Targeted Mutagenesis of Guinea Pig Cytomegalovirus Using CRISPR/Cas9-Mediated Gene Editing

Abstract

The cytomegaloviruses (CMVs) are among the most genetically complex mammalian viruses, with viral genomes that often exceed 230 kbp. Manipulation of cytomegalovirus genomes is largely performed using infectious bacterial artificial chromosomes (BACs), which necessitates the maintenance of the viral genome in Escherichia coli and successful reconstitution of virus from permissive cells after transfection of the BAC. Here we describe an alternative strategy for the mutagenesis of guinea pig cytomegalovirus that utilizes clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9)-mediated genome editing to introduce targeted mutations to the viral genome. Transient transfection and drug selection were used to restrict lytic replication of guinea pig cytomegalovirus to cells that express Cas9 and virus-specific guide RNA. The result was highly efficient editing of the viral genome that introduced targeted insertion or deletion mutations to nonessential viral genes. Cotransfection of multiple virus-specific guide RNAs or a homology repair template was used for targeted, markerless deletions of viral sequence or to introduce exogenous sequence by homology-driven repair. As CRISPR/Cas9 mutagenesis occurs directly in infected cells, this methodology avoids selective pressures that may occur during propagation of the viral genome in bacteria and may facilitate genetic manipulation of low-passage or clinical CMV isolates.

Importance: The cytomegalovirus genome is complex, and viral adaptations to cell culture have complicated the study of infection in vivo Recombineering of viral bacterial artificial chromosomes enabled the study of recombinant cytomegaloviruses. Here we report the development of an alternative approach using CRISPR/Cas9-based mutagenesis in guinea pig cytomegalovirus, a small-animal model of congenital cytomegalovirus disease. CRISPR/Cas9 mutagenesis can introduce the same types of mutations to the viral genome as bacterial artificial chromosome recombineering but does so directly in virus-infected cells. CRISPR/Cas9 mutagenesis is not dependent on a bacterial intermediate, and defined viral mutants can be recovered after a limited number of viral genome replications, minimizing the risk of spontaneous mutation.

FIG 1

Targeted cleavage and repair of GPCMV. (A) Sequence of the GP133 CRISPR site selected for pilot studies. The predicted Cas9 cleavage site and PvuI restriction site are shown. (B) Cells transfected with a gRNA expression plasmid targeting GP133 or a nontargeting control (NT) were infected with GPCMV. An ∼582-bp DNA fragment containing the CRISPR site was PCR amplified from infected cells (−) and subjected to either cleavage detection assay (+) or PvuI digestion (P). (C) Titers of virus-containing media collected 4 d p.i. from cells transfected with either NT or GP133-targeting gRNA expression plasmids. (D) Sequence alignment showing 11 unique GP133 mutants compared to the CRISPR site in the WT virus. The type (I, insertion, or D, deletion) and size (in bp) of each mutation is shown, along with whether the GP133 reading frame is maintained.

FIG 2

Knockout of GP133 by CRISPR/Cas9 editing. (A) GP133 was amplified from virus-containing media collected from cells transfected with gRNA expression constructs targeting the 5′ or 3′ end of GP133 plus a GFP expression cassette or a mix of the two gRNA expression constructs and visualized by electrophoresis. (B) GP133 was amplified from plaques isolated by limiting dilution using PCR. Full-length GP133 was expected to produce a 582-bp fragment, while deletion of sequence between the Cas9 cleavage sites would result in a 278-bp PCR product. (C) Candidate GP133 deletion viruses, indicated by asterisked lanes in panel B, were Sanger sequenced. The unique viruses were aligned to the WT sequence. The CRISPR and predicted Cas9 cleavage sites are shown above the alignment. Due to space limitations, 281 bp of sequence between the CRISPR sites in the wild-type sequence is not shown in the figure. The most frequently recovered mutant is indicated by an asterisk.

FIG 3

Epitope-tagging of GP133 by homology-directed repair. (A) Diagram of the homology repair template (HRT) used to introduce the HA-2A-GPT tag to the 3′ end of GP133. Modifications to the GP133 coding sequence (CDS) were as follows: amino acids overlapping the CRISPR site were codon optimized (black arrow) and sequences encoding a 3' HA-tag, a glycine-serine-glycine (GSG) spacer, a 2AT peptide, and a flippase recognition target (FRT)-flanked GPT were added. (B) PCR amplification of passaged virus-containing media from transfection/infection experimentsusing gRNA plasmids targeting the 3′ end of GP133, with or without the HRT. The initial transfection/infections were carried out in the presence or absence of the DNA ligase IV inhibitor SCR7, and the virus was passaged with or without GPT selection. (C) PCR amplification of GP133 from viruses isolated either immediately after transfection/infection or after a single passage in media containing mycophenolic acid xanthine to enrich for recombinant viruses. (D) Immunoblots of cell lysates from mock-infected GPL cells and cells infected with GPCMV GP133-HA-2A-GPT or its parental virus, RFP GPCMV. Primary antibodies against the HA tag, 2A peptide, GPCMV gB, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were used.

FIG 4

CRISPR/Cas9 editing of US22 family genes. (A) GPL cells were transfected with gRNA expression constructs targeting the indicated genes and infected with GPCMV. Targeted genes were amplified (−) directly from virus-containing media at 96 h p.i. and subjected to cleavage detection analysis (+). The predicted sizes for each cleavage detection reaction are outlined in Table S2 in the supplemental material. (B) Multistep growth curve analysis of GPCMV with targeted disruptions in gp144 or gp145 compared to the parental strain (WT). GPL cells were infected at an MOI of 0.01, virus-containing media were collected, and titers were determined at the indicated time points. Data points represent mean titer of duplicate infections.

FIG 5

Summary of unique GPCMV CRISPR/Cas9 mutants recovered. The net size and number of insertion or deletion mutants recovered in this study is shown. Results for mutations that maintain the open reading frame are shown as hatched columns. The number of terminal repeat recombinant (TRR) viruses is also shown.