Efficient gene replacements in Toxoplasma gondii strains deficient for nonhomologous end joining

Abstract

A high frequency of nonhomologous recombination has hampered gene targeting approaches in the model apicomplexan parasite Toxoplasma gondii. To address whether the nonhomologous end-joining (NHEJ) DNA repair pathway could be disrupted in this obligate intracellular parasite, putative KU proteins were identified and a predicted KU80 gene was deleted. The efficiency of gene targeting via double-crossover homologous recombination at several genetic loci was found to be greater than 97% of the total transformants in KU80 knockouts. Gene replacement efficiency was markedly increased (300- to 400-fold) in KU80 knockouts compared to wild-type strains. Target DNA flanks of only approximately 500 bp were found to be sufficient for efficient gene replacements in KU80 knockouts. KU80 knockouts stably retained a normal growth rate in vitro and the high virulence phenotype of type I strains but exhibited an increased sensitivity to double-strand DNA breaks induced by treatment with phleomycin or gamma-irradiation. Collectively, these results revealed that a significant KU-dependent NHEJ DNA repair pathway is present in Toxoplasma gondii. Integration essentially occurs only at the homologous targeted sites in the KU80 knockout background, making this genetic background an efficient host for gene targeting to speed postgenome functional analysis and genetic dissection of parasite biology.

FIG. 1.

Construction of T. gondii strains in which KU80 is disrupted. (A) Strategy for disrupting the KU80 gene via integration of the HXGPRT marker into strain RHΔhxgprt. Targeting plasmid pΔKUHXFCD targets a ∼4-kb deletion of the KU80 gene (see Materials and Methods). Parasites were selected by positive selection in MPA plus xanthine or by negative selection against the downstream cytosine deaminase (CD) marker in MPA plus xanthine plus flucytosine. Approximate locations of PCR products using primer pairs to verify genotype are depicted (not to scale). The parental strain RHΔhxgprt was positive for PCR 1 (538-bp product) and PCR 2 (639-bp product). A targeted KU80 knockout was positive for the PCR 2 product and was negative for the PCR 1 product. (B) A representative panel of 18 MPA-resistant clones obtained after transfection of plasmid pΔKU80HXFCD and selection in MPA. Clones marked with a * show a pattern consistent with targeted deletion of a region of the KU80 gene (see text). (C) Cleanup of the KU80 locus. The HXGPRT marker was removed from the KU80 locus using the strategy depicted with negative selection in 6TX after transfection of strain RHΔku80::HXGPRT with plasmid pΔKU80B. Approximate locations of PCR products using primer pairs to verify genotype are depicted (not to scale). The parental strain RHΔku80::HXGPRT was positive for PCR 2 (639-bp product) and PCR 6 (373-bp product) and a plasmid pΔKU80B-retargeted KU80 knockout was positive for PCR 6 and was negative for PCR 2. (D) 6TX-resistant clones uniformly had the genotype Δku80 Δhxgprt.

FIG. 2.

Phenotypes of KU80 knockout strains. (A) Virulence of strains was determined by intraperitoneal infection of C57BL/6 mice with 200 tachyzoites. Results of a representative experiment (of two experiments) in which groups of four mice were infected with freshly isolated tachyzoites from strain RHΔhxgprt (triangles), strain RHΔku80::HXGPRT (circles), or strain RHΔku80Δhxgprt (squares). (B) Sensitivity of extracellular tachyzoites to phleomycin. Strains RHΔhxgprt (triangles), RHΔku80::HXGPRT (circles), or RHΔku80Δhxgprt (squares) were treated with phleomycin (two replicate experiments; see text), and survival was determined relative to untreated controls. Results of a representative experiment are shown. (C) Sensitivity of extracellular tachyzoites to γ-irradiation. Strains RHΔhxgprt (triangles), RHΔku80::HXGPRT (circles), and RHΔku80Δhxgprt (squares) were treated with γ-irradiation (two replicate experiments; see text), and survival was determined relative to untreated controls. Results of a representative experiment are shown.

FIG. 3.

Targeted gene replacement at the uracil phosphoribosyltransferase (UPRT) locus. (A) Strategy for disruption of UPRT by a double-crossover homologous recombination event in strain RHΔhxgprt or RHΔku80Δhxgprt by using a fixed 5′ target flank of 1.3 kb and a 3′ target DNA flank of 0.67 kb on plasmid pΔUPT-HXB. The PCR strategy for genotype verification is depicted using primer pairs to assay for products from the PCR (not to scale). (B) PFU assays were performed at various times after transfection to determine the GRF based on the fraction of parasites that had dual resistance to MPA and FUDR (5 μM) compared to the fraction of parasites that were resistant to MPA (Table 2). (C) Genotype verification of clones selected for MPA resistance after transfection with the pΔUPT-HXB plasmid. For parental strain RHΔhxgprt, PCR 1 was positive (304-bp product; lane b), PCR 2 was positive (460-bp product; lane C), and PCR 3 was negative (840-bp product; lane a). For MPA-resistant clones isolated after transfection with plasmid pΔUPT-HXB, PCR 1 was negative, PCR 2 was positive, and PCR 3 was positive. Control lane (c) contained no template and the PCR 1 and PCR 2 primers. Twelve of 12 MPA-resistant clones revealed targeted gene replacement at the UPRT locus.

FIG. 4.

Targeted gene replacement at the carbamoyl phosphate synthetase II (CPSII) locus. (A) Strategy for deletion of ∼24 kb of the endogenous CPSII locus by using a functional CPSII cDNA minigene and a downstream HXGPRT marker flanked by a 1.5-kb 5′ target and a 0.8-kb 3′ target. The 37 exons (shaded rectangles) and 36 introns (white rectangles) of the endogenous CPSII locus are shown (not to scale). PCR primer pairs (PCRs 1 to 4; see Materials and Methods) were used to amplify PCR products to verify genotype (not to scale). In strains RHΔhxgprt and RHΔku80Δhxgprt the expected PCR product size from PCR 1 was 382 bp and from PCR 2* was 1,084 bp. If targeted replacement occurred at the CPSII locus, the expected PCR 2 product would be 363 bp, and no product was produced from PCR 1 because the targeting plasmid CPSII cDNA does not contain the PCR 1 primer sites located in intron 27 (see panel A above). (B) Twelve of 12 MPA-resistant clones revealed targeted gene replacement occurred at the CPSII locus in strain RHΔku80Δhxgprt. To more clearly visualize PCR products those from PCR 1 were resolved for ∼45 min using the markers in lane M¹, and then the same agarose gel was reloaded with PCR products from PCR 2 and using markers in lane M² for the control. Control lane C shows the products from PCR 1 (382-bp product) and PCR 2* (1,084-bp product) using parental RHΔku80Δhxgprt template DNA (endogenous CPSII locus).

FIG. 5.

KU80 knockouts efficiently target gene replacements due to a deficiency in nonhomologous recombination. (A) Strategy for targeted repair of the Δhxgprt locus using different lengths of flanking target DNA that surround a 1.5-kb SalI fragment that is deleted in the Δhxgprt background (see text). (Top) Targeting plasmids with 230-, 450-, 620-, and 910-bp target DNA flanks are shown. Plasmids with 120-bp, 50-bp, or 0-bp targeting DNA flanks are not shown. All pHXH targeting plasmids are MPA sensitive (MPA^s). The termination codon (TAA) of the HXGPRT gene is shown. (Middle) The structure of the Δhxgprt locus for MPA^s strains RHΔhxgprt and RHΔku80Δhxgprt is shown. The gene structure is depicted by exons (dark rectangles), introns (lines), a 5′ UTR (gray rectangle), 3′ UTR (rectangular box with diagonal lines), and a far downstream 3′ UTR (open rectangle). The disrupted locus contains an intact 5′ UTR, intact exon 1 (ATG start is shown) to exon 4, and the deletion in exon 5 and the 3′ UTR contained in the missing 1.5-kb SalI fragment. A hypothetical double-crossover homologous recombination gene targeting event is shown by the dotted lines linking the targeting pHXH plasmids. (Bottom) MPA-resistant (MPA^r) parasites can only arise by the double-crossover homologous recombination gene targeting event shown. (B) The percent maximal homologous recombination was determined in strain RHΔhxgprt (triangles) and strain RHΔku80Δhxgprt (squares) (see Materials and Methods). (C) The percent maximal homologous recombination was determined as a function of targeting DNA concentration.