Development of CRISPR as an Antiviral Strategy to Combat SARS-CoV-2 and Influenza

Abstract

The coronavirus disease 2019 (COVID-19) pandemic, caused by the SARS-CoV-2 virus, has highlighted the need for antiviral approaches that can target emerging viruses with no effective vaccines or pharmaceuticals. Here, we demonstrate a CRISPR-Cas13-based strategy, PAC-MAN (prophylactic antiviral CRISPR in human cells), for viral inhibition that can effectively degrade RNA from SARS-CoV-2 sequences and live influenza A virus (IAV) in human lung epithelial cells. We designed and screened CRISPR RNAs (crRNAs) targeting conserved viral regions and identified functional crRNAs targeting SARS-CoV-2. This approach effectively reduced H1N1 IAV load in respiratory epithelial cells. Our bioinformatic analysis showed that a group of only six crRNAs can target more than 90% of all coronaviruses. With the development of a safe and effective system for respiratory tract delivery, PAC-MAN has the potential to become an important pan-coronavirus inhibition strategy.

Keywords: 2019-nCoV; COVID-19; CRISPR; Cas13; IAV; RdRP; SARS-CoV-2; antiviral; influenza; nucleocapsid.

Graphical abstract

Figure 1

The Hypothetical Life Cycle of SARS-CoV-2 and the PAC-MAN Approach for Inhibiting Coronavirus Using CRISPR-Cas13 (A) A hypothetical life cycle of SARS-CoV-2. Upon SARS-CoV-2 entry and genome RNA release, the positive strand RNA genome serves as a template to make negative strand genomic and subgenomic templates, which are used to produce more copies of the positive strand viral genome and viral mRNAs. (B) Cas13d can inhibit viral function and replication by directly targeting and cleaving all viral positive-sense RNA.

Figure 2

Bioinformatic Analysis of Cas13d Target Sites for SARS-CoV-2 and Construction of the PAC-MAN System (A) Alignment of 47 patient-derived SARS-CoV-2 sequences with SARS-CoV and MERS-CoV. Top: predicted abundance of crRNAs that are able to target SARS-CoV-2 genomes and SARS or MERS. Middle: annotation of genes in the SARS-CoV-2 genomes, along with conserved regions chosen to be synthesized into the SARS-CoV-2 reporters (magenta and purple). Bottom: percentage of conservation between aligned viral genomes. See Table S2 for the designed crRNA sequences and synthesized SARS-CoV-2 fragments. (B) Schematic of the two reporters (SARS-CoV-2-F1/F2) created with synthesized viral sequences. SARS-CoV-2-F1 contains GFP fused to a portion of RdRP (RdRP-F1) and SARS-CoV-2-F2 contains GFP fused to portions of both RdRP (RDRP-F2) and N. (C) Schematics for the constructs used to express Cas13d or crRNAs. (D) Experiment workflow to challenge Cas13d A549 lung epithelial cells with SARS-CoV-2 reporters. See also Figure S1 and Tables S1 and S2.

Figure S1

A Bioinformatic Pipeline to Predict Effective and Specific crRNA Designs, Related to Figure 2A Our bioinformatic method analyzes all possible crRNAs that target regions conserved between reported SARS-CoV-2, SARS-CoV, and MERS-CoV.

Figure 3

PAC-MAN Can Inhibit SARS-CoV-2 Reporters (A and B) Left: schematics of pools of crRNAs targeting transfected (A) SARS-CoV-2-F1 or (B) SARS-CoV-2-F2 reporters. Middle: GFP expression as measured by flow cytometry. Right: mRNA abundance as measured by quantitative real-time PCR. Relative RNA expression is calculated by normalizing to the reporter only sample. p values for each group are included in Table S3. (C) GFP expression as measured by flow cytometry when SARS-CoV-2-F1 (left) or SARS-CoV-2-F2 (right) is delivered via lentiviral transduction. p values for G1 (p = 10^–6) and G4 (p = 7×10^–6) are relative to SARS-CoV-2-F1 only, while p values for G5 (p = 0.001) and G6 (p = 3×10^–6) are relative to SARS-CoV-2-F2 only. MOI = 0.5. (D) mRNA abundance measured by quantitative real-time PCR when SARS-CoV-2-F1 (left) or SARS-CoV-F2 (right) is delivered via lentiviral transduction. Relative RNA expression is calculated by normalizing to the reporter only sample. p values for G1 (p = 10^–5) and G4 (p = 8×10^–8) are relative to SARS-CoV-2-F1 only, while p values for G5 (p = 4×10^–6) and G6 (p = 1×10^–4) are relative to SARS-CoV-2-F2 only. MOI = 0.5. (E) Top: schematic of pools of crRNAs tiling across SARS-CoV-2-F1 and SARS-CoV-2-F2 reporters. Bottom: GFP expression levels as measured by flow cytometry. Red, SARS-CoV-2-F1 reporter; blue, SARS-CoV-2-F2 reporter. NT, non-targeting crRNAs. p values are relative to the NT samples; p = 0.01 for the F1 pool and p = 5×10^–4 for the F2 pool. See also Figures S2 and S3 and Tables S1, S2, and S3.

Figure S2

Flow Cytometry Plots Demonstrating SARS-CoV-2-19 F1 and F2 Transfection and Transduction, Related to Figures 3A–3D (A) Examination of SARS-CoV-2-19 reporter levels in A549 cells 48 hours after transfection or transduction. While transfection led to a much lower percentage of GFP positive cells compared to transduction, transfection also led to much higher levels of GFP expression, giving a better dynamic range for assessing repression. (B) Representative flow histograms of GFP reporter levels after SARS-CoV-2 reporter challenge. Cells shown were gated for GFP+ cells only.

Figure S3

Flow Cytometry Plots Demonstrating Tiled crRNAs with a Strong Repression of the SARS-CoV-2 F1 and F2 Reporters, Related to Figure 3E (A) Representative flow cytometry histograms of GFP expression after SARS-CoV-2-F1 (left) and SARS-CoV-2-F2 (right) reporter challenge. (B) Analysis of different expression levels of Cas13d (marker: mCherry) and crRNA pools (marker: BFP) and the resulting effects on SARS-CoV-2-F1/F2 reporter inhibition. Percent reduction of GFP expression is labeled on the top. Red, SARS-CoV-2-F1 reporter; blue, SARS-CoV-2-F2 reporter. (C) Representative flow cytometry plots of forward versus side scatter (FSC versus SSC) for A549 cells expressing Cas13d, crRNAs (crNT – non-targeting, F1 pool, or F2 pool) and the matching SARS-CoV-2 reporters in A549 cells.

Figure 4

PAC-MAN Can Inhibit Infection in Lung Epithelial Cells (A) Workflow used to challenge Cas13d A549 lung epithelial cells with PR8 mNeon IAV. (B) Microscopy quantification of the percentage of mNeon⁺ cells per field of view (FOV) at MOI = 2.5 (p = 1 × 10⁻⁵, left) and 5 (p = 5 × 10⁻⁹, right). Each dot represents the percentage for a single microscopy FOV. n = 48 FOV with at least a total of 1,100 cells counted per condition. p values for the S6 crRNA pool are relative to the NT condition; p = 1×10^–5 for MOI = 2.5 and p = 5×10^–9 for MOI = 5. NT, non-targeting crRNA pool. (C) Flow cytometry evaluation of the percentage of mNeon⁺ cells at MOI = 2.5 (p = 0.039, left) and 5 (p = 0.0058, right). n = 6. p values for the S6 crRNA pool are relative to the NT condition; p = 5.8×10^–3 for MOI = 2.5 and p = 0.039 for MOI = 5. NT, non-targeting crRNA pool. (D) A histogram showing the predicted minimum number of crRNAs to target 91,600 IAV genomes. Dotted line, 90% of IAVs. See also Figure S4 and Tables S2 and S4.

Figure S4

Screening of Pools of Six crRNAs Targeting Each of the Eight IAV Genome Segments, Related to Figure 4 (A) Each panel shows the percentage of quantified mNeon+ cells under two infection conditions (MOI = 2.5 or 5). Each dot represents the percentage for a single randomly chosen microscopy field of view (FOV). Blue, non-targeting (NT) crRNAs; red, targeting crRNAs; n = 3 infected wells per condition and 9 FOV, > 300 cells counted per condition. (B) Flow cytometry screens of 6-crRNA pools targeting each IAV genome segment. The percentage of mNeon+ cells for each segment using targeting crRNA pooled compared to NT crRNAs at MOI = 2.5 (left) or 5 (right) are shown; n = 3 infected wells pooled together in a single flow cytometry tube for each condition. (C) Microscopy quantification of crRNA pools targeting IAV S4 and S6 at MOI = 0.5; n = 9 randomly chosen FOV, > 300 cells counted per condition with a biological n = 1.

Figure 5

Pan-Coronavirus Targeting Using a Minimal Pool of PAC-MAN crRNAs (A) A phylogenetic tree of sequenced coronaviruses, organized by genus of the strains. The inner ring shows coverage by each of the top six (PAC-MAN-T6) pan-coronavirus crRNAs targeting all human coronaviruses. The outer ring shows current species of coronaviruses that are infectious to humans, including SARS-CoV-2 (red). (B) A histogram showing the predicted minimum number of crRNAs to target all sequenced 3,051 coronavirus genomes. (C) Analysis of experimentally validated crRNA-N18f targeting 1,087 SARS-CoV-2 sequences from GISAID. The Weblogo of the target region is shown on the top. The bar graph shows the number of strains targeted by crRNA-N18f for each type (L/S/Unknown) of SARS-CoV-2. See also Figure S5 and Tables S4 and S5.

Figure S5

A Bioinformatic Pipeline to Predict Minimum Sets of Pan-Coronavirus crRNA Designs, Related to Figure 5B The bioinformatic pipeline analyzed and predicted pan-coronavirus targeting crRNAs. The number of crRNAs at each step of the workflow is denoted in parentheses.