Week 3 Update: Establishing a scalable mammalian cell fluorescence-based assay for SARS-CoV2 RdRP activity

**Note: We are currently searching for SARS-CoV2 genomic RNA or cDNA for use in our project! Our order from ATCC has been backordered until October and we are looking for an alternative. Please reach out to our lab if you may be able to help (rsherwood@rics.bwh.harvard.edu).

Background: SARS-CoV2 is a positive strand RNA virus with a large, 30 kb genome. The first two-thirds of the viral RNA (ORF1ab) is directly translated by the host upon viral infection of a cell to create one giant polyprotein. This polyprotein is posttranslationally cleaved to create 16 non-structural proteins. Many of these proteins are involved in replication, such as nsp 12, the RNA dependant RNA polymerase (RdRP) enzyme.¹ 

The last third of the genome (ORFs 2-10) encodes for the SARS-CoV2 structural proteins. These ORFs are not directly translated by ribosomes upon infection. Rather, they are first transcribed by RdRP into negative strand subgenomic mRNAs (sgmRNAs), which are transcribed again into positive strand mRNAs.¹ This process allows each gene to position itself at the 5’ end of the mRNA to be translated by the host ribosomes.

SARS-CoV2 replication and transcription are major targets for COVID-19 drug candidates. For instance, remdesivir is a current COVID-19 drug candidate that acts as an adenosine analogue, blocking RdRP’s ability to transcribe RNA.² It is also thought that hydroxychloroquine (HCQ), another COVID-19 drug candidate, may at least partially restrict SARS-CoV2 infection through increasing cellular concentration of zinc which impedes RdRP activity. Several other proposed COVID-19 drugs (favipiravir, EIDD-2081) also work by impeding RdRP. However, there are no established methods to specifically investigate SARS-CoV2 transcription in mammalian cells. While viral infectivity assays have been used to demonstrate in vitro efficacy of RdRP-inhibiting drugs such as remdesivir and HCQ, these assays do not allow specific study of RdRP, and they require level 3 biosafety, which most laboratories lack. Therefore, a biosafe assay for SARS-CoV2 RdRP activity would allow laboratories to understand the mechanism of antiviral drugs, develop new drugs targeting SARS-CoV2 RdRP, investigate mutations in RdRP that would lead to drug resistance, and understand genetic variations that may increase or decrease severity of viral infection.

Aim: Our assay will measure RdRP activity through measuring transcription of a reverse GFP subgenomic-like mRNA (revGFP sgmRNA) reporter construct into GFP mRNA and its subsequent translation to GFP protein. In the absence of active SARS-CoV2 RdRP, revGFP will be transcribed into revGFP sgmRNA, which does not encode any detectable protein products (Figure 1a). However, if the RdRP is active, revGFP sgmRNA will be transcribed by RdRP, producing GFP mRNA. This positive sense mRNA will then be translated by the host cell to GFP protein (Figure 1b). The presence of GFP positive cells, in comparison to a negative control, will indicate RdRP activity. The percentage and fluorescence intensity of GFP positive cells can be determined by flow cytometric analysis.

Figure 1: In the presence of active RdRP, RevGFP sgmRNA is transcribed into GFP mRNA, which is translated to GFP protein.

Thus, in order to assess SARS-CoV2 RdRP activity, we must introduce two components into host cells: our revGFP sgmRNA reporter construct, and the genes that encode SARS CoV2 RdRP.

revGFP reporter constructs: We have devised six variants of revGFP reporter constructs that take advantage of different steps in the SARS-CoV2 transcription and replication process. The 5’ UTR, 3’ UTR, and anti-leader sequence act as promoters for RdRP during replication, sgmRNA synthesis, and sgmRNA transcription, respectively. Therefore, we have designed three constructs that end in sense or antisense versions of each of the SARS-CoV2 promoters. Once the host cell transcribes these constructs into RNA, these sequences at the 3’ end of the transcript should serve as binding sites for RdRP to begin transcription, producing an RNA molecule with GFP RNA in the sense orientation. In addition, the last three constructs include a reverse 3’ UTR upstream of the revGFP. This may increase stability of the RdRP-produced GFP mRNA through the addition of a poly-A tail. We expect GFP production only when each construct is co-transfected with the SARS-CoV2 RdRP genes.

1. rev GFP – 3’ UTR
2. rev GFP – rev 5’ UTR
3. rev GFP – rev 5’ anti-leader
4. rev 3’ UTR – rev GFP – 3’ UTR
5. rev 3’ UTR –rev GFP – rev 5’ UTR
6. rev 3’ UTR – rev GFP – rev 5’ antileader

Expression of SARS-CoV2 RdRP genes: Even though SARS-CoV2 nsp12 is known to encode the RdRP enzyme, it does not function efficiently on its own, and it is not well understood which viral nsps are required for efficient RdRP activity (which in our case will be read out as revGFP transcription into GFP). Among the 16 nsps encoded by the SARS-CoV2 genome, nsps 7, 8, 12, and 14 have been reported to be essential for transcription of the viral genome. In addition, nsps 9, 10, 13, 15, and 16 may be important for efficient viral transcription given their known interactions with RNA and/or the four essential nsps previously listed.³ Given the uncertainty of which genes are required to reconstitute RdRP activity and given the diverse goals of establishing this assay, we have come up with two approaches for expressing SARS-CoV2 RdRP in host cells, which we will optimize separately:

Approach #1 – Codon-Optimized Vectors: Stephen Elledge’s lab has graciously given us lentiviral vectors encoding each codon-optimized nsp listed above. To begin to determine whether any combination of nsps can efficiently transcribe our reporter construct, we will transiently transfect Hek293FT cells via TransIT with a revGFP reporter and all nine nsp vectors. Since each nsp is in a separate vector, we can easily add different combinations of nsps. If we succeed in obtaining GFP expression with the addition of all 9 nsps, we will remove each nsp vector one by one during transfection. If GFP is still produced at equivalent levels by the cells after transfection, the nsp removed will be recorded as non-essential for RdRP function. All treatments will be compared to a control well of cells transfected only with revGFP. We will continue this process until we determine the minimal nsps required for efficient transcription of our revGFP construct by RdRP. Because these vectors are in a lentiviral backbone, it is easy to make stable host cell lines co-expressing all required nsps once we have determined the appropriate combination.

Approach #2 – WT Sequence – Rather than using the individual codon optimized vectors, another approach utilizes the wild-type nsp sequences as they occur in the viral genome. Because it involves viral sequences in their native genomic order, this approach would allow the study of specific mutations in SARS-CoV2 RdRP or other nsps. We ordered viral SARS-CoV2 RNA from BEI Resources (NR-52285), which we will reverse transcribe into cDNA. Next, we will amplify the SARS-CoV2 genomic region encompassing nsps 5-16 using PCR. Given the long length of this sequence (~11kb), we will amplify the region both all at once and in 3+ kb segments as an alternative. Next, we will perform an NEBuilder to insert the amplified region(s) into a digested Tol2 transposon-based CAG-Cas9NG-P2A-mChe-BlastR vector. The resulting CAG-nsp5-16-P2A-mChe plasmid can be cotransfected into Hek293T cells along with the revGFP reporter construct, and it also allows us to construct stable RdRP-expressing cell lines through transposon integration.

Once we establish a fluorescence-based assay for SARS-CoV2 RdRP using either or both of these approaches, we plan to investigate the following areas:

• Evaluating remdesivir/HCQ-resistant mutations in SARS-CoV2 RdRP that may arise during treatment of SARS-CoV2 patients.
• Screening and evaluating other drugs with RdRP-inhibiting activity without the need for a BSL3 live virus assay.
• Understanding host genes that interact with SARS-CoV2 RdRP which may help us identify markers and mechanisms of genetic susceptibility and resistance to SARS-CoV2 infection and remdesivir treatment.

We have already begun prepping the nsp lentiviral vectors and cloning the revGFP reporters. However, we are still waiting for our SARS-CoV2 genomic RNA to arrive. Over the next few weeks I will be posting our results, so check here and our Twitter (@SherwoodLab) for updates!

1. Fehr, A. R. and Perlman, S. (2015). Coronaviruses: an overview of their replication and pathogenesis. Methods in molecular biology, 1282, 1–23.
2. Warren, T., Jordan, R., Lo, M. et al. (2016). Therapeutic efficacy of the small molecule GS-5734 against Ebola virus in rhesus monkeys. Nature, 531, 381–385.
3. Isabel Sola, Pedro A. Mateos-Gomez, Fernando Almazan and Sonia Zuñiga & Luis Enjuanes (2011). RNA-RNA and RNA-protein interactions in coronavirus replication and transcription, RNA Biology, 8:2, 237-248.