The Malarial Serine Protease SUB1 Plays an Essential Role in Parasite Liver Stage Development

Transmission of the malaria parasite to its vertebrate host involves an obligatory exoerythrocytic stage in which extensive asexual replication of the parasite takes place in infected hepatocytes. The resulting liver schizont undergoes segmentation to produce thousands of daughter merozoites. These are released to initiate the blood stage life cycle, which causes all the pathology associated with the disease. Whilst elements of liver stage merozoite biology are similar to those in the much better-studied blood stage merozoites, little is known of the molecular players involved in liver stage merozoite production. To facilitate the study of liver stage biology we developed a strategy for the rapid production of complex conditional alleles by recombinase mediated engineering in Escherichia coli, which we used in combination with existing Plasmodium berghei deleter lines expressing Flp recombinase to study subtilisin-like protease 1 (SUB1), a conserved Plasmodium serine protease previously implicated in blood stage merozoite maturation and egress. We demonstrate that SUB1 is not required for the early stages of intrahepatic growth, but is essential for complete development of the liver stage schizont and for production of hepatic merozoites. Our results indicate that inhibitors of SUB1 could be used in prophylactic approaches to control or block the clinically silent pre-erythrocytic stage of the malaria parasite life cycle.

. Detection of PbSUB1 in P. berghei blood and liver stage schizonts using an anti-PbSUB1 antibody. (A) SDS extracts of P. berghei ANKA schizonts were probed with a rabbit antibody raised against the putative catalytic domain (residues Ser196-Asn599) of PbSUB1 (PBANKA_110710). By analogy with the known proteolytic maturation profile of parasite-derived P. falciparum SUB1 [1,2], as well as the processing pattern observed with recombinant SUB1 from P. falciparum, P. vivax, P. knowlesi and P. berghei [3,4], the signal at ~70 kDa likely corresponds to full-length unprocessed PbSUB1 zymogen, which includes its N-terminal prodomain. P. falciparum SUB1 undergoes autocatalytic maturation in two steps, involving first loss of the prodomain then a further smaller N-terminal truncation to form a terminal product comprising predominantly the catalytic domain [1,2]. The doublet band at 44-47 kDa therefore likely corresponds to the two processed forms of PbSUB1. (B) Detection of PbSUB1 in infected hepatoma cells. Hepa1-6 cells infected with sporozoites of the marker-free, GFP-expressing P. berghei 507m6cl1 (RMgm-7) clone [5] (also see the  the single-crossover homologous recombination strategy used to fuse a single hemagglutinin (HA) epitope tag (red) to the pbsub1 coding sequence (orange). The targeting plasmid, called pPbSUB1-HA, contained 1,248 bp of targeting sequence (hatched pale orange) fused to the tag, followed by the 3' UTR sequence from the P. berghei dihydrofolate reductase thymidylate synthase (pbdhfr-ts) gene to ensure correct transcription termination and polyadenylation of the modified gene. The presence of the human dihydrofolate reductase (hDHFR) cassette allowed selection of integrant parasites with pyrimethamine. The integration construct was linearised by restriction digestion at a unique Hind III site prior to transfection into the GFP-expressing P. berghei 507m6cl1 (RMgm-7) clone [5]. Positions of hybridization of primers used for diagnostic PCR analysis of the wild-type and modified loci are indicated (red arrows), as well as the predicted size of PCR amplicons (red dotted lines).
Primers used were Fprom_PbSUB1 (a), R_HA (b), F2_PbSUB1 (c), F1_PbSUB1 (d) and R1_3'utr (e) (sequences of all oligonucleotide primers used in this study are provided in 4 Supplemental Table S1 at the end of this file). (B) Southern hybridisation of pulse field gelseparated chromosomes of cloned PbSUB1-HA and control non-transfected parental parasites confirmed integration of the tagging construct into the expected chromosome 11 location. For detection, a 452 bp probe hybridising to the pbdhfr-ts 3' UTR was generated with primers F_3'utr_pbdhfr_probe and R_3'utr_pbdhfr_probe (Table S1). Note that the signal at the position of chromosome 7 observed in both the parental and PbSUB1-HA tracks corresponds to the endogenous pbdhfr-ts gene. The stronger signal observed at chromosome 11 in the PbSUB1-HA lane is due to hybridisation of the probe to both the modified pbsub1 gene as well as the integrated hDHFR cassette, which also contains the pbdhfr-ts 3' UTR (not shown).    Table S1). (B) Southern blot. Asexual blood stage genomic DNA of condSUB1 short , condSUB1 and parental control parasites was digested with Stu I and Nci I and hybridised with a 1.3 kb probe corresponding to an internal segment of pbsub1. (C) Southern hybridisation on pulse field gel-separated chromosomes, using a 452 bp probe corresponding to the pbdhfr-ts 3' UTR. This sequence is present in the endogenous pbdhfr-ts locus as well as in the modified pbsub1 locus, as it is used as a 3' UTR for the flirted pbsub1 gene in all the transgenic parasites. Data are shown only for condSUB1 clone A and the condSUB1 short clone, but were identical for condSUB1 clone B. infected with condSUB1 short parasites were subjected to a temperature shift 18 days post transmission to ensure optimal activity of the FlpL recombinase. Oocysts, salivary glands and sporozoites from these insects displayed strong GFP expression at 26 days (d26), when the insects were allowed to feed on naive mice (bite-back). The resulting blood stage parasites were collected and analysed by PCR using the indicated primer pairs, which were expected to produce a ~1.2 kb amplicon from the non-excised modified pbsub1 locus, or a ~300 bp or a ~600 bp product from the excised locus. Primers used were: (a) F3_PbSUB1 with R_out GFP (product 1203 bp; 'non-excised-specific' band); (b) F_out_hsp70 with R2_out GFP (product 281 bp; 'excised-specific' band); and (c) F_out_hsp70 with R_out GFP (product 605 bp, 'excised-specific' band) (see Table S1 for primer sequences). Results are also shown from analysis of the condSUB1 clone B (right hand side gel). Excision occurred efficiently in the d26 sporozoites, though in this case the non-excised locus was still detectable by PCR. However, as with the condSUB1 clone A (see Figure 2

Generation of recombineering tools for modification of the PbSUB1 locus PCR template pColE1 5'hsp70-ATG-FRT-zeo-pheS-FRT for Step 1
The plasmid p5'hsp70-GFP (a kind gift from R. Menard, Institut Pasteur, Paris) containing GFP under the control of 5' and 3' regulatory sequences from P. berghei hsp70 [7] was linearised immediately downstream of the 5' UTR with Nhe I. A PCR product on the zeo-pheS bacterial selection cassette of plasmid pR6K attR1-zeo-pheS-attR2 was generated using oligonucleotides F_FRT_Zeo_infu_pStep1 and R_FRT_Zeo_infu_pStep1 (Table S1), which contained 5' extensions to introduce FRT sites on either end, followed by 15 bp sequences for in-fusion cloning, and inserted into the linearised p5'hsp70-GFP plasmid.

DNA fragment sub1-HA-attR1-zeo-pheS-attR2-3'sub1 for Step 3
To create a DNA fragment for modifying the 3' end of pbsub1 in the gDNA library clone (Step 3) by Red/ET mediated recombination we placed a bacterial cassette for positive and negative selection in E. coli (zeo-pheS) next to the 3' end of the pbdhfr-ts derived 3' UTR within the HA-tagging vector for pbsub1. For this a PCR amplicon that also retained the attR sites flanking zeo-pheS was generated from plasmid pR6K attR1-zeo-pheS-attR2 using the oligonucleotides F_Zeo_infu_pStep3 and R_Zeo_infu_pStep3 (Table S1). The reverse primer encoded a 5' extension corresponding to 54 bp homologous to the pbsub1 3' UTR, which were followed by a Hind III site and 15 bp extensions for in-fusion cloning into plasmid pPbSUB1-HA, which had been linearised downstream of the pbdhfr-ts 3' UTR with Bst BI.
Digestion of the resulting plasmid with Hind III released a DNA fragment flanked by sequences from the pbsub1 locus that was gel purified and used for Red/ET mediated homologous recombination in E. coli in Step 3.

Gateway donor pR6K attL1-FRT-GFP-hdhfr-yfcu-attL2 for Step 4.
We generated a gateway donor cassette to serve as a universal tool for introducing FRT sites downstream of genes. Using in-fusion cloning we assembled three DNA fragments in a plasmid backbone with a tetracycline resistance cassette and an R6K origin, replication of which is restricted to pir + strains of E. coli, thus preventing replication in E. coli TSA, which we used to propagate the product of the Gateway reaction in Step 4. Between the appropriate attL clonase recognition sites we assembled a promoterless GFP with a terminator from hsp70, which was followed in the same orientation by an expression cassette for hdhfr-yfcu for positive and negative selection in P. berghei. The P. berghei marker contained a promoter from eef1αa and a terminator from hsp70. The directly repeated 3' UTR from hsp70 was designed to allow marker recycling in P. berghei by negative selection against yfcu. The resulting plasmid, pR6K attL1-FRT-GFP-hdhfr-yfcu-attL2, was verified by sequencing.

Red/ET mediated genetic engineering
All in vivo recombination reactions were carried out in 4 ml cultures following the protocol of Pfander et al. [8]. E. coli TSA harbouring the gDNA library clone PbG01-2474a09 that contains the pbsub1 gene, were first electroporated with the recombinase plasmid pSC101gbaA-tet and the culture was grown overnight at 30˚C. The next day, the culture was

Flp recombinase reactions
The plasmid encoding the enhanced Flp recombinase, pSC101 708-Flp-e cm R (Gene Bridges) was used in a similar way to the recombineering plasmid. To induce Flp-e expression and concomitant loss of the plasmid, the temperature was shifted to 37˚C when the culture was in log phase. After induction overnight, the 'flipped' clones were selected on YEG-Cl plates containing p-chlorophenylalanine to select for loss of the zeo-pheS cassette.
After each recombineering and flipping step, the resulting clones were genotyped by PCR and confirmed by nucleotide sequencing. After an overnight culture, miniprep DNA was extracted before being extensively diluted and re-electroporated into TSA cells to eliminate carry-over of undesirable products that might interfere with the next modification step.