A synthetic genomics-based African swine fever virus engineering platform

Abstract

African swine fever (ASF) is a deadly viral disease in domestic pigs that has a large global economic impact for the swine industry. It is present in Africa, Europe, Asia, and in the Caribbean island of Hispaniola. There are no effective treatments or broadly licensed vaccines to prevent disease. Efforts to counteract ASF have been hampered because of the lack of convenient tools to engineer its etiological agent, ASF virus (ASFV), largely due to its large noninfectious genome. Here, we report the use of synthetic genomics methodology to develop a reverse genetics system for ASFV using a CRISPR-Cas9-inhibited self-helper virus to reconstitute live recombinant ASFV from synthetic genomes to rapidly generate a variety of combinatorial mutants of ASFV. The method will substantially facilitate the development of therapeutics or subunit and live-attenuated vaccines for ASF. This synthetic genomics-based approach has wide-ranging impact because it can be applied to rapidly develop reverse genetics tools for emerging viruses with noninfectious genomes.

Fig. 1.. Reconstitution of live ASFV from transfected genomic DNA using a self-helper virus.

(A) The ASFV genome is by itself not infectious since transfection of the viral genome into mammalian cells does not produce virus progeny. (B) Replication of ASFV genotype I or II is inhibited in mammalian cells when the p30 gene is targeted for cleavage by a gRNA-directed Cas9 nuclease. (C) Replication of ASFV genotype IX is not inhibited in mammalian cells when the p30 gene is targeted by the same gRNA-directed Cas9 nuclease and produces virus progeny. This is due to single-nucleotide polymorphisms at the target site between the different genotypes. (D) Hypothesis: An ASFV genome can be booted-up by a self-helper virus (either a heterologous or homologous ASFV) whose replication is inhibited by CRISPR-Cas9 cleavage activity as long as the donor genome is immune to the same cleavage activity. (E) Fluorescence images of representative plaques produced by the reconstituted virus [ΔCD2v::DsRed (D)] and the parental virus [ΔCD2v::DsRed (V)] at 6 days postinfection of WSL cells. ns, not significant. (F) Plaque areas (square millimeter) produced by the same viruses. They were measured 6 days after infection of WSL cells. Shown are the mean sizes of 30 plaques per virus with positive standard deviations. (G) Replication of the reconstituted virus [ΔCD2v::DsRed (D)] on WSL cells. Progeny virus titers [plaque-forming unit (PFU) per milliliter] of ΔCD2v::DsRed (D) and parental virus [ΔCD2v::DsRed (V)] were determined at the indicated times after infection at an MOI of 0.03. Shown are the mean results of at least three parallel experiments.

Fig. 2.. Schematic depiction of the synthetic genomics assembly of the ASFV-Kenya-IX-1033 genome.

(A) Deconstruction of ASFV-Kenya genome. Twelve approximately 15-kb overlapping genomic fragments span the entire sequence of the ASFV-Kenya genome. (B) TAR-assembly of third genomes. Third genomes of ASFV-Kenya were assembled separately from fragments 1 to 4, fragments 5 to 8, and fragments 9 to 12. The corresponding fragments were pooled with a linear YCpBAC vector that contained terminal homology to each end of the pertinent third genome and flanked by a unique restriction enzyme (RE) site and then transformed into yeast cells. Transformants were screened by PCR amplification for correct assembly. DNA from the positive transformants was isolated and transformed into E. coli cells to produce high concentration DNA stocks. (C) TAR assembly of full-length genome and reconstitution of live recombinant virus. Plasmid DNAs of the three third genomes were isolated and digested to release the ASFV-Kenya fragments. The third fragments were transformed into yeast cells together with a linear YCpBAC vector that contained homology to the terminal repeats of ASFV flanked by unique restriction sites and a different marker for selection in yeast that was used for the third genome assemblies. DNA from the positive transformants was isolated and transformed into E. coli. DNA was isolated from E. coli, and the viral genome was released by digestion with an RE targeting the flanking sites. Hairpin loops were ligated to the viral genome and transfected into WSL cells that contained the CRISPR-Cas9 system. The transfected cells were then infected by a self-helper virus to reconstitute live recombinant ASFV-Kenya. (D) Schematic of genes targeted for modification. ASFV genes that were modified are shown. The numbered gray segments represent the fragments containing the targeted genes. The blue rectangles represent the terminal hairpin loops at the ends of the ASFV genome.

Fig. 3.. Characterization of the TAR-assembled synthetic ASFV-Kenya viruses.

(A) Junction PCR amplification analysis of assembled full-length DCD2v::mCh genome. Representative agarose gel after PCR amplification showing the presence of all the appropriate junctions between the ASFV-Kenya fragments (1, 2, etc.). For the lane headings, numbers give the ASFV-Kenya fragment numbers and V refers to vector for the junctions tested. The positions of the molecular size markers in the M lane are given to the left of the gel. (B) Junction PCR amplification analysis of assembled full-length ΔCD2v::mCh/p12M genome. Representative agarose gel after PCR amplification showing the presence of all the appropriate junctions between the ASFV-Kenya fragments (1, 2, etc.). For the lane headings, numbers give the ASFV-Kenya fragment numbers and V refers to vector for the junctions tested. PCR amplification of the mutated segment of the p12 gene is also shown in the rightmost lane. The positions of the molecular size markers in the M lane are given to the left of the gel. (C) Plaque areas (square millimeter) produced by the TAR-assembled synthetic viruses. They were measured 6 days after infection of WSL cells and compared to booted-up ΔCD2v::DsRed (D) and parental ΔCD2v::DsRed (V) viruses. Shown are the mean sizes of >50 plaques per virus with positive standard deviations. For non-parametric multiple groups analyses, we used a Kruskal-Wallis H test with Dunn’s post hoc analysis. Asterisks denote the level of significance observed: ****P ≤ 0.0001. (D) Replication of the reconstituted synthetic viruses on WSL cells. Progeny virus titers (plaque-forming unit per milliliter) of ΔCD2v::mCh, ΔCD2v::mCh/p12M, ΔCD2v::DsRed (D) and ΔCD2v::DsRed (V) were determined at the indicated times after infection at an MOI of 0.03. Shown are the mean results of at least three parallel experiments.

Fig. 4.. Characterization of combinatorial ASFV-Kenya mutant viruses.

(A) Fluorescence images of representative plaques produced by the viruses reconstituted from synthetic assembled genomes, ΔK145R::mCh/p12M and ΔK145R::mCh/ΔA238L/ΔI329L/p12M, compared to the self-helper virus, ΔDUT::GFP, at 6 days postinfection of WSL cells. (B) Plaque areas (square millimeter) produced by the analyzed viruses. They were measured 6 days after infection of WSL cells and compared to ΔDUT::GFP. Shown are the mean sizes of >50 plaques per virus with positive standard deviations. For non-parametric multiple groups analyses, we used a Kruskal-Wallis H test with Dunn’s post hoc analysis. Asterisks denote the level of significance observed: *P ≤ 0.05. (C) Replication of the reconstituted viruses on WSL cells. Progeny virus titers (plaque-forming unit per milliliter) of ΔK145R::mCh/p12M, ΔK145R::mCh/ΔA238L/p12M, ΔK145R::mCh/ΔI329L/p12M, ΔK145R::mCh/ΔA238L/ΔI329L/p12M, and control virus ΔDUT::GFP were determined at the indicated times after infection at an MOI of 0.03. Shown are the mean results of at least three parallel experiments.