Efficient CRISPR-Cas13d-Based Antiviral Strategy to Combat SARS-CoV-2

Abstract

The current SARS-CoV-2 pandemic forms a major global health burden. Although protective vaccines are available, concerns remain as new virus variants continue to appear. CRISPR-based gene-editing approaches offer an attractive therapeutic strategy as the CRISPR-RNA (crRNA) can be adjusted rapidly to accommodate a new viral genome sequence. This study aimed at using the RNA-targeting CRISPR-Cas13d system to attack highly conserved sequences in the viral RNA genome, thereby preparing for future zoonotic outbreaks of other coronaviruses. We designed 29 crRNAs targeting highly conserved sequences along the complete SARS-CoV-2 genome. Several crRNAs demonstrated efficient silencing of a reporter with the matching viral target sequence and efficient inhibition of a SARS-CoV-2 replicon. The crRNAs that suppress SARS-CoV-2 were also able to suppress SARS-CoV, thus demonstrating the breadth of this antiviral strategy. Strikingly, we observed that only crRNAs directed against the plus-genomic RNA demonstrated antiviral activity in the replicon assay, in contrast to those that bind the minus-genomic RNA, the replication intermediate. These results point to a major difference in the vulnerability and biology of the +RNA versus -RNA strands of the SARS-CoV-2 genome and provide important insights for the design of RNA-targeting antivirals.

Keywords: COVID-19; CRISPR-Cas13d; RNA; SARS-CoV-2 genome.

Figure 1

Replication cycle of SARS-CoV-2 and genome organization. (A) The SARS-CoV-2 replication cycle. Binding of the viral S protein to the ACE-2 receptor triggers SARS-CoV-2 infection and the plus genomic RNA (+gRNA) is translated into polyprotein (pp)1a and pp1ab. These proteins are proteolytically cleaved to generate 16 non-structural proteins (nsps), including some that form the replication/transcription complex that drives the synthesis of minus genomic RNA (−gRNA). The −gRNA is converted into genomic +gRNA that is packaged into new virion particles. Discontinuous transcription generates a set of 3′ co-terminal sub-genomic mRNAs (sgmRNAs) with identical 5′-leader and 3′-trailer ends. The sgmRNAs are translated into structural and accessory proteins that are required for the assembly of infectious virions, which takes place at the ER-Golgi intermediate compartment (ERGIC). Nascent virions are released from the cell via exocytosis. (B) The SARS-CoV-2 RNA genome. Human coronaviruses contain the largest viral genome (27–31 kb) among the RNA viruses and they share a similar genome organization. At the 5′-terminus two large overlapping open reading frames (ORFs), ORF 1a and ORF 1b, encode non-structural proteins and the 3′-terminal ORFs encode four structural proteins: spike (S), envelope (E), membrane (M) and nucleocapsid (N). The SARS-CoV-2 genome encodes eight accessory proteins: 3, 6, 7a, 7b, 8, 9, 14 and 10. Accessory proteins are indicated in grey.

Figure 2

Targeting of SARS-CoV-2 RNA with CRISPR/Cas13d. (A) SARS-CoV-2 RNA genome diversity and the crRNA targets. Distribution of crRNAs along the SARS-CoV-2 RNA genome (MN908947). The Shannon entropy along the RNA genome varies from 0 to 1, where lower genetic diversity gives values closer to zero.The position of the crRNAs targeting conserved RNA sequences of SARS-CoV-2 is indicated with black triangles. The SARS-CoV-2 RNA genome diversity is shown by plotting Shannon entropy along the whole genome. The values vary between 0 and 1, where more conserved sequences give lower values (closer to 0); (B) Luciferase multitarget reporter construct. To measure the efficiency of the designed crRNAs, a pGL3 control plasmid-based multi-target luciferase reporter construct was designed with SARS-CoV-2 conserved target sequences cloned downstream of the firefly luciferase gene. Two constructs were designed: one with the + and one with the −RNA target sequences. SV40: simian virus 40 promoter, pA: polyA sequence; (C) Engineering a SARS-CoV-2 replicon. Scheme of the SARS-CoV-2 cDNA cloned in a BAC (upper panel) and the SARS-CoV-2 replicon (lower panel). The letters indicate the viral genes: S, spike; E, envelope; M, membrane; N, nucleocapsid; or the reporter gene mNeonGreen (mNG). UTR, untranslated regions; CMV, cytomegalovirus promoter; pA, polyA sequence; Rz, hepatitis delta virus ribozyme; BGH, bovine growth hormone polyadenylation and termination signals.

Figure 3

The efficiency of the designed anti-SARS-CoV-2 crRNAs. (A,B) The efficiency of 29 crRNAs targeting either the + or the -RNA strand with Cas13d was tested in HEK293T cells by co-transfecting them with the luciferase construct encoding SARS-CoV-2 +RNA (A) and −RNA (B) target sequences, respectively. To quantify viral gene expression, luciferase level was measured at two days after transfection. Mean values (±SD) of three experiments in duplicate are shown. The average luciferase levels are expressed as relative percentage level (%), setting the negative control (NC), a crRNA not targeting any SARS-CoV-2 encoding sequence, at 100%. Statistical significance was determined using one-way ANOVA, * p ≤ 0.05.

Figure 4

Titration of SARS-CoV-2 targeting crRNAs. The efficiency of 14 crRNAs was tested in HEK293T cells by co-transfecting them with the luciferase construct encoding SARS-CoV-2 +RNA target sequences. A titration of the crRNA constructs was performed (75, 150 and 300 ng). To quantify viral gene expression, luciferase level was measured at two days after transfection. Mean values (±SD) of three experiments in duplicate are shown. The average luciferase levels are expressed as percentage (%) of luciferase expression, setting the negative control (NC) at 100%. Statistical significance was determined using one-way ANOVA, * p ≤ 0.05.

Figure 5

The anti-SARS-CoV-2 efficiency of single versus dual crRNA therapy. The efficiency of six crRNAs was tested in HEK293T cells by co-transfecting them with the luciferase construct encoding SARS-CoV-2 +RNA target sequences. The crRNAs were tested as single inhibitor versus a dual-combinatorial approach. To quantify viral gene expression, luciferase level was measured at two days after transfection. Mean values (±SD) of three experiments in duplicates are shown. The average luciferase levels are expressed as percentage (%) of luciferase expression, setting the negative control (NC) at 100%. Statistical significance was determined using one-way ANOVA, **** p ≤ 0.0001.

Figure 6

The efficiency of crRNAs targeting SARS-CoV-2 and SARS-CoV replicon. (A) The efficiency of five selected single crRNAs targeting SARS-CoV-2 sequences was tested in HEK293T cells by co-transfecting them with a SARS-CoV-2 replicon. crRNAs targeting both SARS-CoV-2 + and −RNA sequences were included. Viral RNA replication was quantified by measuring levels of mNeonGreen. Mean values (±SD) of three experiments in duplicates are shown. The average mNeonGreen levels are expressed as percentage (%) of fluorescent cells, setting the negative control (NC) at 100%; (B) The same experimental set-up was used with a SARS-CoV replicon. Here, viral RNA replication was quantified by measuring levels of GFP. Statistical significance was determined using one-way ANOVA, **** p ≤ 0.0001.

Figure 7

The impact of Cas13 on the intermediate minus sense transcripts. (A) Schematic representation of differences between +RNA and –RNA strands during SARS-CoV-2 RNA replication. +RNA strands (black arrows) are far more abudant than the –RNA strands (black dashed arrows). The less abundant –RNA strands are present as double-stranded replicative forms during RNA replication. (B) The efficiency of crRNAs targeting -RNA targets when annealed to complementary +RNA targets. The efficiency of six crRNAs was tested in HEK293T cells by co-transfecting them with a mixture of different ratios of − and + luciferase constructs (1:5, 1:25, 1:50, 1:75 and 1:100 of − luciferase construct: + luciferase construct) or with the − luciferase construct alone (1:0). To quantify viral gene expression, luciferase level was measured at two days after transfection. Mean values (±SD) of three experiments in duplicates are shown. The average luciferase levels are expressed as percentage (%) of luciferase expression, setting the negative control (NC) at 100%. Statistical significance was determined using one-way ANOVA, * p ≤ 0.05.