Mutant/mutation
The mutant expresses (a flag-tagged) CAS9 under control of the strong and constitutive hsp70 promoter. The mutant does not contain a drug-selectable marker.
The flag-cas9 expression cassette is introduced into the silent 230p gene locus of the reference GIMO-PbANKA line (RMgm-687). The GIMO mother line is used for introduction of transgenes into the modified 230p locus through transfection with constructs that target the 230p locus. These constructs insert into the 230p locus (‘gene insertion’), thereby removing the hdhfr::yfcu selectable marker (‘marker out’) from the genome of the mother lines. Transgenic parasites that are marker-free are subsequently selected by applying negative drug selection using 5-FC. This selection procedure is performed in vivo in mice.

Protein (function)
A human codon-optimised spCas9. The RNA-guided endonuclease Cas9 has applied as an efficient gene-editing method in malaria parasite Plasmodium.

Phenotype
FLAG-CAS9 expression in blood stages (nuclear localization); normal growth/multiplication of asexual blood stages.

Additional information
Using Cas9 to precisely engineer a double-strand break (DSB) enhances the efficiency of gene editing in Plasmodium when using a standard length (≤1000 bp) homology region (HR). Since Plasmodium lacks the pathway for canonical non-homologous end-joining (c-NHEJ), any CRISPR-Cas9 mediated edit requires a homology directed repair (HDR) template to facilitate DSB repair. This prohibits the adoption of standard CRISPR-Cas9 disruption screens that rely on c-NHEJ, which introduces insertion and deletion mutations during repair. The HDR template must instead be delivered into the parasite together with the corresponding guide RNA (gRNA) and can be supplied in the same genetic vector that carries the gRNA or on a separate linear or circular DNA molecule. The development of CRISPR-Cas9 screens in Plasmodium requires a scalable system where the gRNA and the HR are physically linked to ensure that each parasite receives a matched gRNA and HR during pooled transfections. Recently, the development of a T. gondii high-throughput tagging CRISPR-Cas9 system, which physically couples the gRNA and HR in a single scalable vector, has paved the way for HDR-mediated CRISPR-Cas9 screens. We have developed a scalable CRISPR system called PbHiT for the rodent malaria parasite P. berghei, which uses a single cloning step to generate targeting vectors with 100 bp homology arms physically linked to a guide RNA (gRNA) that effectively integrate into the target locus.

To develop an improved P. berghei CRISPR-Cas9 editing system, we first modified the existing Plasmodium yoelii pYCm CRISPR-Cas9 vector by replacing the P. yoelii U6 promoter with the endogenous P. berghei U6 promoter to drive gRNA expression. The resulting vector, pPbU6-hdhfr/yfcu-Cas9, encodes Streptococcus pyogenes Cas9 (spCas9) together with the dual positive/negative selection marker human dihydrofolate reductase/yeast cytosine deaminase and uridyl phosphoribosyl transferase (hdhfr/yfcu).

Generation and evaluation of a vector lacking Cas9, pPbU6-hdhfr/yfcu in combination with the PbCas9 parasite line (RMgm-5555).
We introduced a 3xHA tag into the 3’ of rap2/3 (RMgm-5556). The Cas9 expressed from the genome in the new PbCas9 background line facilitated integration of the 3xHA tag using both the one-plasmid and PCR-template approaches (with ~500 bp homology arms). This one-plasmid approach was efficient, with mice reaching a parasitemia above 0.5%, six days post-transfection.

The pPbHiT vector containing the gRNA barcode and 100 bp homology arms efficiently integrates into the target locus
Scaling-up CRISPR-Cas9 in organisms lacking the c-NHEJ pathway requires the gene-specific gRNA and HR to be physically linked in the same plasmid to enable pooled transfections. Having established that the P. berghei genome can be effectively modified using short homology arms in a one-plasmid approach enabled us to adopt a high-throughput tagging strategy previously used in T. gondii. The CRISPR-Cas9 P. berghei High-Throughput strategy (PbHiT) developed here, relies on single-step cloning of a 320 bp synthetic fragment carrying the gRNA and homology arms into the pPbU6-hdhfr/yfcu-HiT (referred to as pPbHiT) vector, which we generated by modifying the pPbU6-hdhfr/yfcu vector. Before transfection into Cas9-expressing parasites, the final pPbHiT vector containing the synthetic fragment is linearised, resulting in the homology arm sequences that drive integration flanking the entire plasmid. To reduce the likelihood of unintegrated plasmids maintained as episomes, the linearised vector was gel extracted prior to transfection. When a Cas9-mediated DSB is repaired through HDR, the entire vector is inserted into the target locus and facilitates the editing of the target gene. The gene-specific gRNA is thereby stably integrated into the genome and serves as a molecular barcode to identify the edited parasites by NGS. The pPbHiT vector can be used for epitope tagging and gene knockout by adapting the position of the homology regions and gRNA. For epitope tagging, the endogenous 3’ untranslated region (UTR) of the target gene is replaced by the 3’UTR of the constitutively expressed gene hsp70.
To test this strategy, the pPbHiT vector containing a triple cMyc (3x-cMyc) epitope tag was used to evaluate the efficiency of tagging genes using 50 and 100 bp homology arms. We targeted both rap2/3 and PBANKA_1224200 (RMgm-5557), a gene predicted to be localised along the secretory pathway. 100 bp homology arms resulted in efficient editing for both targets. Furthermore, the wild type target loci were undetectable by PCR for both targets when using 100 bp homology arms. To further evaluate the performance of the PbHiT system, five other genes were tagged with 3x-cMyc: PBANKA_1225600 (RMgm-5558), PBANKA_0914500 (RMgm-5559), PBANKA_0622900 (RMgm-5560), PBANKA_1451000 (RMgm-5561), and PBANKA_0829400 (RMgm-5562) using 100 bp homology arms and two guides per gene. For PBANKA_0829400 we only obtained one gRNA within the accepted distance from the editing site, which is restricted to the 3’UTR for epitope tagging. Parasites emerged at different days post-transfection (days four to six), with a marked difference between the same gene targeted by the different gRNA, likely reflecting gRNA efficiency. However, this was not directly correlated to either the proximity of the guide to the editing site or the on/off gRNA target score. Genotyping PCRs confirmed the correct integration of the epitope tag in the 3’ UTR of the target gene. One gene (PBANKA_0914500) did not show any 3’ integration product when edited by guide one, however the 5’ integration was confirmed. The expression of all but two of the tagged proteins (PBANKA_1451000 and PBANKA_0829400) was confirmed by Western blot for guide one. PBANKA_0914500-cMyc was detectable by Western blot, despite that it did not show a positive 3’ integration band by PCR. We also confirmed by IFAs that RAP2/3 and PBANKA_0622900 are expressed in schizonts, in agreement with transcriptomic data. This shows that changing the gene’s 3’UTR does not affect mRNA stability and facilitates protein expression at the expected stage. In the case of RAP2/3, the 3x-cMyc tagged protein was observed in the apical end of the parasites, which is consistent with rhoptry localisation29 and demonstrates that the 3x-cMyc tag is not altering protein localisation. We also tested if transfections could be done without gel extracting the final pPbHiT linearised vector and saw no evidence of episomes by PCR.

PbHiT enables pooled vector transfections that recapitulate published knockout phenotypes
Having established the efficiency of the pPbHiT vector for editing single genes, we assessed the performance of the PbHiT system in pooled knockout vector transfections. We selected 12 target genes with in vivo blood-stage growth knockout phenotypes assigned with high confidence in the PlasmoGEM screen and classified as essential (n = 4), dispensable (n = 4), or slow growers (n = 4). Most genes were targeted by two gRNAs except for PBANKA_0515000 (ookinete surface protein P25, p25) and PBANKA_0933700 (mitogen-activated protein kinase 2, map2k), which had one guide each. A total of 22X knockout vectors were individually generated before pooling together in equal amounts, linearised, and transfected into the PbCas9 parasite line. Blood samples were taken at days four to eight post-transfection, genomic DNA was extracted and NGS sequencing libraries were prepared by nested PCR. Results revealed a high concordance between the PbHiT and PlasmoGEM data.

PbHiT facilitates pooled transfection CRISPR screens in P. berghei
We provide evidence of dispensability for four unstudied genes: PBANKA_0103700 (RMgm-5563), PBANKA_0812900 (RMgm-5564), PBANKA_0409000 (RMgm-5565), PBANKA_0821200 (RMgm-5566) that were reported non-mutable in the piggyBac screen but here we were able to knockout using PbHiT

Other mutants