-
Enes Nicolaisen posted an update 6 years, 2 months ago
The first plasmid lacks a protospacer specific to the T. gondii genome, and the Cas9 nuclease has been replaced with a pyrimethamine resistance cassette. The second plasmid only differs from the pU6- SAG1 plasmid in that it lacks the protospacer. All DNA transfections decreased the viability of parasites, although this effect was smallest in wild-type parasites transfected with either of the control plasmids. The DKU80 strain was more susceptible than wild type to transfection with either the control plasmids or pU6-SAG1. Expression of the nuclease without a targeting protospacer had a marginally increased cost to viability that only achieved significance in the DKU80 strain. In both strains, the greatest loss of viability was Vorinostat 149647-78-9 observed upon transfection with pU6-SAG1, which, as opposed to the controls, is expected to efficiently introduce double-stranded breaks at the SAG1 locus. This suggests that loss of viability in response to CRISPR/Cas9 is multifactorial, including a modest effect from nuclease expression that is significantly exacerbated by the presence of a protospacer targeting the genome. At present, we cannot explain the decrease in DKU80 viability upon transfection with the control plasmids. However, consistent with its increased susceptibility to DNA damage, the DKU80 strain may be unable to recover from CRISPR/Cas9-targeted double-stranded breaks in the absence of a template for homologous recombination, effectively resulting in a reduced frequency of observed gene disruption. Current methods for genome engineering of T. gondii rely on the generation of complex constructs for homologous recombination, which yield significant efficiency only in specific parasite strains and require antibiotic selection. This has restricted this type of analysis to a relatively small number of parasite genes. Here, we present a genome engineering method based on the CRISPR/Cas9 system that permits, with a single plasmid, the rapid and efficient disruption of genomic loci. We demonstrate that DKU80 parasite strains are significantly more vulnerable to these targeted disruptions than wild-type parasites, and we postulate that this vulnerability is due to the defect in this strain’s ability to repair double-strand breaks using NHEJ. We further exploit the susceptibility of the DKU80 strain as a genetic background in which to introduce targeted point mutations and epitope tags, and we show that genome editing can be achieved in 15-30% of the manipulated parasite population without the need for any form of selection. Together, these methods enhance our ability to manipulate the T. gondii genome and will enable highthroughput manipulation of a broad range of genetic loci. Available tools for genome editing in T. gondii rely on antibiotic selection and are therefore difficult to compare to CRISPR/Cas9. However, our observations suggest that, beyond the obvious advantage of performing these manipulations in the absence of selection, CRISPR/Cas9 increases the rates of gene disruption by at least 100-fold over methods relying on homologous recombination. We repeatedly observed knockout rates of approximately 20% when targeting CRISPR/Cas9 to the SAG1 locus, which would make it possible to isolate knockouts directly from a transfected population. Not only will this method greatly expedite genetic manipulations, but it will also enable the generation of knockouts with reduced fitness, which would be lost in the course of selection using traditional methods. The single plasmid system used here can be easily reprogrammed to target virtually any genetic locus where a unique 20 bp sequence followed by a PAM can be identified, although ideal sequences should differ at two or more positions from other genomic loci. We have also generated a universal plasmid that allows direct ligation of semi-complementary DNA oligos, thus streamlining the construction of the targeting constructs. We compared the viability of wild-type and DKU80 parasites transfected with CRISPR/Cas9 plasmids or with control plasmids lacking a targeting protospacer in the presence or absence of the Cas9 nuclease. We observed decreased viability in response to transfection of the SAG1-targetting plasmid, although the loss was more pronounced in DKU80 parasites, which exhibited a 96% decrease, compared to the 78% decrease found in wild-type parasites. In both strains, viability was improved in the absence of a targeting protospacer, and further improved in the absence of the nuclease. However, even in the absence of Cas9, the DKU80 strain showed a significant decrease in viability, which we cannot attribute to the effects of CRISPR/Cas9. This effect may be caused by homology of the control plasmids to the genome or a more general response to foreign DNA in the DKU80 strain, although the relatively low rates of homologous recombination in the absence of antibiotic selection do not agree with the significant decrease in viability we report. The adverse effect of plasmid transfection and the predicted susceptibility of the DKU80 strain to double-stranded breaks both contribute to the decrease in gene disruption frequency observed in the absence of NHEJ. It is not surprising that double-stranded DNA breaks are also detrimental to the wild-type strain, given that not all cells will be able to repair the lesions despite the efficiency of NHEJ. Following the initial crisis, parasites that underwent modification by CRISPR/Cas9 showed no secondary phenotypes in growth rate, plaque morphology, or replication that would suggest off-target effects of the CRISPR/Cas9 system. However, as with the severe bottlenecks that occur during antibiotic selection, genetic complementation will need to be performed to account for the presence of secondary mutations or polar effects.