	ATTCTGGGTCTGCCGACCGCGACCGTTGAT	CGCGGTCGGCAGACCCAGAATTGCGCTCGC
S539A/Y541A	AGCAGCCAAGGCGCCGAAGCCGACTATGTGATCTT	GGCTTCGGCGCCTTGGCTGCTATCAACGGTCTGGG

TABLE 2.

Chemicals

Chemical	Manufacturer	Catalog no.
PEI MAX Transfection Grade Linear Polyethylenimine Hydrochloride (MW 40,000)	Polysciences (Hirschberg an der 129 Bergstrasse, Germany)	24765-1
Lipofectamine 2000	Invitrogen (Carlsbad, CA, USA)	11668019
Actinomycin D	Selleck (Houston, TX, USA)	50-76-0
MG132	Sigma-Aldrich (St. Louis, MO, USA)	M7449
Bafilomycin A1	MedChemExpress (St. Louis, MO, USA)	HY-100558
Bismuth subcitrate	MedChemExpress	HY-B0796
NTP set	Thermo Fisher Scientific (Carlsbad, CA, USA)	R0481
Hoechst 33258	Invitrogen	H1398

Luciferase reporter assay.

(i) IFN-β promoter luciferase assay. HEK293T cells, seeded in a 24-well plate at a density of 1 × 10⁵ cells per well, were co-transfected with 100 ng of IFN-β promoter firefly luciferase reporter plasmid (pIFN-β-luc), 10 ng Renilla luciferase control plasmid (pRL-TK, normalizing transfection efficiency), and 50 to 200 ng of protein expression plasmid using PEI MAX (24765-1, Polysciences, Warrington, PA, USA) transfection reagent at a 1:3 ratio. Empty pXJ40-HA vector was used to ensure the same total amount (310 ng) of plasmids in each well. HEK293T cells were activated by co-transfection with 100 ng stimulator-expressing plasmids (RIG-IN, MDA5, VISA, TBK1, or IKKε) for 24 h or stimulated with SeV (MOI = 100) for 10 h. Luciferase activity was measured using a Dual-Luciferase Reporter Assay System kit (cat no. E1960, Promega, San Luis Obispo, CA, USA) according to the manufacturer’s instructions. Data are shown as relative IFN-β firefly luciferase activity after normalization to Renilla luciferase signals.

(ii) Luciferase assay. HEK293T cells, seeded in a 24-well plate at a density of 1 × 10⁵ cells per well, were co-transfected with firefly luciferase plasmid (pEF1α-Luc) and protein expression plasmid or empty vector plasmid using PEI MAX transfection reagent at a ratio of 1:3. At 24 h post-transfection, firefly luciferase activity was measured using a Dual-Luciferase Reporter Assay System kit (cat no. E1960, Promega) according to the manufacturer’s instructions. The protein expression of nsp13-truncated mutants such as 101–601 aa, 235–601 aa, and 440–601 aa was significantly different. To better study the effects of these mutations on episomal DNA expression, we adjusted the quality of the transfected nsp13 mutant plasmids to normalized protein expression. The nsp13 101–601 aa plasmid was 400 ng, the nsp13 235–601 aa plasmid was 100 ng, and the nsp13 440–601 aa plasmid was 500 ng, and pXJ40-HA vector was used to ensure the same total amount of transfected plasmids in each well.

Immunofluorescence microscopy.

In a 12-well plate, HEK293T cells at 40% to 50% confluence were transfected with the indicated plasmids (1,000 ng) for 24 h, then cells were washed twice with phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin (BSA) and fixed in 4% paraformaldehyde at room temperature for 10 min. Following permeabilization with 0.1% Triton X-100 buffer for 5 min, cells were washed three times with PBS containing 0.1% BSA, then blocked with PBS containing 5% BSA for 1 h. The cells were incubated with the primary antibody overnight at 4°C, followed by incubation with DyLight 488-conjugated anti-rabbit IgG and DyLight 568-conjugated anti-mouse IgG for 1 h. After three washes with PBS (containing 0.1% BSA), the cells were counterstained with Hoechst-33258 solution (Invitrogen, Waltham, MA, USA) for 5 min and then washed three more times with PBS. Finally, the cells were analyzed using a confocal laser scanning microscope (Leica TCS SP8 STED, Leica, Germany). The antibodies used are listed in Table 3.

TABLE 3.

Antibodies^a

Antibody	Source	Catalog no.	Dilution(s)
Anti-HA	Sigma-Aldrich (St. Louis, MO, USA)	H6908	WB, 1:5,000
			ChIP, 1:100
			IP, 1:200
			IF, 1:1,000
Anti-HA	Beyotime Biotechnology (Shanghai, China)	AH158	WB, 1:1,000
Anti-Flag	Sigma-Aldrich	F1804	WB, 1:1,000
Anti-firefly luciferase	Proteintech (China)	27986-1-AP	WB, 1:1,000
Anti-RNA Pol II	Beyotime Biotechnology	AF7788	WB, 1:1,000
Anti-RNA Pol II	This antibody was kindly provided by Kaiwei Liang (Wuhan University) (PMID: 34400476)	NA	ChIP, 1:50
Anti-GAPDH	Sigma-Aldrich	G9295	WB, 1:10,000
Anti-Amin A/C	Beyotime Biotechnology	AF7350	WB, 1:1,000
Anti-mouse IgG	Invivogen (San Diego, CA, USA)	mabg1-ctrlm	CHIP, 1:200
Anti-rabbit IgG	Cell Signaling Technology (Danvers, MA, USA)	2729S	CHIP, 1:100
Anti-mouse IgG (HRP-linked)	Cell Signaling Technology	7076S	WB, 1:5,000
Anti-rabbit IgG (HRP-linked)	Cell Signaling Technology	7074S	WB, 1:5,000
Anti-mouse IgG (Alexa Fluor 568)	Invitrogen (Waltham, MA, USA)	A-11031	IF, 1:1,000
Anti-rabbit IgG (Alexa Fluor 488)	Invitrogen	A-11034	IF, 1:1,000

^aWB, Western blotting; IP, immunoprecipitation; IF, immunofluorescence; HRP, horseradish peroxidase.

RT-qPCR, qPCR, and RNA-seq.

Total cellular RNA was extracted by an RNApure Tissue & Cell kit (Cowin Biotech, China), followed by cDNA preparation using ReverTra Ace qPCR RT Master Mix (Toyobo Co., Ltd.; Osaka, Japan) according to the manufacturer’s instructions. RNA extracted from the RNApure Tissue & Cell kit was treated with DNase I and was directly used for RT-qPCR and Northern blotting. Residual DNA was removed by gDNA remover from ReverTra Ace qPCR RT Master Mix. Total intracellular genomic DNA was extracted using a TIANamp Genomic DNA kit (cat no. DP304-03, Tiangen Biotech, Beijing, China) according to the manufacturer’s instructions. RT-qPCR/qPCR was performed using the Roche LC480 and FastStart Essential DNA Green Master (cat no. 06924204001, Roche, Basel, Switzerland). The reaction mixture contained 5 μL SYBR Green PCR Master Mix, 4 μL diluted DNA template, and 1 μL of the corresponding primers. The RT-qPCR/qPCR primers used are listed in Table 4. The RNA samples were sent to Annoroad Gene Technology company (Beijing, China) for RNA-seq analysis. Briefly, a total of 2 μg RNA per sample was used as input material for RNA sample preparation. Sequencing libraries were generated using the NEBNext Ultra RNA Library Prep kit for Illumina (cat no. E7530L, New England Biolabs, Ipswich, MA, USA) following the manufacturer’s recommendations, and index codes were added to attribute sequences to each sample. mRNA was purified from total RNA using poly(T) oligonucleotide-attached magnetic beads. The cDNA library was generated and then purified (AMPure XP system). The clustering of the index-coded samples was performed on a cBot Cluster Generation System using HiSeq PE Cluster kit v4-cBot-HS (Illumina) according to the manufacturer’s instructions. After cluster generation, the library preparations were sequenced on an Illumina platform and 150-bp paired-end reads were generated.

TABLE 4.

Primer sequences for RT-qPCR/qPCR

Oligonucleotide	Forward primer sequence	Reverse primer sequence
Nsp13	CTGCGACTGGACCAACGCGG	ACCGGTGAACACGTAGTTAC
Firefly luciferase	CTATTCTCCTTCTTCGCCAA	TATCCAGATCCACAACCTTC
GFP	CCCGACAACCACTACCTGAG	GTCCATGCCGAGAGTGATCC
GAPDH	GACAAGCTTCCCGTTCTCAG	GAGTCAACGGATTTGGTGGT

Southern/Northern blotting.

The Random Primer DNA Labeling kit v2 (TaKaRa Bio, Tokyo, Japan) and Northern Max kit (Invitrogen) were used in these experiments. Briefly, for Southern blotting, episomal DNA was extracted following a Hirt protein-free DNA extraction procedure, and 20 μg DNA was subjected to a Southern blot assay as described previously (27, 28). For Northern blotting, the cultured cells and liver tissues were assayed according to the manufacturer’s instructions. The total cellular RNA was extracted by TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. A total of 20 μg RNA was used for the Northern blot assay according to the manufacturer’s protocols (Invitrogen). For the mouse liver tissues, 50-mg tissue samples were minced by a Tissue Cell-Destroyer (DS1000, NovaStar, China) and homogenized in TRIzol reagent, and 30 μg RNA was used for the Northern blot assay according to the manufacturer’s protocols. The hybridization signal was collected by a Typhoon FLA 9500 imager (GE Healthcare Lifesciences).

Western blot.

Cell lysates were prepared with lysis buffer (50 mM Tris-HCl [pH 7.5], 300 mM NaCl, 1% Triton X-100, 5 mM EDTA, and 10% glycerol). After protein concentration was measured by a Bradford assay (500-0205, Bio-Rad, Hercules, CA), 25 μg of cell lysates was electrophoresed in 10% SDS-PAGE gel and transferred to a polyvinylidene difluoride (PVDF) membrane (cat no. IPVH00010, Millipore, Burlington, MA, USA). PVDF membranes were blocked with 5% skim milk in Tris-buffered saline with 0.05% Tween 20 (TBST) and incubated with specific primary and secondary antibodies. Protein bands were detected using a Genegome XRQ chemiluminescence imaging system (Genegome XRQ-NPC, China). The antibodies used are listed in Table 3.

Chromatin immunoprecipitation assay.

The ChIP assay was performed as described previously (28, 29). Briefly, 1 × 10⁷/mL cells were fixed in 1% formaldehyde (Biosharp, Hefei, China) for 20 min at room temperature, and 125 mM glycine (BioFroxx, China) was added to quench formaldehyde. After a 30-min incubation in ChIP lysis buffer (50 mM Tris-HCl [pH 8.0], 5 mM EDTA, 150 mM NaCl, 1% NP-40, 0.1% SDS, protease inhibitor) on ice, the lysates were sheared by sonication using a Bioruptor Plus (Diagenode, Liège, Belgium). Cross-linked chromatin samples were incubated with the indicated antibodies or normal rabbit IgG in a rotator at 4°C overnight. Subsequently, protein A/G-conjugated agarose beads (Smart-Lifesciences, Changzhou, China) were added, and the samples were incubated overnight at 4°C in a rotator, collected, and washed three times. To elute DNA fragments, immunocomplexes were incubated with elution buffer (50 mM Tris-HCl [pH 8.0], 10 mM EDTA, 1.0% SDS) for 2 h at 65°C. One half of the eluted immunocomplexes was saved as the immunoprecipitation sample, while the other half was treated with proteinase K (Tiangen Biotech) overnight at 55°C. Finally, the DNA was purified with a TIANamp Genomic DNA kit (Tiangen Biotech). The purified DNA was detected by qPCR. The antibodies used for ChIP and the primers used for qPCR are listed in Tables 3 and 4.

Run-off transcription.

HEK293T cells were transfected with pHA-nsp13 or vector for 48 h. Cell transcription mixture production was performed as described previously (30). Run-off transcription was performed as described previously (31). Briefly, the following components were assembled on ice in a 1.5-mL sterile siliconized microcentrifuge tube: 90 μg of HEK293T cell transcription mixture, 6 μL of transcription buffer (20 mM HEPES [pH 7.9], 100 mM KCl, 0.2 mM EDTA, 0.5 mM dithiothreitol, 20% [vol/vol] glycerol), 1.5 μL of 50 mM MgCl₂, 1 μL of 25-μM ribonucleotide mixture, and 0.5 μL (100 ng) of plasmid templates. Nuclease-free water was added for a final volume of 25 μL after addition of the nuclear extract. After incubation at 30°C for 45 min, transcription was stopped by adding 175 μL of stop buffer (0.3 M Tris HCl [pH 7.4], 0.3 M sodium acetate, 0.5% SDS, 2 mM EDTA). The RNA sample was analyzed by RT-qPCR.

Statistical analyses.

Statistical significance was analyzed by a one-way analysis of variance with Dunnett’s multiple-comparison test or t test using GraphPad Prism 8.

Data availability.

All raw and processed sequencing data generated in this study have been submitted to the NCBI Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) under accession no. GSE211851.