Summary
Cell cycle transitions are generally triggered by variation in the activity of cyclin-dependent kinases (CDKs) bound to cyclins. Malaria-causing parasites have a life cycle with unique cell-division cycles, and a repertoire of divergent CDKs and cyclins of poorly understood function and interdependency. We show that Plasmodium berghei CDK-related kinase 5 (CRK5), is a critical regulator of atypical mitosis in the gametogony and is required for mosquito transmission. It phosphorylates canonical CDK motifs of components in the pre-replicative complex and is essential for DNA replication. We also provide evidence for indirect regulation of the concomitant M-phase progression. During a replicative cycle, CRK5 stably interacts with a single Plasmodium-specific cyclin (SOC2), although we obtained no evidence of SOC2 cycling by transcription, translation or degradation. Our results provide evidence that during Plasmodium male gametogony, this unique cyclin/CDK pair fills the functional space of multiple eukaryotic cell-cycle kinases controlling DNA replication and M-phase progression.
Introduction
Progression through the cell cycle critically relies upon post-translational mechanisms including changes in activity of cell cycle kinases and phosphatases, and ubiquitin-mediated degradation of specific components once their function is complete. Central components of these networks are the cyclin-dependent kinases (CDKs). Many different families of CDK exist1. In most model organisms however, those primarily regulating DNA replication and mitosis belong to a monophyletic family that includes human CDK1-CDK4/6, Cdc2 and CDC28 in fission and budding yeasts, respectively, and CDKA in plants2,3. In malaria parasites, seven CDKs have been described4–6. Of these, five are divergent CDK-related kinases (CRK), i.e. they are clearly related to CDKs but have no clear orthologues in the yeast and human kinomes. These include CRK5 (PBANKA_123020), which forms a distinct branch within the CDK cluster with atypical sequence motifs in the activation loop and an additional 148 amino acid C-terminal extension found only in Plasmodium7. The gene is required, but not essential, for proliferation of asexual blood stages of the human parasite P. falciparum7.
The primary regulator of CDK activity is the cyclin subunit. Cyclins were originally named because of their oscillation in level that reached a threshold required to drive cell-cycle transition8,9. Cyclins are now defined as a family of evolutionarily related proteins encoding a cyclin box motif that is required for binding to the CDK catalytic subunit3. Plasmodium genomes contain no sequence-identifiable G1-, S-, or M-phase cyclins, and only three proteins have sequence homology with cyclin family members in other eukaryotes10,11. Cyc1 is important for cytokinesis in P. falciparum blood stage replication, possibly regulating the CDK7 homolog MRK12. P. berghei Cyc3 is dispensable for blood-stage replication but important for oocyst maturation in the mosquito midgut11. The paucity of putative cyclins and the diversity of CDKs, has led to suggestions that some of these kinases function without a cyclin partner13.
The malaria parasite has several proliferative phases in its life cycle. Male gametogony or gametogenesis is a proliferative sexual stage in the mosquito vector that is essential for parasite transmission. Circulating mature male gametocytes in the vertebrate host are arrested in a G0-like phase and resume development in the mosquito midgut following a blood meal, activated by the presence of xanthurenic acid (XA) and a drop in temperature14. In about ten minutes the haploid male gametocyte completes three rounds of genome replication and closed endomitosis, assembles the component parts of eight axonemes, and following nuclear division, produces eight flagellated motile male gametes in a process called exflagellation15. The organisation and regulation of the cell cycle during male gametogony is unclear. Current evidence suggests that certain canonical cell-cycle checkpoints are absent16. For example, compounds that interfere with mitotic spindle formation do not prevent DNA replication proceeding17,18, while spindle formation is not affected in a mutant that is unable to replicate DNA18,19. Recently, we observed that proteins involved in DNA replication and cytoskeletal reorganisation are phosphorylated during the first seconds of gametogony20. Interestingly, CRK5 clusters with both groups of proteins, suggesting it has a key regulatory role during male gametogony20. Here, we show CRK5 is part of a unique and divergent CDK/cyclin complex required for progress through male gametogony and essential for parasite transmission.
Results
CRK5 is a key regulator of gametogony and sporogony in the mosquito
Previous attempts to disrupt P. berghei crk5 had suggested the gene is essential for asexual blood-stage proliferation5. However, using long sequence homology regions to replace crk5 with a T. gondii DHFR/TS resistance marker (Fig. 1A and Fig. S1A), we obtained resistant parasites and cloned them following a two-step enrichment. The gene deletion in the resulting CRK5-knockout (KO) clone was confirmed by PCR (Fig. S1B) and RNAseq analysis (Fig. 1A and Table S1). There was no significant growth defect during erythrocytic asexual multiplication (Fig. 1B), nor an inability to produce morphologically normal gametocytes (Fig. 1C). However, upon XA activation only a few microgametocytes formed active exflagellation centres (Fig. 1D). While no major transcriptional changes were detected (Fig. 1E), a subtle trend in increased expression of multiple regulators of gametogony was observed in CRK5-KO gametocytes (Table S1). Quantitative PCR analysis confirmed these observations but in most cases, changes were not statistically significant (Fig. 1F). To test the possibility that the exflagellation defect was due to multiple subtle transcriptional changes, we used an additional approach to study the effect of rapid CRK5 degradation in mature gametocytes. We tagged the endogenous crk5 with an AID/HA epitope tag (Fig. S1B) to degrade the fusion protein in presence of auxin in a strain expressing the Tir1 protein21 (Fig. 1G). Addition of the AID/HA tag to the CRK5 C-terminus imposed a significant fitness cost, with a 2-fold decrease in exflagellation in the absence of auxin, but importantly, depletion of CRK5-AID/HA by auxin treatment of mature gametocytes resulted in a dramatic reduction in exflagellation (Fig. 1H).
Since a residual number of CRK5-KO male gametocytes produced active exflagellation centres, we asked whether the resulting microgametes were fertile by measuring the number of ookinetes per activated female gametocyte. Consistent with the dramatic reduction in male gametogony, CRK5-KO parasites showed a significant reduction in ookinete conversion compared to the parental control line (Fig. 1I). This reduction led to a > 30-fold decrease in the number of CRK5-KO oocysts present on the midgut epithelium of Anopheles stephensi mosquitoes allowed to feed on infected mice (Fig. 1J). CRK5-GFP is expressed in early oocysts (Fig. S1C and S1D) and residual CRK5-KO ookinetes produced non-viable oocysts (Fig. 1K), suggesting an additional role during sporogony and important for parasite transmission (Fig. 1L). Altogether, these results demonstrate a critical requirement for CRK5 during male gametogony and an essential function for CRK5 in parasite development and transmission.
Phosphoproteome kinetics point to direct phosphorylation of the pre-replicative complex by CRK5
To elucidate the processes regulated by CRK5 during gametogony, we compared the proteomes and phosphoproteomes of WT and CRK5-KO gametocytes over the first minute of gametogony covering the first S-phase and metaphase of mitosis I (Fig. 2A). Apart from a significant CRK5 reduction in the CRK5-KO line, no significant differences were observed at the proteome level between mutant and parental lines (Table S2). Similarly, non-activated CRK5-KO gametocytes showed limited differences in phosphorylation compared to the parental counterpart. However, 20 phosphopeptides showed significant down-regulation upon activation (Fig. 2B) with most enriched GO terms corresponding to proteins important in DNA replication including the origin recognition complex (ORC - Fig. 2C and Table S3). This group includes a protein with similarities to the licensing factor CDT1 (chromatin licensing and DNA replication factor 1, PBANKA_1356300), a possible orthologue of the DNA replication factor CDC6 (PBANKA_1102900), and two ORC components (ORC2 - PBANKA_0803000 - and ORC4 – PBANKA_1348800). Interestingly, eight down-regulated phosphorylation sites map to the S/T*PxK CDK consensus motif (Fig. 2D). This motif shows a significant > 600 fold enrichment compared to all other phosphorylated sequences identified in this study (p-value<1.10−14).
Multiple phosphopeptides were more abundant following activation (Fig. 2B) suggesting rapid secondary effects on CRK5 phospho-dependent pathways, probably via the regulation of protein phosphatases. Consistently, the phosphatase PPM11 in the mutant line showed a significantly different phosphorylation profile before and after activation. Components of the ribosome were the most significantly enriched (Fig. 2C), suggesting CRK5 can contribute indirectly to the activation of translation that is de-repressed upon gametocyte activation22. Interestingly, four phosphopeptides from dynein subunits were also more abundant, suggesting that the microtubule-based movement required during gametogony may also be dependent on CRK5.
The set of enriched GO terms associated with CRK5-regulation is similar to those associated with calcium dependent protein kinase (CDPK) 4 and CRK4 (Fig. S2), two kinases previously shown to be crucial for DNA replication and its initiation during gametogony19 and schizogony23, respectively. When compared with a set of human kinase-associated GO terms, the CRK5 set shows similarities to those of CDK1 and CDK2 as well as other cell-cycle kinases including Check1, ATM and Aurora B. Altogether this phosphoproteomic survey suggests that CRK5 is a genuine CDK that likely phosphorylates components of the pre-replicative complex. It may also indirectly regulate phosphorylation of proteins involved in translation and microtubule-based movement.
CRK5 is required for both S- and M- phases during P. berghei gametogony
Given CRK5 similarity to canonical CDKs and its possible role in initiating DNA replication, we quantified the DNA content by flow cytometry of microgametocytes lacking crk5 or upon CRK5-AID/HA degradation mediated by auxin. In both cases, we observed an average 2-fold fewer diploid (2N) gametocytes 1 min post-activation (pa) with XA (Fig. 2E), demonstrating a dependency on CRK5 for DNA replication during male gametogony.
It has been suggested that M-phase is initiated independently of S-phase completion during Plasmodium gametogony19,20. As multiple dynein-related proteins are phosphorylated in a CRK5-dependent manner, we quantified the formation of spindle-like structures and axonemes in CRK5-KO parasites. Cells depleted of CRK5 are unable to correctly assemble or maintain the spindle, with a greater than two-fold decrease of visible spindles at 1 min pa (Fig. 2F). Tubulin staining at 15 min pa showed that axoneme formation is not affected but axonemal beating is not initiated (Fig. 2G). Ultrastructural analysis by electron microscopy confirms the typical (9+2) axoneme structure however at 6 min pa there appears to be more nuclear poles present in WT compared to mutant parasites (0.56/cell compared to 0.32/cell; random sample of 50 cells of each). The cytoplasm contained electron dense basal bodies, which were often, but not always, associated with nuclear poles (Figs 2Hb and g). CRK5 deletion appeared to prevent activated microgametocytes developing to form multiple nuclear poles and they showed no chromatin condensation, cytokinesis or flagellum formation (Fig. 2H).
These results indicate that CRK5 is a critical primary regulator of DNA replication initiation and progression and, possibly, an indirect regulator of microtubule-based processes required for spindle and nuclear pole formation in the nucleus, but not for axoneme assembly in the cytosol.
CRK5 is part of an atypical nuclear cyclin/CDK complex
The binding of cyclins to CDKs is required for progression through cell-cycle phases in model organisms. In the absence of clear S- and M-phase cyclin homologues, we searched for possible CRK5 regulatory subunits. To do this we tagged CRK5 with GFP or 3xHA at its C-terminus by modification of the endogenous gene (Fig. S3A), to facilitate localisation studies and identify any proteins bound to it. The tagged protein was localised to the nucleus in non-activated gametocytes and at 1 min pa showed an additional location at the mitotic spindle (Fig. 3A and Fig. S3D). Mass-spectrometry analyses of affinity-purified tagged protein identified 234 proteins (Table S4). Consistent with a role for CRK5 in the initiation of DNA replication, GO term enrichment analyses of proteins co-purifying with CRK5 highlighted ORC, minichromosome maintenance (MCM), RepC and alpha DNA polymerase primase complexes (Fig. 3B and C).
One of the most abundant proteins co-purified with CRK5 was SOC2 (PBANKA_144220), a protein we had previously found to be important for cell-cycle progression during gametogony20. SOC2 has very low homology (<13 %) with a cyclin domain in a protein of the starfish Asterina pectinifera10, but the protein lacks key residues across most of the cyclin box10 and has no detectable cyclin-like function in in vitro biochemical assays11. Analysis of CRK5 immunoprecipitates also identified multiple peptides from PBANKA_082440 (CDK regulatory subunit, CDKrs), a protein related to Cks1 and CksHs2, two CDK-associated proteins in Saccharomyces cerevisiae24 and human25, respectively. To confirm the putative interaction between CRK5, SOC2 and CDKrs, we tagged both SOC2 and CDKrs at the C-terminus by integration of 3×HA19 or GFP coding sequence into the endogenous gene (Fig. S3B) and affinity purified the tagged proteins. Reciprocal enrichment of CRK5, SOC2 and CDKrs components was observed, consistent with the formation of a complex (Fig. 3C). In addition, each protein also co-purified the ORC, MCM, RepC and alpha DNA polymerase primase complexes (Table S4).
Previous bioinformatics analyses had identified only three bona fide Plasmodium cyclins11. Here, the identification of SOC2 as a putative CRK5 binding partner suggests it is also a cyclin. We revisited the phylogenetic analyses of apicomplexan cyclins and CDKs, and found that CRK5 and CDKrs are conserved in Haemosporidia, Coccidia and Piroplasma, whilst SOC2 is found only in Haemosporidia (Fig. 3D and Fig. S3C).
CRK5, SOC2 and CDKrs have a similar location and complementary functions during gametogony
Given the biochemical evidence that CRK5, SOC2 and CDKrs form a complex, we wanted to compare their subcellular location. Live fluorescence microscopy identified CDKrs-GFP in the nucleus of non-activated microgametocytes and throughout male gametogony, with a spindle-like localisation during mitoses, as found for CRK5 (Fig. 3E). By indirect immunofluorescence of fixed cells, we confirmed the pattern of SOC2-HA distribution (Fig. 3F), as previously described19.
SOC2-KO microgametocytes showed defects in the ploidy transitions during gametogony19. In agreement with the notion of complementary roles for CRK5, SOC2 and CDKrs, deletion of both SOC2 and CDKrs encoding genes (Fig. S3E) also resulted in a rapid and severe reduction in exflagellation (Fig. 3G). The defect was already apparent at 1 min pa, with a >3-fold reduction of cells with 2N DNA content (Fig. 3H) and a 5-fold reduction in mitotic spindles (Fig. 3I). Ultrastructural analysis by electron microscopy of SOC2-KO and CDKrs-KO activated gametocytes confirms an early arrest during male gametogony as observed in CRK5-KO parasites (Fig. S3F). Altogether, these data strongly suggest that the CRK5/SOC2/CDKrs complex is a functional CDK/cyclin complex that controls entry into S-phase and possibly controls progression through the following or concurrent M-phase during gametogony. We propose the name ‘CSC complex’ for this CRK5/SOC2/CDKrs complex.
SOC2 expression does not follow a temporal cyclin pattern during gametogony and the CSC complex is stable during the first round of mitosis
Cyclin turnover mediated by the ubiquitin-dependent proteasome pathway causes oscillations in CDK activity and therefore we wished to examine the stability of the CSC complex. Western blot and fluorescent microscopy analysis revealed that CRK5, SOC2 and CDKrs are expressed in non-activated gametocytes (Fig. 3A, 3E and 4A) and therefore we examined whether the CSC complex is degraded during the course of gametogony. Western blot analysis of CRK5-HA and SOC2-HA over the 10 minutes of gametogony revealed no evidence for degradation of either protein (Fig. 4A). To confirm that levels of CRK5 and CDKrs are maintained during gametogony in single cells, we monitored by flow cytometry the fluorescence of gametocytes expressing CRK5-GFP or CDKrs-GFP during gametogony. This analysis confirmed that there was no change in the level of CRK5-GFP or CDKrs-GFP fluorescence following male gametocyte activation and gametogony (Fig. S4A).
To examine whether the apparent stable levels of SOC2-HA and CRK5-HA result from a balance of de novo translation and protein degradation, we pre-treated non-activated gametocytes with the proteasome inhibitor MG132 for one hour. Although MG132 significantly reduced exflagellation (Fig. S4B), it did not affect DNA replication (Fig. S4C) nor lead to a significant accumulation of SOC2, CRK5, or CDKrs as assessed by western blot (Fig. 4A) and flow cytometry (Fig. S4A). Together, these results suggest no oscillation in levels of the CSC complex resulting from differential temporal translation and degradation during gametogony.
CDKs are usually characterised by low activity in the absence of a bound cyclin, therefore we reasoned that dynamic assembly of the CSC complex might underlie CRK5 regulation. Development of activated gametocytes is highly synchronous and we used this property to investigate the interaction between CRK5 and SOC2 in the first minute after activation. We immunoprecipitated SOC2-HA and CRK5-HA at 0, 15, 30 and 60 seconds pa and measured the relative abundance of immunoprecipitated proteins in the CSC complex by mass-spectrometry (Table S4). We detected no difference in the levels of SOC2 or CRK5 protein/protein interactions over this time course (Fig. 4B). However, we noticed that a CDT1-like protein became less abundant in CRK5 and SOC2 immunoprecipitates following gametocyte activation (Fig. 4C). Pre-treatment of gametocytes with BKI-294, a CDPK4 inhibitor prevented this reduction. These results suggest that SOC2-CRK5 binding is not dynamic during the first round of DNA replication. However, a CDT1-like protein is either degraded or released from the complex following CDPK4 activation.
CSC is dynamically phosphorylated during the first round of replication
Reversible phosphorylation of cyclin-CDK complexes regulates kinase activity and therefore we examined whether this mechanism may regulate the activity of the CSC complex. We had previously identified 13 phosphorylated SOC2 residues20, with S5088/5089 phosphorylated in a CDPK4-dependent manner during early gametogony19. Here, S5088 was also phosphorylated in a CDPK4-dependent manner in the immunoprecipitated SOC2 (Fig. S4D), but S5088/5089 substitution with alanine (Fig. S4E) did not cause a significant defect in DNA replication or exflagellation (Fig. S4F and 4G) suggesting that SOC2-S5088/5089 phosphorylation is either not required or insufficient to regulate CSC complex function.
CRK5 is also phosphorylated dynamically during gametogony20. We show here that inhibition of CDPK4 with BKI-1294 affected the phosphorylation at positions S654 and S693 of immunoprecipitated CRK5 (Fig. 4D). Interestingly, all eight detected phosphorylation sites in CRK5 are located in the 148-residue C-terminal extension (Fig. 4E), which is only found in Plasmodium CRK57. We considered that this extension may have an important role in the regulation of CRK5 activity, and to test this hypothesis we generated a parasite clonal line, CRK51-562, with the last 148 amino acids of CRK5 deleted, leaving an intact catalytic domain (Fig. S4H). This CRK5 truncation produced the same phenotype as that of the CRK5-KO line, with a significant reduction in exflagellation, DNA replication and mitotic spindles (Fig. 4F and G). These results suggest the Plasmodium-specific C-terminus extension of CRK5 is essential for its function and may act as a regulatory platform to coordinate cell cycle progression during gametogony.
Discussion
In this study, we have identified an atypical cyclin-CDK pair that is essential for male gametogony in a malaria parasite. The divergent CDK-related kinase, CRK5, interacts with a Plasmodium-restricted cyclin, SOC2, and with another putative regulatory subunit to form a complex we have named CSC. Given our poor understanding of the mechanisms that regulate the rapid DNA synthesis and the three rounds of mitosis, followed by chromosome condensation, nuclear division and exflagellation of gametes that comprise male gametogony, it is unclear whether CSC is required for a single checkpoint upstream of S-phase or for two independent checkpoints, necessary to enter S- and M-phases, respectively. We provide evidence of direct phosphorylation by CRK5 of multiple components of the pre-replicative complex, but further studies are necessary to understand the consequences of CRK5-dependent phosphorylation. The abundance of phosphorylated dyneins increased upon CRK5 deletion, suggesting that CRK5-dependent regulation of M-phase may be indirect, affecting microtubule-dependent processes such as formation or segregation of nuclear poles or mitotic spindle organisation26,27 (Fig. 4H).
CRK5, SOC2 and CDKrs have similar transcriptional profiles during asexual replication in erythrocytes with a peak in late trophozoites28, suggesting the CSC complex may have conserved control functions in cell cycle progression during schizogony. Consistent with this idea, CRK5 deletion reduced nuclear division during erythrocytic schizogony in P. falciparum7, while SOC2 deletion led to slower proliferation of P. berghei in erythrocytes19. Given the versatile nature of cyclin-CDK complex function in other organisms, it is possible that other CDKs or cyclins partially compensate for the loss of CRK5 or SOC2 in asexual blood stages. For example, CRK4 activity may compensate for the lack of CRK5 during schizogony, because it is also important for DNA replication during this developmental stage23.
Usually there are significant oscillations in the abundance of cyclins during the cell cycle. However, SOC2 shows no evidence of oscillation or changes in the level of its interaction with CRK5. Such changes may be undetectable in the experimental conditions and we cannot exclude more subtle changes that may regulate the CSC complex. The observed absence of cycling may also reflect the constraints imposed by the extremely rapid nature of gametogony. No marked Cyc1 oscillations were detected during the slower P. falciparum blood stage schizogony and the Cyc1/MRK complex was detected in mature segmented schizonts after Cyc1 is required to complete cytokinesis29. Therefore, it is also possible that non-oscillating cell-cycle cyclins and stable cyclin/CDK complexes are a conserved feature of the various cycles of division within Plasmodium parasites. However, PbCyc3 showed fluctuating expression patterns during oocyst development11. It thus remains unclear whether non-cycling cyclins and stable cyclin/CDK complexes represent conserved features of the various cell cycles of Plasmodium parasites.
We provide evidence that the C-terminal extension of CRK5 is important in the control of its function. Interestingly, the C-terminal extensions of both SOC2 and CRK5 are Haemosporidia-specific. For example, they are absent in the related Apicomplexa parasite Toxoplasma gondii in which the homologue of CRK5 is also involved in DNA replication13, highlighting a specific evolution of this cyclin/CDK system in Plasmodium. It is tempting to speculate that the CRK5 C-terminal extension and its phosphorylation are important in the regulation of CRK5 activity by controlling its interaction with SOC2, but this hypothesis remains untested. Very few studied CDKs contain additional domains with known function. For example, phosphorylation of the Cdk16 N-terminal domain blocks binding to cyclin Y30. Multiple CRK5 and SOC2 residues are phosphorylated in a CDPK4-dependent manner and it is possible that CDPK4 works as a Plasmodium-specific link to translate a calcium signal triggering gametogony to prime the CSC complex (Fig. 4H). In addition, CDPK4-independent phosphorylation of CRK5 is also observed20 and other kinases or phosphatases may control CSC activity during the rounds of DNA replication and mitosis. However, given the absence of canonical checkpoint kinases5,6,31, it is possible that the successive mitotic cycles of gametogony are separated from each other by timing rather than by checkpoints that rely on signals from one event in the cell cycle to regulate the next.
Plasmodium species lack clear CDK orthologues3 that are usually highly conserved across Eukarya and indispensable in yeasts and humans32. This study indicates that CRK5, despite its divergence from canonical CDKs, acts as a genuine CDK by binding a cyclin, phosphorylating targets at classical CDK motifs and regulating proteins that are targeted by conventional CDKs in human and yeast. However, it remains difficult to identify a specific direct functional orthologue due to functional similarities with multiple CDKs and even non-CDK cell-cycle kinases, and it is tempting to speculate that CRK5 and SOC2 cover the functional space of multiple cell-cycle regulators in other organisms, to satisfy the specificities of the Plasmodium cell cycle.
Methods
Ethics statement
The animal work performed in the United Kingdom passed an ethical review process and was approved by the UK Home Office. Work was carried out under UK Home Office Project Licenses (40/3344 and 30/3248) in accordance with the United Kingdom ‘Animals (Scientific Procedures) Act 1986’. All animal experiments performed in Switzerland were conducted with the authorisation numbers GE/82/15 and GE/41/17, according to the guidelines and regulations issued by the Swiss Federal Veterinary Office.
Parasite maintenance and transfection
P. berghei ANKA strain33-derived clones 2.3434, 507cl1, 820cl135, and 6156, together with derived transgenic lines, were grown and maintained in CD1 outbred mice. Six to ten week-old mice were obtained from Charles River laboratories, and females were used for all experiments. Mice were specific pathogen free and subjected to regular pathogen monitoring by sentinel screening. They were housed in individually ventilated cages furnished with a cardboard mouse house and Nestlet, maintained at 21 ± 2 °C under a 12 hr light/dark cycle, and given commercially prepared autoclaved dry rodent diet and water ad libitum. The parasitaemia of infected animals was determined by microscopy of methanol-fixed and Giemsa-stained thin blood smears.
For gametocyte production, parasites were grown in mice that had been phenyl hydrazine-treated three days before infection. One day after infection, sulfadiazine (20mg/L) was added in the drinking water to eliminate asexually replicating parasites. Microgametocyte exflagellation was measured three or four days after infection by adding 4 μl of blood from a superficial tail vein to 70 μl exflagellation medium (RPMI 1640 containing 25 mM HEPES, 4 mM sodium bicarbonate, 5% fetal calf serum (FCS), 100 μM xanthurenic acid, pH 7.4). To calculate the number of exflagellation centres per 100 microgametocytes, the percentage of red blood cells (RBCs) infected with microgametocytes was assessed on Giemsa-stained smears. For gametocyte purification, parasites were harvested in suspended animation medium (SA; RPMI 1640 containing 25 mM HEPES, 5% FCS, 4 mM sodium bicarbonate, pH 7.20) and separated from uninfected erythrocytes on a Histodenz/Nycodenz cushion made from 48% of a Histodenz/Nycodenz stock (27.6% [w/v] Histodenz/Nycodenz [Sigma/Alere Technologies] in 5.0 mM TrisHCl, 3.0 mM KCl, 0.3 mM EDTA, pH 7.20) and 52% SA, final pH 7.2. Gametocytes were harvested from the interface. To induce CRK5-AID/HA degradation, 500 μM auxin dissolved in ethanol was added to purified gametocytes for one hour.
Schizonts for transfection were purified from overnight in vitro culture on a Histodenz cushion made from 55% of the Histodenz/Nycodenz stock and 45% PBS. Parasites were harvested from the interface and collected by centrifugation at 500 g for 3 min, resuspended in 25 μL Amaxa Basic Parasite Nucleofector solution (Lonza) and added to 10-20 μg DNA dissolved in 10 μl H2O. Cells were electroporated using the FI-115 program of the Amaxa Nucleofector 4D. Transfected parasites were resupended in 200 μl fresh RBCs and injected intraperitoneally into mice. Parasite selection with 0.07 mg/mL pyrimethamine (Sigma) in the drinking water (pH ~4.5) was initiated one day after infection. Each mutant parasite was genotyped by PCR using three combinations of primers, specific for either the WT or the modified locus on both sides of the targeted region (experimental designs are shown in Supplemental Figures). For allelic replacements, sequences were confirmed by Sanger sequencing using the indicated primers. Controls using wild type DNA were included in each genotyping experiment; parasite lines were cloned when indicated.
Generation of DNA targeting constructs
The oligonucleotides used to generate and genotype the mutant parasite lines are in Table S5.
Restriction/ligation cloning
The C-termini of CRK5 and CDKrs were tagged with GFP by single crossover homologous recombination in the parasite. To generate these GFP lines, a region of the crk5 or cdkrs gene downstream of the ATG start codon was amplified using primers T1851/T1852 and T2461/T2462, respectively, ligated into p277 vector, and transfected in ANKA line 2.34 as described previously36. Diagnostic PCRs were performed with primer 1 (IntT185 or intT246) and primer 2 (ol492) to confirm integration of the GFP targeting constructs. Schematic representations of the endogenous crk5 and cdkrs loci, the constructs and the recombined loci are found in Fig. S1 and S3 respectively.
The gene-deletion targeting vector for crk5 was constructed using the pBS-DHFR plasmid, which contains polylinker sites flanking a T. gondii dhfr/ts expression cassette conferring resistance to pyrimethamine, as described previously5. PCR primers N1001 and N1002 were used to generate a fragment of crk5 5’ upstream sequence from genomic DNA, which was inserted into ApaI and HindIII restriction sites upstream of the dhfr/ts cassette of pBS-DHFR. A fragment generated with primers N1003 and N1004 from the 3’ flanking region of crk5 was then inserted downstream of the dhfr/ts cassette using EcoRI and XbaI restriction sites. The linear targeting sequence was released using ApaI/XbaI and the construct was transfected into the ANKA line 507cl1 expressing GFP. Diagnostic PCR was performed with primer 1 (IntN100) and primer 2 (ol248) to confirm integration of the targeting construct, and primer 3 (N100 KO1) and primer 4 (N100 KO2) were used to confirm deletion of the crk5 gene. A schematic representation of endogenous crk5, the constructs and the recombined locus is in Fig. S1.
PlasmoGEM vectors
3×HA and AID/3×HA tagging of CRK5, CDKrs-KO, CRK51-562 and SOC2 allelic replacement construct were generated using phage recombineering in Escherichia coli TSA strain with PlasmoGEM vectors (http://plasmogem.sanger.ac.uk/). For final targeting vectors not available in the PlasmoGEM repository, generation of knock-out and tagging constructs was performed using sequential recombineering and gateway steps as previously described37,38. For each gene of interest (goi), the Zeocin-resistance/Phe-sensitivity cassette was introduced using oligonucleotides goi HA-F × goi HA-R and goi KO-F × goi KO-R for 3×HA tagging and KO targeting vectors, respectively. Insertion of the GW cassette following gateway reaction was confirmed using primer pairs GW1 × goi QCR1 and GW2 × goi QCR2. The modified library inserts were then released from the plasmid backbone using NotI. The CRK5-AID/HA targeting vector was transfected into the 615 parasite line21, the CDKrs-KO, and CRK51-562 vectors into the 820cl1 line and the CRK5-HA vector into the 2.34 line.
Substitutions of SOC2S5088/5089A residues were introduced with a two-step strategy using λ Red-ET recombineering as described in reference39. The first step involved the insertion by homologous recombination of a Zeocin-resistance/Phe-sensitivity cassette flanked by 5’ and 3’ sequences of the codon of interest, which is amplified using the soc2-delF × soc2-delR primer pair. Recombinant bacteria were then selected on Zeocin. The recombination event was confirmed by PCR and a second round of recombination replaced the Zeocin-resistance/Phe-sensitivity cassette with a PCR product containing the S5088/5089A substitutions amplified using soc2-mut-F with soc2-mutR primer pairs, respectively. Bacteria were selected on YEG-Cl kanamycin plates. Mutations were confirmed by sequencing vectors isolated from colonies sensitive to Zeocin with primers soc2-seqF to soc2-seqR. The modified library insert was then released from the plasmid backbone using NotI and the construct transfected into the 801cl1 line.
Parasite phenotype analyses
Blood containing approximately 50,000 parasites of the CRK5-KO line was injected intraperitoneally into mice to initiate infections. Four to five days post infection, exflagellation and ookinete conversion were examined as described previously36 with a Zeiss AxioImager M2 microscope (Carl Zeiss, Inc) fitted with an AxioCam ICc1 digital camera. When indicated, gametocytes were pre-treated 5 min in SA supplemented with 1 μM BKI-1294 or 60 min in SA supplemented with 1 μM MG132. To analyse mosquito transmission, 30 to 50 Anopheles stephensi SD 500 mosquitoes were allowed to feed for 20 min on anaesthetized, infected mice whose asexual parasitaemia had reached 15% and were carrying comparable numbers of gametocytes as determined on Giemsa-stained blood films. To assess mid-gut infection, approximately 15 guts were dissected from mosquitoes on day 7, 14, and 21 post-feeding and oocysts were counted using the Zeiss AxioImager M2 microscope and a 63× oil immersion objective. Mosquito bite-backs were performed 21 days post-feeding using naive mice and blood smears were examined after 3-4 days.
Immunofluorescence labelling
Gametocyte immunofluorescence assays were performed as previously described40. For HA and α-tubulin staining, purified cells were fixed with 4% paraformaldehyde and 0.05% glutaraldehyde in PBS for 1 h, permeabilised with 0.1% Triton X-100/PBS for 10 min and blocked with 2% BSA/PBS for 2 h. Primary antibodies were diluted in blocking solution (rat anti-HA clone 3F10, 1:1000; mouse anti-α-tubulin clone DM1A, 1:1000, both from Sigma-Aldrich). Anti-rat Alexa594, anti-mouse Alexa488, anti-rabbit Alexa 488, Anti-rabbit Alexa594 were used as secondary antibodies together with DAPI (all from Life technologies), all diluted 1:1000 in blocking solution. Confocal images were acquired with a LSM700 or a LSM800 scanning confocal microscope (Zeiss).
The CRK5-KO gametocytes were purified and activated in ookinete medium, then fixed at 15 min pa with 4% paraformaldehyde (PFA, Sigma) diluted in microtubule stabilising buffer (MTSB) for 10-15 min and added to poly-L-lysine coated slides. Immunocytochemistry was performed using primary mouse anti-α-tubulin mAb (Sigma-T9026; 1:1000) and secondary antibody Alexa 488 conjugated anti-mouse IgG (Invitrogen-A11004; 1:1000). The slides were then mounted in Vectashield 19 with DAPI (Vector Labs) for fluorescence microscopy. Parasites were visualised on a Zeiss AxioImager M2 microscope fitted with an AxioCam ICc1 digital camera (Carl Zeiss, Inc).
Flow cytometry analysis of gametocyte DNA content and CRK5/CDKrs-GFP fluorescence
DNA content of microgametocytes was determined by flow cytometry measurement of fluorescence intensity of cells stained with Vybrant dye cycle violet (life Technologies). Gametocytes were purified and resuspended in 100 μl of SA. Activation was induced by adding 100 μl of modified exflagellation medium (RPMI 1640 containing 25 mM HEPES, 4 mM sodium bicarbonate, 5% FCS, 200 μM xanthurenic acid, pH 7.8). To rapidly block gametogony, 800 μl of ice cold PBS was added and cells were stained for 30 min at 4⁰C with Vybrant dye cycle violet and analysed using a Beckman Coulter Gallios 4. Microgametocytes were selected on fluorescence by gating on GFP positive microgametocytes when the 820cl1 parasite line or its derivatives were used. In this case, cell ploidy was expressed as a percentage of male gametocytes only. When the 2.34 line or its derivatives were analysed, gating was performed on both micro- and macro-gametocytes and cell ploidy was expressed as a percentage of all gametocytes. For each sample, >50,000 cells were analysed.
Electron microscopy
Gametocytes samples at 1, 3, 6, 15 and 30 min pa were fixed in 4% glutaraldehyde in 0.1 M phosphate buffer. Samples were post fixed in osmium tetroxide, treated en bloc with uranyl acetate, dehydrated and embedded in Spurr’s epoxy resin. Thin sections were stained with uranyl acetate and lead citrate prior to examination in a JEOL1200EX electron microscope (Jeol UK Ltd).
Phylogenetic analyses
Annotated protein sequences were downloaded for the following organisms: P. falciparum (pfal), P. reichenowi (prei), P. knowlesi (pkno), P. vivax (pviv), P. berghei (pber), P. chabaudi (pcha), and P. yoelii (all from PlasmoDB, [plasmodb.org]). Theileria annulata (tann), T. parva (tpar), and Babesia bovis (bbov) (all from PiroplasmaDB, [piroplasmadb.org]). Toxoplasma gondii (tgon), Neospora caninum (ncan), and Eimeria tenella (eten) (all from ToxoDB, [toxodb.org]). Cryptosporidium hominis (chom), C. parvum (cpar), Chromera velia (cvel), and Vitrella brassicaformis (vbra) (all from CryptoDB, [cryptodb.org]). Perkinsus marinus (pmar), Ichthyophthirius multifiliis (imul), and Tetrahymena thermophilia (tthe) (all from EnsemblProtists, [protists.ensembl.org]). Symbiodinium minutum (smin) (from OIST, [marinegenomics.oist.jp]).
Reciprocal best hits to P. berghei CRK sequences were determined from pair-wise protein BLAST searches. For each CRK protein, reciprocal best BLAST matches were retrieved and aligned using mafft41. Alignments were trimmed using trimal42 and HMM models were produced with HMMer (hmmer.org). These models were then used to search the genomes for which no reciprocal best BLAST hit was detected.
A neighbour-joining tree was then constructed using four combined data sets: 1) the P. berghei CRK proteins, 2) all protein sequences with a reciprocal best BLAST hit to P. berghei, 3) genes that were best BLAST hit for a P. berghei protein, but did not have the CRK protein as best hit in P. berghei, 4) all protein sequences with a score below 1×10−12 to an HMM model. Sequences and identifiers are provided in the source data files of Fig. S3.
Sequence distances were calculated using clustalw243,44, which was also used to construct the NJ tree. Protein sequences from 1) and 2) were divided into their CRK groups (i.e. CRK1-7, CDKrs, and SOC2), and all pair-wise distances were calculated within each CRK group. For all protein sequences from 3) and 4), the distance to all members in all CRK groups were then measured. A protein sequence was included in a CRK group if the average distance between the protein sequence and all members of a given CRK group did not exceed 40% of all pair-wise distances within that CRK group, or if the CRK group to which a protein sequence had the lowest average distance was lower than the distance to the CRK group with the second-lowest average distance minus 5 standard deviations. Final trees were produced by PhyML45,46 with the GTR substitution model selected by SMS47. Branch support was evaluated with the Bayesian-like transformation of approximate likelihood ratio test, aBayes48. Trees were visualized using FigTree (tree.bio.ed.ac.uk/software/figtree/).
Transcriptomics
Gametocytes (non-activated −0 min- and activated −15 min-) were collected from CRK5-KO and WT lines. Total RNA was isolated from purified parasites using an RNeasy purification kit (Qiagen). RNA was vacuum concentrated (SpeedVac) and transported using RNA stable tubes (Biomatrica). Strand-specific mRNA sequencing was performed from total RNA using TruSeq Stranded mRNA Sample Prep Kit LT (Illumina) according to the manufacturer’s instructions. Briefly, polyA+ mRNA was purified from total RNA using oligo-dT dynabead selection. First strand cDNA was synthesised primed with random oligos followed by second strand synthesis where dUTPs were incorporated to achieve strand-specificity. The cDNA was adapter-ligated and the libraries amplified by PCR. Libraries were sequenced in an Illumina Hiseq machine with paired-end 100bp read chemistry.
RNA-seq read alignments were mapped onto the P. berghei ANKA genome (May 2015 release in GeneDB— http://www.genedb.org/) using Tophat2 (version 2.0.13) with parameters “-library-type fr-firststrand–no-novel-juncs–r 60”. Transcript abundances were extracted as raw read counts using the Python script ‘HTseq-count’49 (model type – union, http://www-huber.embl.de/users/anders/HTSeq/). Counts per million (cpm) values were obtained from count data and gene were filtered if they failed to achieve a cpm value of 1 in at least 30% of samples per condition. Library sizes were normalized by the TMM method using EdgeR software50 and further subjected to linear model analysis using the voom function in the limma package51. Differential expression analysis was performed using DESeq2 in R version 3.2.152. Genes with fold change greater than 2 and p-value less than 0.05 were considered as significantly differentially expressed. P. berghei GO terms (Gene Ontology) were downloaded from GeneDB (http://www.genedb.org/; May 2015 release) and gene ontology enrichment analysis was performed for the DEG lists using GOstats R package. All analyses and visualizations were done with R packages-cummeRbund and ggplot2.
Quantitative Real Time PCR (qRT-PCR) analyses
RNA was isolated from purified gametocytes using an RNA purification kit (Stratagene). cDNA was synthesised using an RNA-to-cDNA kit (Applied Biosystems). Gene expression was quantified from 80 ng of total RNA using SYBR green fast master mix kit (Applied Biosystems). All the primers were designed using primer3 (Primer-blast, NCBI). Analysis was conducted using an Applied Biosystems 7500 fast machine with the following cycling conditions: 95°C for 20 s followed by 40 cycles of 95°C for 3 s; 60°C for 30 s. Three technical replicates and two biological replicates were performed for each assayed gene. The hsp70 (PBANKA_081890) and arginyl-t RNA synthetase (PBANKA_143420) genes were used as endogenous control reference genes. The primers used for qPCR can be found in Table S5.
Protein immunoprecipitation and identification
Sample preparation
Co-immunoprecipitations (IPs) of protein were performed with purified gametocytes. For HA-based IPs, samples were fixed for 10 min with 1% formaldehyde. Parasites were lysed in RIPA buffer (50 mM Tris HCl pH 8, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) and the supernatant was subjected to affinity purification with anti-HA antibody (Sigma) or anti-GFP antibody (Invitrogen) conjugated to magnetics beads. Beads were re-suspended in 100 μl of 6 M urea in 50 mM ammonium bicarbonate (AB). Two μl of 50 mM dithioerythritol (DTE) were added and the reduction was carried out at 37°C for 1h. Alkylation was performed by adding 2 μl of 400 mM iodoacetamide for 1 h at room temperature in the dark. Urea was reduced to 1 M by addition of 500 μl AB and overnight digestion was performed at 37 °C with 5 μl of freshly prepared 0.2 μg/μl trypsin (Promega) in AB. Supernatants were collected and completely dried under speed-vacuum. Samples were then desalted with a C18 microspin column (Harvard Apparatus) according to manufacturer’s instructions, completely dried under speed-vacuum and stored at −20°C.
Liquid chromatography electrospray ionisation tandem mass spectrometry (LC-ESI-MSMS)
Samples were diluted in 20 μl loading buffer (5% acetonitrile [CH3CN], 0.1% formic acid [FA]) and 2 μl were injected onto the column. LC-ESI-MS/MS was performed either on a Q-Exactive Plus Hybrid Quadrupole-Orbitrap Mass Spectrometer (Thermo Fisher Scientific) equipped with an Easy nLC 1000 liquid chromatography system (Thermo Fisher Scientific) or an Orbitrap Fusion Lumos Tribrid mass Spectrometer (Thermo Fisher Scientific) equipped with an Easy nLC 1200 liquid chromatography system (Thermo Fisher Scientific). Peptides were trapped on an Acclaim pepmap100, 3 μm C18, 75 μm × 20 mm nano trap-column (Thermo Fisher Scientific) and separated on a 75 μm × 250 mm (Q-Exactive) or 500 mm (Orbitrap Fusion Lumos), 2μm C18, 100 Å Easy-Spray column (Thermo Fisher Scientific). The analytical separation used a gradient of H2O/0.1% FA (solvent A) and CH3CN/0.1 % FA (solvent B). The gradient was run as follows: 0 to 5 min 95 % A and 5 % B, then to 65 % A and 35 % B for 60 min, then to 10 % A and 90 % B for 10 min and finally for 15 min at 10 % A and 90 % B. Flow rate was 250 nL/min for a total run time of 90 min.
Data-dependant analysis (DDA) was performed on the Q-Exactive Plus with MS1 full scan at a resolution of 70,000 Full width at half maximum (FWHM) followed by MS2 scans on up to 15 selected precursors. MS1 was performed with an AGC target of 3 × 106, a maximum injection time of 100 ms and a scan range from 400 to 2000 m/z. MS2 was performed at a resolution of 17,500 FWHM with an automatic gain control (AGC) target at 1 × 105 and a maximum injection time of 50 ms. Isolation window was set at 1.6 m/z and 27% normalised collision energy was used for higher-energy collisional dissociation (HCD). DDA was performed on the Orbitrap Fusion Lumos with MS1 full scan at a resolution of 120,000 FWHM followed by as many subsequent MS2 scans on selected precursors as possible within a 3 sec maximum cycle time. MS1 was performed in the Orbitrap with an AGC target of 4 × 105, a maximum injection time of 50 ms and a scan range from 400 to 2000 m/z. MS2 was performed in the Ion Trap with a rapid scan rate, an AGC target of 1 × 104 and a maximum injection time of 35 ms. Isolation window was set at 1.2 m/z and 30% normalised collision energy was used for HCD.
Database searches
Peak lists (MGF file format) were generated from raw data using the MS Convert conversion tool from ProteoWizard. The peak list files were searched against the PlasmoDB_P.berghei ANKA database (PlasmoDB.org, release 38, 5076 entries) combined with an in-house database of common contaminants using Mascot (Matrix Science, London, UK; version 2.5.1). Trypsin was selected as the enzyme, with one potential missed cleavage. Precursor ion tolerance was set to 10 ppm and fragment ion tolerance to 0.02 Da for Q-Exactive Plus data and to 0.6 for Lumos data. Variable amino acid modifications were oxidized methionine and deamination (Asn and Gln) as well as phosphorylated serine, threonine and tyrosine. Fixed amino acid modification was carbamidomethyl cysteine. The Mascot search was validated using Scaffold 4.8.4 (Proteome Software) with 1% of protein false discovery rate (FDR) and at least 2 unique peptides per protein with a 0.1% peptide FDR.
Proteomics and phosphoproteomics
Sample preparation (SDS buffer-FASP procedure)
Activated and non-activated purified gametocytes were snap frozen in liquid nitrogen. Cell lysis was performed in 1 ml of 2% SDS, 25 mM NaCl, 2.5 mM EDTA, 20 mM TCEP, 50 mM TrisHCl (pH 7.4), supplemented with 1x Halt™ protease and phosphatase inhibitor (Thermo Fisher). Samples were vortexed and then heated at 95°C for 10 min with mixing at 400 rpm on a thermomixer. DNA was sheared via four sonication pulses of 10 s each at 50% power. Samples were then centrifuged for 30 min at 17,000 g and the supernatant was collected. Three-hundred μl sample was incubated with 48 μl 0.5 M iodoacetamide for 1h at room temperature. Protein was digested based on the FASP method53 using Amicon® Ultra-4, 30 kDa cut-off as centrifugal filter units (Millipore). Trypsin (Promega) was added at 1:100 enzyme/protein ratio and digestion was performed at 37°C overnight. The resulting peptides were treated as described above.
TMT10plex-labelling
Peptide concentration was determined using a colorimetric peptide assay (Pierce). A pool made from 1/20 of each sample was created and used as reference for each tandem mass tag (TMT) experiment. Briefly, 100 μg of each sample was labelled with 400 μg of corresponding TMT-10plex reagent previously dissolved in 110 μl 36% CH3CN, 200 mM 4-(2-hydroxyethyl)-1-piperazinepropanesulfonic acid (EPPS, pH 8.5). The reference sample was labelled with TMT11-131C label reagent. The reaction was performed for 1h at room temperature and then quenched by adding hydroxylamine to a final concentration of 0.3% (v/v). Labelled samples of each TMT10 experiment were pooled together, dried and desalted with a peptide desalting spin column (Pierce) according to manufacturer’s instructions.
Phosphopeptide enrichment
Phosphopeptides were enriched using the High-Select Fe-NTA Phosphopeptide Enrichment Kit (Thermo Fisher Scientific) following manufacturer’s instructions. The phosphopeptide fraction as well as the flow-through fraction were desalted with a C18 macrospin column (Harvard Apparatus) according to manufacturer’s instruction and then completely dried under speed-vacuum.
High pH Reverse-Phase fractionation
The flow-through fraction of each TMT10 experiment was fractionated into 13 fractions using the Pierce™ High pH Reversed-Phase Peptide Fractionation Kit (Thermo Fisher Scientific) according to manufacturer’s instructions.
LC-ESI-MSMS
For all samples, peptide concentration was determined using a colorimetric peptide assay (Pierce). Phosphopeptides were reconstituted in loading buffer (5% CH3CN, 0.1% FA) and 2 μg injected on to the column. LC-ESI-MS/MS was performed on an Orbitrap Fusion Lumos Tribrid mass spectrometer (Thermo Fisher Scientific) equipped with an Easy nLC1200 liquid chromatography system (Thermo Fisher Scientific). Peptides were trapped on a Acclaim pepmap100, C18, 3 μm, 75 μm × 20 mm nano trap-column (Thermo Fisher Scientific) and separated on a 75 μm × 500 mm, C18, 2 μm, 100 Å Easy-Spray column (Thermo Fisher Scientific). The analytical separation was run for 125 min using a gradient of 99.9% H2O/ 0.1% FA (solvent A) and 80% CH3CN/0.1% FA (solvent B). The gradient was run as follows: 0 to 2 min 92 % A and 8 % B, then to 72 % A and 28 % B for 105 min, to 58% A and 42% B for 20 min, and finally to 5 % A and 95 % B for 10 min with a final 23 min at this composition. Flow rate was of 250 nL/min. DDA was performed with MS1 full scan at a resolution of 120,000 FWHM followed by as many subsequent MS2 scans on selected precursors as possible within a 3 sec maximum cycle time. MS1 was performed in the Orbitrap with an AGC target of 4×105, a maximum injection time of 50 ms and a scan range from 375 to 1500 m/z. MS2 was performed in the Orbitrap using HCD at 38% Normalised collision Energy (NCE). Isolation window was set at 0.7 u with an AGC target of 5×104 and a maximum injection time of 86 ms. A dynamic exclusion of parent ions of 60 s with 10 ppm mass tolerance was applied.
High pH Reversed-Phase Peptide fractions were reconstituted in the loading buffer (5% CH3CN, 0.1% FA) and 1 μg was injected onto the column. LC-ESI-MS/MS was performed as described above. The analytical separation was run for 90 min using a gradient of 99.9% H2O/0.1% FA (solvent A) and 80% CH3CN/0.1%FA (solvent B). The gradient was run as follows: 0 to 5 min 95 % A and 5 % B, then to 65 % A and 35 % B for 60 min, and finally to 5% A and 95% B for 10 min with a final 15 min at this composition. Flow rate was 250 nL/min. DDA was performed using the same parameters, as described above.
Database search
Raw data were processed using Proteome Discoverer (PD) 2.3 software (Thermo Fisher). Briefly, spectra were extracted and searched against the P berghei ANKA database (PlasmoDB.org, release 38, 5076 entries) combined with an in-house database of common contaminants using Mascot (Matrix Science, London, UK; version 2.5.1). Trypsin was selected as the enzyme, with one potential missed cleavage. Precursor ion tolerance was set to 10 ppm and fragment ion tolerance to 0.02 Da. Carbamidomethyl of cysteine (+57.021) as well as TMT10plex (+229.163) on lysine residues and on peptide N-termini were specified as fixed modifications. Oxidation of methionine (+15.995) as well as phosphorylated serine, threonine and tyrosine were set as variable modifications. The search results were validated with a Target Decoy PSM validator. PSM and peptides were filtered with a FDR of 1%, and then grouped to proteins, again with a FDR of 1% and using only peptides with high confidence level. Both unique and razor peptides were used for quantitation and protein and peptides abundances value were based on signal to noise (S/N) values of reporter ions. Abundances were normalised on “Total Peptide Amount” and then scaled with “On Controls Average” (i.e. using the reference sample channel). All the protein ratios were calculated from the medians of the summed abundances of replicate groups and associated p-values were calculated with an ANOVA test based on individual protein or peptides. GO term enrichment was performed using PlasmoDB. The reference set of GO terms for human kinases was obtained from 23.
Statistical analysis
All statistical analyses were performed using GraphPad Prism 8 (GraphPad Software). Statistical tests are mentioned in the figure legends.
Data availability
All relevant data are available from the authors on request. Mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD017283 (proteomics and phosphoproteomics) and PXD017308 (HA immunoprecipitations). RNAseq data have been deposited to the Gene Expression Omnibus under accession number GSE144743.
Supplementary material
Figure S1. Generation of CRK5-KO, CRK5-AID/HA, and CDKrs-GFP transgenic lines and CRK5-GFP expression in oocysts. A-C. Genetic modification vectors and genotyping data for CRK5-KO (A), CRK5-AID/HA (B), and CRK5-GFP (C). Oligonucleotides used for PCR genotyping are indicated and agarose gels for corresponding PCR products from genotyping reactions are shown. C. Western blot detection of CRK5-GFP. D. CRK5-GFP expression and localisation in midgut oocysts at day 7, 14, 17, and 21 post-infection. Scale bar = 10 μm.
Figure S2. Percentage of GO terms shared between CRK5 and a set of Plasmodium and human kinases. For CRK5 GO terms, proteins with a >2 fold change (p-value <0.05) in the CRK5-KO were retained.
Figure S3. Identification and characterisation of the CSC complex components. A-B. Genetic modification vectors and genotyping data for CRK5-HA (A) and CDKrs-GFP (B). Oligonucleotides used for PCR genotyping are indicated and agarose gels for corresponding PCR products from genotyping reactions are shown. C. Maximum likelihood tree of the CRK, SOC2 and CDKrs proteins detected across Apicomplexa (see Methods). Branch support values evaluated with aBayes are shown on branches. D. Localisation of CRK5-HA in 1 min activated gametocytes. Scale bar = 5 μm. E. Genetic modification vectors and genotyping data for CDKrs-KO. F. Electron micrographs of the male gametocytes of the SOC2-KO (a, e), CDKrs-KO (b, f), CRK5-KO (c, g) and wild type (d,h) parasites at 15 min pa. The three KO parasites consist of predominately early stages (95-100%) while the WT showed a number of late stages (~40%) with exflagellation and free microgametes. Bars represent 1 μm in a-d and 100 nm in all other micrographs. (a to c) SOC2-KO, CDKrs, and CRK5 male gametocytes showing the early stage of development with a lobed nucleus (N) and cytoplasmic axonemes (A). Inserts. Detail of cross-sectioned axonemes. (d) WT showing a late stage in development with nucleus (N) containing small clumps of condensed chromatin (Ch). Note the free microgamete flagella (F). Insert. Cross-sectioned axoneme. (e) Detail of a similar stage parasite to that in (a) showing a nuclear pole (NP) but note the disorganized arrangement of spindle microtubules (Mt) and attached kinetochores (K). (f) Detail of the nucleus in (b) showing the nuclear pole (NP) with a closely associated basal body (BB) in the cytoplasm. K – Kinetochore. (g) Detail of a similar stage to that in (c) showing the nuclear pole (NP) and associated basal body (BB). (h) Detail of a similar parasite to that in (d) undergoing exflagellation showing the flagellum (F) protruding from the microgametocyte. Ch – condensed chromatin. Insert. Cross section through a free microgamete showing the flagellum (F) and electron dense nucleus (N).
Figure S4. Study of the CSC complex regulation. A. Flow cytometry analysis of CRK5-GFP and CDKrs-GFP fluorescence over the course of gametogony in haploid or multiploid (>1N) cells in the presence or absence of 1 μM of the proteasome inhibitor MG132 (error bars show standard deviation from the mean; 3 or 4 independent infections; two-way ANOVA). B. Pre-treatment of gametocytes with 1 μM MG132 leads to reduced exflagellation (error bars show standard deviation from the mean; technical replicates from 3 independent infections; unpaired two-tailed t-test). C. Treatment with 1 μM MG132 has no effect on the number of male gametocytes replicating their DNA. The proportion of male gametocytes undergoing DNA replication was determined at 1 min pa and is expressed as a percentage of cells that are polyploid (>1 nucleus) (error bars show standard deviation from the mean; technical replicates from 3 independent infections; unpaired two-tailed t-test). D. SOC2 phosphorylation levels show that S5088 is phosphorylated in a CDPK4-dependent manner 15 sec after activation. ND = not detected. BKI = CDPK4 inhibitor. E. Genetic modification vector and genotyping data for SOC2S5088/5089A mutagenesis. Oligonucleotides used for PCR genotyping are indicated and agarose gels for corresponding PCR products from genotyping reactions are shown. F. S5088/5089A substitution in SOC2-3xHA does not impair exflagellation (error bars show standard deviation from the mean; technical replicates from 2 independent infections; two-way ANOVA). G. There is no reduction in the number of male gametocytes replicating their DNA in SOC2S5088/5089A-3×HA gametocytes. The proportion of male gametocytes undergoing DNA replication was determined at 1 min pa and is expressed as a percentage of cells that are polyploid (>1N) (error bars show standard deviation from the mean; from 2 independent infections; two-way ANOVA). H. Genetic modification vector and genotyping data for CRK5 mutagenesis. Oligonucleotides used for PCR genotyping are indicated and agarose gels for corresponding PCR products from genotyping reactions are shown.
Table S1. RNAseq data.
Table S2. Peptides detected in WT and CRK5-KO gametocytes.
Table S3. Phosphopeptides detected in WT and CRK5-KO gametocytes.
Table S4. Proteins identified in co-immunoprecipitates by mass-spectrometry.
Table S5. Oligonucleotides used in this study.
Acknowledgements
We thank Julie Rodger (Nottingham Universitiy) for her assistance in the insectary maintenance and Zineb Rchiad (KAUST) for RNAseq library preparation. We thank the excellent service at the bioimaging and flow-cytometry core facilities at the Faculty of Medicine of the University of Geneva. We also would like to thank Nisha Philip (University of Edinburgh) for sharing the 615 Tir1-expressing line as well as Wesley Van Voorhis and Kayode Ojo (University of Washington) for sharing compound BKI-1294. We thank Markus Ganter for sharing the reference GO term set used in this work.
This work was supported by the Swiss National Science Foundation grant BSSGI0_155852 and 31003A_179321 to MB. MB is an INSERM and EMBO young investigator. The work in RT lab is supported by Medical Research Council UK (G0900109, G0900278, MR/K011782/1) and Biotechnology and Biological Sciences Research Council (BB/N017609/1). The work in AB lab is supported by a faculty baseline fund (BAS/1/1020-01-01) and a Competitive Research Grant (CRG) award from OSR (OSR-2018-CRG6-3392) from the King Abdullah University of Science and Technology.