Establishment of a Rapid Detection System for ISG20-Dependent SARS-CoV-2 Subreplicon RNA Degradation Induced by Interferon-α

Inhaled nebulized interferon (IFN)-α and IFN-β have been shown to be effective in the management of coronavirus disease 2019 (COVID-19). We aimed to construct a virus-free rapid detection system for high-throughput screening of IFN-like compounds that induce viral RNA degradation and suppress the replication of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). We prepared a SARS-CoV-2 subreplicon RNA expression vector which contained the SARS-CoV-2 5′-UTR, the partial sequence of ORF1a, luciferase, nucleocapsid, ORF10, and 3′-UTR under the control of the cytomegalovirus promoter. The expression vector was transfected into Calu-3 cells and treated with IFN-α and the IFNAR2 agonist CDM-3008 (RO8191) for 3 days. SARS-CoV-2 subreplicon RNA degradation was subsequently evaluated based on luciferase levels. IFN-α and CDM-3008 suppressed SARS-CoV-2 subreplicon RNA in a dose-dependent manner, with IC50 values of 193 IU/mL and 2.54 μM, respectively. HeLa cells stably expressing SARS-CoV-2 subreplicon RNA were prepared and treated with the IFN-α and pan-JAK inhibitor Pyridone 6 or siRNA-targeting ISG20. IFN-α activity was canceled with Pyridone 6. The knockdown of ISG20 partially canceled IFN-α activity. Collectively, we constructed a virus-free rapid detection system to measure SARS-CoV-2 RNA suppression. Our data suggest that the SARS-CoV-2 subreplicon RNA was degraded by IFN-α-induced ISG20 exonuclease activity.


Introduction
Coronavirus disease (COVID-19) is a worldwide pandemic, and the development of therapeutic agents that specifically target severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) molecules is long awaited. Although the pandemic has been suppressed by the development of vaccines, it persists, and there are concerns regarding the emergence of mutant strains that weaken vaccine efficacy. In recent years, new coronaviruses, such as Middle East respiratory syndrome coronavirus (MERS-CoV), severe acute respiratory syndrome coronavirus (SARS-CoV), and SARS-CoV-2, have emerged at a frequency of approximately once per decade [1,2]. The development of therapeutic agents that are effective against any coronavirus is important. Therefore, we focused on small chemical compounds that enhance the degradation of SARS-CoV-2 genomic RNA.
It has been reported that IFN signaling is suppressed by the nsp1, nsp6, nsp13, and ORF6 proteins of SARS-CoV-2 [8], and analysis of severe COVID-19 patients has shown that autoantibodies against IFNs are produced, which weakens their effect [9,10]. Additionally, mutations in IRF9, IFNAR1, and IFNAR2 were observed in severe patients, which implies weakened IFN signaling and may have contributed to disease severity due to the failure to suppress SARS-CoV-2 replication [11]. A genome-wide association study (GWAS) of severe COVID-19 patients showed a significant association between disease severity and decreased IFNAR2 expression levels [12]. The therapeutic efficacy of IFN-α and IFN-β has been debated, and it has been reported that IFN-β1a has no therapeutic effect when administered subcutaneously or intravenously [13]. Meanwhile, inhaled nebulized IFN-α2b or IFN-β1a has a therapeutic effect and accelerates recovery from COVID-19 [14][15][16]. A phase II clinical trial using inhaled nebulized IFN-β1a (SNG001) for the treatment of COVID-19 showed that the SNG001-treated group of patients recovered more rapidly than the placebo group [14]. Moreover, a phase III clinical trial of SNG001 (NCT04732949) is ongoing, and it will confirm whether SNG001 can accelerate the recovery of hospitalized patients receiving oxygen. Therefore, IFN-α and IFN-β are useful for the treatment of COVID-19, but they cannot exert their anti-SARS-CoV-2 activity without supplying sufficient amounts to respiratory epithelial cells to induce ISG expression.
Moreover, to perform experiments using SARS-CoV-2, we need to use a biosafety level 3 facility (BSL-3), which is not suitable for the large-scale screening of small chemical compounds. The replicon system has been shown to be a powerful tool for screening small chemical compounds [17,18]. SARS-CoV and SARS-CoV-2 replicons have been constructed, and anti-SARS-CoV-2 drugs have been confirmed by activity [19][20][21][22].
Therefore, we developed a virus-free method to measure the degradation of SARS-CoV-2 RNA by IFN-α within a short interval. Using this method, we were able to screen new compounds that induce the degradation of SARS-CoV-2 genomic RNA.

Results
SARS-CoV-2 genomic RNA forms a stem-loop containing double-stranded RNA regions at the 5 -end and 3 -end, similar to SARS-CoV and MARS-CoV ( Figure 1A) [23,24]. IFN-α binds to heterodimers of IFNAR1 and IFNAR2 and induces the expression of ISGs. Among the ISGs, OAS1, 2, and 3 and ISG20 are thought to modify SARS-CoV-2 genomic RNA and induce its degradation [25]. In this study, we constructed a system for measuring the SARS-CoV-2 genomic RNA degradation activity of IFN-α within 1 day.
The production of SARS-CoV and SARS-CoV-2 replicons has also been reported [20,26]. Since the RNA genome of SARS-CoV-2 is approximately 30 kb, it is difficult to prepare a replicon that stably expresses the full-length viral RNA in cultured cells. Therefore, it is considered unsuitable as a screening system. To specialize in SARS-CoV-2 genomic RNA degradation induced by IFN-α, we focused on virus-specific stem-loops containing double-stranded RNA structures at the 5 -and 3 -ends ( Figure 1A). The minimal sequence required for coronaviruses' genomic RNA replication has been shown in previous studies [27], and based on coronaviruses' minimal sequence, we constructed a SARS-CoV-2 subreplicon RNA expression vector. SARS-CoV-2 subreplicon RNA contains 5 -UTR, a part of ORF1a, the luc2 gene, nucleocapsid, ORF10, and 3 -UTR and is expressed under the control of the CMV promoter. The total length obtained was 3.7 kb, and it was possible to insert it into an expression vector under the control of a CMV promoter. In addition, the expression vector was ≤10 kb, and stable cell lines were easily produced. After the transcription of SARS-CoV-2 subreplicon RNA, the fusion protein containing a part of ORF1, luciferase, and nucleocapsid was transcribed ( Figure 1B). The expression levels of SARS-CoV-2 subreplicon RNA were measured using a luciferase assay. In addition, SARS-CoV-2 subreplicon RNA replication could be achieved by adding RNA-dependent RNA polymerase to this sequence. sequence required for coronaviruses' genomic RNA replication has been shown in previous studies [27], and based on coronaviruses' minimal sequence, we constructed a SARS-CoV-2 subreplicon RNA expression vector. SARS-CoV-2 subreplicon RNA contains 5′-UTR, a part of ORF1a, the luc2 gene, nucleocapsid, ORF10, and 3′-UTR and is expressed under the control of the CMV promoter. The total length obtained was 3.7 kb, and it was possible to insert it into an expression vector under the control of a CMV promoter. In addition, the expression vector was ≤10 kb, and stable cell lines were easily produced.
After the transcription of SARS-CoV-2 subreplicon RNA, the fusion protein containing a part of ORF1, luciferase, and nucleocapsid was transcribed ( Figure 1B). The expression levels of SARS-CoV-2 subreplicon RNA were measured using a luciferase assay. In addition, SARS-CoV-2 subreplicon RNA replication could be achieved by adding RNA-dependent RNA polymerase to this sequence.  2 subreplicon RNA expression vector was transfected into Calu-3 cells. After 1 day of transfection, the cells were treated with IFN-α or the IFNAR2 agonist CDM-3008 for 3 days, and cell viability and luminescence intensity were measured ( Figure 2A). As a result, the IC 50 of IFN-α was 193 IU/mL, and the CC 50 was more than 10 µM ( Figure 2B), while the IC 50 of CDM-3008 was 2.54 µM, without showing cytotoxicity ( Figure 2C). Therefore, similar to IFN-α, CDM-3008 showed anti-SARS-CoV-2 activity through the degradation of SARS-CoV-2 genomic RNA.
Thus, we selected Calu-3 cells derived from lung adenocarcinoma, and the SARS-CoV-2 subreplicon RNA expression vector was transfected into Calu-3 cells. After 1 day of transfection, the cells were treated with IFN-α or the IFNAR2 agonist CDM-3008 for 3 days, and cell viability and luminescence intensity were measured ( Figure 2A). As a result, the IC50 of IFN-α was 193 IU/mL, and the CC50 was more than 10 μM ( Figure 2B), while the IC50 of CDM-3008 was 2.54 μM, without showing cytotoxicity ( Figure 2C). Therefore, similar to IFN-α, CDM-3008 showed anti-SARS-CoV-2 activity through the degradation of SARS-CoV-2 genomic RNA. The growth rate of Calu-3 cells was not higher than that of HeLa cells, and Calu-3 cells were not used for electroporation. Thus, we changed the Calu-3 cells to HeLa cells to construct a screening system. The SARS-CoV-2 subreplicon RNA expression vector was electroporated into HeLa cells and simultaneously treated with IFN-α or the IFNAR2 agonist CDM-3008 for 1 day, and cell viability and luminescence intensity were measured ( Figure 3A). The IC50 values of IFN-α and CDM-3008 were 667 IU/mL, without showing cytotoxicity ( Figure 3B), and the IC50 and CC50 of CDM-3008 were 0.78 μM and >100 μM, respectively ( Figure 3C). Therefore, the activity of IFN-α and CDM-3008 in HeLa cells was comparable to that in Calu-3 cells, and the assay was performed within 1 day. However, The growth rate of Calu-3 cells was not higher than that of HeLa cells, and Calu-3 cells were not used for electroporation. Thus, we changed the Calu-3 cells to HeLa cells to construct a screening system. The SARS-CoV-2 subreplicon RNA expression vector was electroporated into HeLa cells and simultaneously treated with IFN-α or the IFNAR2 agonist CDM-3008 for 1 day, and cell viability and luminescence intensity were measured ( Figure 3A). The IC 50 values of IFN-α and CDM-3008 were 667 IU/mL, without showing cytotoxicity ( Figure 3B), and the IC 50 and CC 50 of CDM-3008 were 0.78 µM and >100 µM, respectively ( Figure 3C). Therefore, the activity of IFN-α and CDM-3008 in HeLa cells was comparable to that in Calu-3 cells, and the assay was performed within 1 day. However, CDM-3008 reduced cell viability to 72.8% at 100 µM, while IFN-α did not. Thus, we selected IFN-α as a positive control for chemical screening of HeLa cells.
CDM-3008 reduced cell viability to 72.8% at 100 μM, while IFN-α did not. Thus, we selected IFN-α as a positive control for chemical screening of HeLa cells. Next, stable SARS-CoV-2 subreplicon RNA-expressing HeLa cells were prepared using an expression vector containing a neomycin-resistant gene. After selection with 1 mg/mL G418, the cells were plated and simultaneously treated with IFN-α for 1 day (Figure 4A). The luminescence intensity was suppressed by IFN-α in a dose-dependent manner, with an IC50 of 1875 IU/mL without cytotoxicity ( Figure 4B). IFN-α activity was suppressed by the pan-JAK inhibitor Pyridone 6 ( Figure 4C), suggesting that IFN-α induced the expression of ISGs by activating JAK/STAT, and ISGs enhanced the degradation of SARS-CoV-2 subreplicon RNA. The cells treated with 10 μM Pyridone 6 had approximately 80% cell viability, suggesting that 10 μM Pyridone 6 had weak cytotoxicity in HeLa cells. Next, stable SARS-CoV-2 subreplicon RNA-expressing HeLa cells were prepared using an expression vector containing a neomycin-resistant gene. After selection with 1 mg/mL G418, the cells were plated and simultaneously treated with IFN-α for 1 day ( Figure 4A). The luminescence intensity was suppressed by IFN-α in a dose-dependent manner, with an IC 50 of 1875 IU/mL without cytotoxicity ( Figure 4B). IFN-α activity was suppressed by the pan-JAK inhibitor Pyridone 6 ( Figure 4C), suggesting that IFN-α induced the expression of ISGs by activating JAK/STAT, and ISGs enhanced the degradation of SARS-CoV-2 subreplicon RNA. The cells treated with 10 µM Pyridone 6 had approximately 80% cell viability, suggesting that 10 µM Pyridone 6 had weak cytotoxicity in HeLa cells. To identify the target molecule for chemical screening, we performed knockdown of RNA-editing enzymes within ISGs, which was able to induce SARS-CoV-2 subreplicon RNA degradation. We assumed that ISGs such as OAS1, 2, and 3 and ISG20, which detect virus-derived RNA and promote degradation, enhance the degradation of SARS-CoV-2 subreplicon RNA. We knocked down OAS1, 2, and 3 and ISG20 using siRNAs in a SARS-CoV-2 subreplicon RNA-expressing stable cell line, and then the cells were treated with IFN-α for 1 d ( Figure 5A). Treatment with ISG20 siRNA, even without IFN-α, increased the luminescence intensity. ISGs were thought to be induced by the IFN-α-like substance contained in the cell culture medium, and SARS-CoV-2 subreplicon RNA was constantly degraded. The degradation was suppressed by ISG20 siRNA ( Figure 5B). Furthermore, the luminescence intensity was suppressed by 1000 IU/mL IFN-α, and this effect was partly canceled by ISG20 siRNA. These results suggest that SARS-CoV-2 genomic RNA is degraded by ISG20. To identify the target molecule for chemical screening, we performed knockdown of RNA-editing enzymes within ISGs, which was able to induce SARS-CoV-2 subreplicon RNA degradation. We assumed that ISGs such as OAS1, 2, and 3 and ISG20, which detect virus-derived RNA and promote degradation, enhance the degradation of SARS-CoV-2 subreplicon RNA. We knocked down OAS1, 2, and 3 and ISG20 using siRNAs in a SARS-CoV-2 subreplicon RNA-expressing stable cell line, and then the cells were treated with IFN-α for 1 d ( Figure 5A). Treatment with ISG20 siRNA, even without IFN-α, increased the luminescence intensity. ISGs were thought to be induced by the IFN-α-like substance contained in the cell culture medium, and SARS-CoV-2 subreplicon RNA was constantly degraded. The degradation was suppressed by ISG20 siRNA ( Figure 5B). Furthermore, the luminescence intensity was suppressed by 1000 IU/mL IFN-α, and this effect was partly canceled by ISG20 siRNA. These results suggest that SARS-CoV-2 genomic RNA is degraded by ISG20.

Discussion
In this study, we developed an experimental system that easily shows the anti-SARS-CoV-2 activity of IFN-α within 1 day. Using this experimental system, we were able to identify small-molecule compounds that promote the degradation of SARS-CoV-2 genomic RNA. We confirmed the inhibitory effect of the IFNAR2 agonist CDM-3008, a small chemical compound, in Calu-3 and HeLa cells. This experimental system is optimal for screening compounds that promote the degradation of SARS-CoV-2 genomic RNA.

Therapeutic Effect of IFNs against SARS-CoV-2 In Vivo and In Vitro
To analyze the pharmacological effects of IFNs in vivo, several coronavirus-infected animal models were used. Using MERS-CoV-infected mice expressing hDPP4, IFN-β and IFN-γ suppressed MERS-CoV [29]. In the MERS-CoV-infected common marmoset model, MERS-CoV levels were lower in the IFN-β1b-treated group than those in untreated animals [30]. In MERS-CoV-infected rhesus macaques, a combinational treatment with ribavirin and IFN-α2b reduced MERS-CoV levels as compared to untreated animals [31]. stably expressing SARS-CoV-2 subreplicon were treated with OAS1, 2, and 3 and ISG20 siRNA for 1 day, and then the cells were treated with 1000 IU/mL IFN-α for 1 day. Luciferase intensity (blue bars) and cell viability (orange line) were measured and are shown as % of control siRNA-treated DMSO control. Error bars indicate SD (n = 3). * p < 0.05 (two-tailed t-test).

Discussion
In this study, we developed an experimental system that easily shows the anti-SARS-CoV-2 activity of IFN-α within 1 day. Using this experimental system, we were able to identify small-molecule compounds that promote the degradation of SARS-CoV-2 genomic RNA. We confirmed the inhibitory effect of the IFNAR2 agonist CDM-3008, a small chemical compound, in Calu-3 and HeLa cells. This experimental system is optimal for screening compounds that promote the degradation of SARS-CoV-2 genomic RNA.

Therapeutic Effect of IFNs against SARS-CoV-2 In Vivo and In Vitro
To analyze the pharmacological effects of IFNs in vivo, several coronavirus-infected animal models were used. Using MERS-CoV-infected mice expressing hDPP4, IFN-β and IFN-γ suppressed MERS-CoV [29]. In the MERS-CoV-infected common marmoset model, MERS-CoV levels were lower in the IFN-β1b-treated group than those in untreated animals [30]. In MERS-CoV-infected rhesus macaques, a combinational treatment with ribavirin and IFN-α2b reduced MERS-CoV levels as compared to untreated animals [31]. Using ACE2-transgenic mice, IFN-λ1a showed an inhibitory effect on SARS-CoV-2 [32].
These in vivo analyses of animal models support the use of IFNs in the treatment of MERS and COVID-19.
In the activity measurement using the SARS-CoV-2 subreplicon RNA in Calu-3 cells, the IC 50 value of IFN-α was 193 IU/mL. The IC 50 of IFN-α in SARS-CoV-2-infected Calu-3 cells was less than 100 U/mL [33]. This may have occurred because the activity measurement using SARS-CoV-2-infected Calu-3 cells was performed after 16 h of pretreatment with IFN-α before infection, whereas IFN-α treatment of SARS-CoV-2 subreplicon RNA was performed 24 h after transfection. Collectively, the in vitro analysis of IFN-α using Calu-3 cells also supports the use of IFN-α for the treatment of COVID-19.

SARS-CoV-2 Replicon System
Replicons have been produced for high-throughput screening of SARS-CoV and SARS-CoV-2. Hertzig et al. constructed SARS-CoV replicon RNA consisting of 5 -UTR, ORF1a, ORF1b, GFP or luciferase, nucleocapsid, and 3 -UTR and used this system for IFN-α activity assays (IC 50 = 10 U/mL) [21]. Similar replicon RNA was constructed using the SARS-CoV-2 genomic RNA sequence and was used for high-throughput screening [20,26,34,35]. In addition, Luo et al. constructed three expression vectors containing SARS-CoV-2 genes nsp1-16 and expressed these genes with SARS-CoV-2 replicon RNA consisting of 5 -UTR, IRES, GFP, transcription regulatory sequence, luciferase, and 3 -UTR [22]. Vial et al. showed that the SARS-CoV-2 proteins required for the synthesis of positive strands from negativestrand replicon RNA are Nsp7, 8, and 12. Using these systems, they confirmed the anti-SARS-CoV-2 activity of remdesivir [36]. From these studies, it was concluded that the 5 -UTR, 3 -UTR, and reporter genes such as GFP and luciferase are required to produce SARS-CoV-2 replicon RNA, and that RNA-dependent RNA polymerases (RdRp, nsp12) are required for replication. In contrast, in our experimental system for measuring the activity of IFNs, only the positive strand of SARS-CoV-2 subreplicon RNA is necessary because we are measuring the degradation of RNA, and the action of RdRp is unnecessary. Thus, our SARS-CoV-2 subreplicon system is simpler and more compact than the previously reported SARS-CoV-2 replicon systems.

SARS-CoV-2 Subreplicon RNA Degradation Mechanism
In this study, we showed that ISG20 is involved in the degradation of the SARS-CoV-2 RNA genome ( Figure 5). ISG20 is a 20 kDa exonuclease that degrades singlestranded RNA and DNA [37]. Imam et al. found that the epsilon stem-loop of N6methyladenosine (m 6 A)-modified HBV pregenomic RNA (pgRNA) is recognized by the m6A reader protein YTH-domain family 2 (YTHDF2), and that degradation is induced by a complex of YTHDF2 and ISG20 [38]. Additionally, Liu et al. assessed the SARS-CoV-2 m 6 A methylome and showed that m 6 A sites are mainly accumulated in the regions of nucleocapsid, ORF10, and 3 -UTR in positive-sense genomic RNA. M 6 A methylation is increased in negative-sense genomic RNA in the 5 -UTR and ORF1a regions from 24 to 56 h after infection [39]. SARS-CoV-2 subreplicon RNA is a positive strand because transcription by RNA-dependent polymerases does not occur in our system. Thus, it is considered that the nucleocapsid, ORF10, and 3 -UTR regions of SARS-CoV-2 subreplicon RNA are modified by m 6 A, recognized by YTHDF2, and degraded by ISG20. In addition, during the IFN-β response, many mRNAs encoding ISGs are modified by m 6 A. YTHDF1 binds to m 6 A-modified mRNA and enhances transcription [40]. This suggests that type I IFNs and CDM-3008 induce ISG20 mRNA expression and simultaneously promote translation of ISG20, and that degradation of SARS-CoV-2 genomic RNA could be enhanced by ISG20 and YTHDF1 complexes. The development of a SARS-CoV-2 treatment method that promotes m 6 A modification and ISG20 activation induced by type I IFNs and CDM-3008 should be further explored.

Stable Cell Line for SARS-CoV2 Subreplicon RNA
The pCMV-SARS-CoV-2 subreplicon RNA expression vector was digested with BamHI overnight. After purification of the vector using Qiaquick PCR purification (Qiagen, Venlo, The Netherlands), 2 µg of the digested vector and 6 µL of X-tremeGene 360 transfection reagent (Roche) were mixed and added to HeLa cells cultured on a 35 mm plate. After 3 days, the HeLa cells were replated and selected with 1 mg/mL G418-containing medium. After 2 weeks of selection, the SARS-CoV2 subreplicon 3C5 cell line was selected.

Luciferase Assay
After treatment, the luciferase intensity was measured using the Steady-Glo Luciferase Assay System (Promega), according to the manufacturer's instructions. Briefly, the cells were mixed with 50 µL of cell culture medium and Steady Glo reagent (Promega) for 10 min at 25 • C using a plate shaker. Luciferase intensity was measured using a multimode plate reader (EnSight, PerkinElmer, Waltham, MA, USA).

Cell Viability Assay (XTT Assay)
After treatment, cell viability was assayed using the Cell Proliferation Kit II (Roche) according to the manufacturer's instructions. Briefly, 100 µL of cell culture medium and 50 µL of XTT reagent (Cell Proliferation Kit II, Roche) under 5% CO 2 at 37 • C for 45 min were added to the cells. The absorbance at 492 nm was measured using a multimode plate reader (EnSight, PerkinElmer).

Statistical Analysis
Statistical analysis was performed using Microsoft Office 2016 (Microsoft, Redmond, WA, USA).

Conclusions
The advantages of small-molecule compounds with IFN-like activity that are available in oral and inhalation formulations include innate immune response activation in the host and the inhibition of various viruses. To measure anti-SARS-CoV-2 activity through RNA genome degradation, we constructed a SARS-CoV-2 subreplicon RNA expression vector. IFN-α and CDM-3008 showed anti-SARS-CoV-2 activity in SARS-CoV-2 subreplicon RNA-expressing Calu-3 and HeLa cells. We constructed a virus-free screening system to select small chemical compounds with IFN-like activity. It is important to screen for new compounds using this system and to develop several therapeutic drugs for future pandemics.

Data Availability Statement:
The data presented in this study are available in the article.