Kinesin-8B controls basal body function and flagellum formation and is key to malaria parasite transmission

Eukaryotic flagella are conserved microtubule-based organelles that drive cell motility. Plasmodium, the causative agent of malaria, has a single flagellate stage: the male gamete in the mosquito. Three rounds of endomitotic division together with an unusual mode of flagellum assembly rapidly produce eight motile gametes. These processes are tightly coordinated but their regulation is poorly understood. To understand this important developmental stage, we studied the function and location of the microtubule-based motor kinesin-8B, using gene-targeting, electron microscopy and live cell imaging. Deletion of the kinesin-8B gene showed no effect on mitosis but disrupted 9+2 axoneme assembly and flagellum formation during male gamete development and also completely ablated parasite transmission. Live cell imaging showed that kinesin-8B-GFP did not colocalise with kinetochores in the nucleus but instead revealed dynamic, cytoplasmic localisation with the basal bodies and the assembling axoneme during flagellum formation. We thus uncovered an unexpected role for kinesin-8B in parasite flagellum formation that is vital for the parasite life cycle.


Introduction
Eukaryotic flagella, also known as motile cilia, are conserved microtubule (MT)-based organelles that protrude from cells and drive motility of single cells, or the movement of fluid across ciliated tissue (Mirvis et al., 2018). In a number of organisms, the mechanisms by which flagella are built and maintained are gradually being revealed, and defects in flagellum assembly and function are associated with numerous human diseases (Baker and Beales, 2009;Croft et al., 2018) In general, flagella grow from mother centrioles -also known as basal bodies -which are recruited from the centrosome to act as templates for the characteristic nine-fold symmetry of the MT array known as the flagellum axoneme (Mirvis et al., 2018).
Specialised axonemal dynein motors are arranged according to the underlying axonemal MT organisation and these motors power flagella beating (Lin and Nicastro, 2018).Thus, flagella-driven motility depends on axonemal organisation defined by basal body function. However, the mechanisms that coordinate all these facets of flagella action are complex and less well understood.
Malaria is a disease caused by the unicellular parasite Plasmodium spp., which infects many vertebrates and is transmitted by female Anopheles mosquitoes (WHO, 2018).
The parasite has a complex life cycle, with distinct asexual and sexual developmental stages. Extracellular, motile and invasive stages move by gliding motility powered by an actomyosin motor, except for male gametes which use flagellar movement (Boucher and Bosch, 2015;Frenal et al., 2017;Sinden et al., 1976;Sinden et al., 2010). Male gamete development from haploid gametocyte is a very rapid process with endomitotic division and flagellum formation completed in 10 to 15 minutes . Specifically, three successive rounds of DNA replication produce an 8N nucleus and are accompanied by synthesis and assembly of basal bodies and axonemes; then cytokinesis and exflagellation release eight haploid flagellated male gametes (Sinden et al., 1976).
The Plasmodium male gamete has a very simple structure with no subcellular organelles except an axoneme with associated dyneins, elongated nucleus and surrounding flagellar membrane (Creasey et al., 1994;Okamoto et al., 2009;Sinden et al., 1976). The axoneme -essential for flagellar unit motility -consists of a pair of central MTs (C1 and C2) encircled by nine doublet MTsa so-called 9+2 organisation -as in many other organisms (Mitchell, 2004). However, the Plasmodium flagellum differs in several key respects in the mechanism of axoneme formation from that of other organisms including trypanosomes, Chlamydomonas and humans (Bastin, 2010;Briggs et al., 2004;Soares et al., 2019), as well as other Apicomplexa with flagellated gametes like Toxoplasma and Eimeria (Ferguson, 2009;Ferguson et al., 1977;Ferguson et al., 1974). The differences in the mechanism of axoneme formation are: 1) the Plasmodium basal body, from which the flagellum is assembled, does not exhibit the classical nine triplet MT (9+0) (Mirvis et al., 2018), but instead consists of an electron dense amorphous structure within which nine peripherally arranged MTs can sometimes be resolved (Francia et al., 2015); 2) basal bodies/centrioles are not present during asexual stages of the Plasmodium life cycle, unlike Toxoplasma and Eimeria where they play a role during asexual division (Ferguson DJP, 2013), but assemble de novo in male gametocytes (Sinden et al., 1976); 3) the axoneme itself is assembled from the basal bodies within the male gametocyte cytoplasm, and only after its assembly does the axoneme protrude from the membrane to form of the flagellum itself; as a result and very unusually, axoneme construction occurs independently of intra-flagellar transport (IFT) (Briggs et al., 2004;Sinden et al., 1976).
Mechanisms that regulate Plasmodium male gamete biogenesis and control axoneme length are largely unknown, although the involvement of some individual proteins has been described. For example, PF16 is a flagellar protein with an essential role in flagellar motility and the stability of the central MT pair in Chlamydomonas and Plasmodium (Smith and Lefebvre, 1996;Straschil et al., 2010), while SAS6, a wellconserved basal body protein, has a role in Plasmodium flagellum assembly (Marques et al., 2015;Nigg and Stearns, 2011). Proteomic analysis has revealed many potential regulators of axoneme assembly present in male gametes, including members of the kinesin superfamily (Talman et al., 2014). Kinesins are molecular motor proteins that have essential roles in intracellular transport, cell division and motility (Cross and McAinsh, 2014;Verhey and Hammond, 2009;Wittmann et al., 2001). Some kinesin families are known to contribute to ciliary structure and function, either by transporting ciliary components along axonemal MTs by IFT like kinesin-2s (Scholey, 2008), or by destabilizing MTs, like kinesin-9 and kinesin-13 (Blaineau et al., 2007;Dawson et al., 2007).
Here we have analysed the location and function of the kinesin-8B protein (PBANKA_0202700) present in male gametes in the rodent malaria parasite, Plasmodium berghei. Deletion of the gene results in complete impairment of 9+2 axoneme formation because of misregulation of the basal body, and no effect on nuclear division. Its unprecedented involvement in basal body function prevents assembly of axonemes and the motile flagellum, and thereby blocks parasite transmission. Live cell imaging showed that kinesin-8B-GFP is expressed only during male gamete development and that strikingly, it is not associated with the nuclear kinetochore but is present only in the cytoplasm. Here, kinesin-8B-GFP dynamically associated with the basal bodes and with axonemes during their formation and assembly, consistent with an unexpected role in flagellum genesis.

Kinesin-8B is essential for male gamete formation and its deletion blocks parasite transmission
Based on its known expression in male gametes and a potential function during male gamete development, we first assessed the function of kinesin-8B throughout the Plasmodium life cycle by using a double crossover homologous recombination strategy to delete its gene (S1A Successful deletion of the kinesin-8B gene indicated that it is not essential for asexual blood stage development, consistent with the protein's expression and localisation only during male gametogenesis, and with global functional studies (Bushell et al., 2017). Phenotypic analysis of the Δkinesin-8B parasite was carried out at other developmental stages, comparing the wild type (WT-GFP) parasite with two independent gene-knockout parasite clones generated by two different transfections. Since Δkinesin-8B parasites undergo asexual blood stage development, form gametocytes in mice, and exhibit no change in morphology or parasitemia, an essential role for kinesin-8B during these stages is unlikely. We then examined male and female gametocytes following activation in exflagellation/ookinete medium for the development and emergence of male and female gametes. This analysis revealed that there was no exflagellation in either of the Δkinesin-8B parasite clones (Clone 1 and clone 4), whereas exflagellation frequency for the WT-GFP parasites was normal ( Figure 1A).
To investigate this defect further, we examined both zygote formation and ookinete Δkinesin-8B parasites neither at 14 days post infection (dpi) nor 21 dpi ( Figure 1D).

Kinesin-8B is not required for DNA replication but its deletion impairs male exflagellation
To understand the defect in more detail, we analysed the development of MT structures in Δkinesin-8B and WT-GFP male gametocytes following activation in vitro by treatment with xanthurenic acid and decreased temperature at 0, 8 or 15 minutes post activation (mpa). We observed no very clear early differences, but the MT distribution in the Δkinesin-8B gametocytes gradually diverged from that of the WT-GFP parasites over time, and these mutants did not form flagella (as evidenced from tubulin labelling) at 15 mpa (arrow, Figure 1E). To assess the effect of the mutations on DNA replication during male gametogenesis, we analysed the DNA content (N) of Δkinesin-8B and WT-GFP male gametocytes by fluorometric analyses after DAPI staining. We observed that Δkinesin-8B male gametocytes had a haploid DNA content (1N) at 0 min (non-activated) and were octaploid (8N) 8-10 min post activation, similar to WT-GFP, indicating the absence of kinesin-8B had no effect on DNA replication ( Figure 1F).

Δkinesin-8B phenotype
To investigate further the defect in male gamete development in Δkinesin-8B parasites, we analysed the transcript profile of both non-activated and activated Δkinesin-8B and control WT-GFP gametocytes. RNAseq analysis was performed on WT-GFP and Δkinesin-8B gametocytes, just before activation (0 mpa) and after exflagellation (30 mpa)) for four pairs of biological replicates (WT, 0 mpa; WT, 30 mpa; Δkinesin-8B, 0 mpa; and Δkinesin-8B, 30 mpa) and the read coverages exhibited Spearman correlation coefficients of 0.97, 0.98, 0.96, and 1.00, respectively, demonstrating the reproducibility of this experiment. As expected, no significant reads mapped to the Δkinesin-8B region (S1D Fig). Only 7 genes were upregulated, and 40 genes were downregulated compared to the WT-GFP control, but none was known to be important in gametogenesis ( Figure 1G, S1 Table). In a qRT-PCR analysis of a specific set of genes coding for proteins identified either as molecular motors (kinesin family) or involved in male gamete development (Billker et al., 2004;Ferguson DJP, 2013;Tewari et al., 2005) (Figure 1H) showed no significant change in transcript level.
These data suggest that the observed phenotype resulted directly from the absence of kinesin-8B rather than through an indirect effect on another gene transcript. In summary, deletion of the kinesin-8B gene did not appear to affect the initiation or growth of single and duplet MTs but association of basal bodies with the nuclear pole appeared to be reduced and coordinated formation of the 9+2 organisation of functional axonemes did not occur. This defect prevented exflagellation and the formation of viable male gametes.

Spatio-temporal profiles of kinesin-8B and the kinetochore protein Ndc80 show that kinesin-8B is absent from the mitotic spindle
To understand more precisely the role of kinesin-8B in building the male gametocyte flagellum, we investigated the subcellular location of kinesin-8B by live cell imaging in P. berghei. We generated a transgenic parasite line by single crossover recombination at the 3' end of the endogenous kinesin-8B locus to express a C- In order to establish whether kinesin-8B is part of the mitotic assembly during the three rounds of chromosome replication in male gamete development, we examined its location together with that of the kinetochore protein Ndc80. Parasite lines expressing kinesin-8B-GFP and Ndc80-cherry were crossed and used for live cell imaging of both markers to establish their spatio-temporal relationship. 1-2 min after gametocyte activation, kinesin-8B was observed close to the nucleus and adjacent to Ndc80, but not overlapping with it ( Figure 3A). This pattern of adjacent location is also seen in later stages of development. This shows that both axoneme formation and chromosome division begin at very early stage of gametogenesis and continue side by side ( Figure 3A). Completion of both processes is coordinated in a spatio-temporal manner before the onset of exflagellation, but kinesin-8B's role is specifically related to axoneme formation in the cytoplasm.

Kinesin-8B localizes to cytoplasmic MTs and its movement is blocked by antimalarial compounds
To further explore the location of kinesin-8B, we investigated its co-localization with MTs (using α-tubulin as a marker) by indirect immunofluorescence assay (IFA) using gametocytes fixed at different time points after activation. Kinesin-8B was localized on cytoplasmic MTs rather than on spindle MTs during male gamete formation. The pattern of both MTs and kinesin-8B was more distinct in later stages where their distribution adopted an axoneme-like pattern around the nucleus ( Figure 3B). Further we tested two molecules with known antimalarial properties (TCMDC-123880 and DDD01028076) with a putative role in targeting MT dynamics during male gamete development (Delves et al., 2018;Gamo et al., 2010) Addition of these molecules at 4 mpa blocked the dynamic distribution of kinesin-8B in over 80% of male gametocytes had no significant effect on Ndc80 dynamics ( Figure 3C). This suggest that these compounds are most effective against cytoplasmic MT, and serve to distinguish the cytoplasmic and nuclear MT-based processes.

Kinesin-8B associates dynamically with basal bodies and growing axonemes
Since the Δkinesin-8B parasite appeared to have defects in basal body interactions with nuclear pole and axoneme assembly, we analysed further the dynamic location of kinesin-8B by live imaging. We observed a diffuse cytoplasmic localization in nonactivated gametocytes, and following activation kinesin-8B accumulated at one side of the nucleus, while retaining a cytoplasmic location. Within 1 to 2 minutes after activation, the distribution of kinesin-8B showed four clear foci close to the nucleus, suggesting its association with the 'tetrad of basal bodies' which is the template for axoneme polymerization ( Figure 4A) (Sinden et al., 1978;Sinden et al., 1976). These four foci duplicated further within 2-4 min to form another set of four kinesin-8B foci, which later moved apart from each other ( Figure 4A). As gametogenesis proceeds, these kinesin-8B foci continued to move apart forming fibre-like structures around the nucleus by 4-6 mpa. The growth of these kinesin-8B-tagged fibres resembles the growing cytoplasmic axonemes (Marques et al., 2015;Straschil et al., 2010). The duplication of the four foci and emergence of these fibre-like structures occurred within 2-4 min ( Figure 4B, Supplementary videos SV1 and SV2). By 6-8 mpa, these fibrelike structures decorated by kinesin-8B were arranged in a specific pattern with the basal bodies around the nucleus and forming a basket-like structure, similar to that described in earlier ultrastructural studies (Sinden et al., 1978;Sinden et al., 1976).
This structure likely corresponds to completed axonemes. Kinesin-8B was also observed with a continuous distribution along the length of the gametes as they emerged from the residual body of the gametocyte. Overall, the spatial and temporal profile of kinesin-8B revealed a localisation to basal bodies and axonemes throughout gametogenesis.
Based on the ultrastructural studies, live cell imaging, and immunofluorescence microscopy, we suggest a model for the dynamic location of kinesin-8B on basal bodies and axonemes during male gamete development, which is depicted in  (Briggs et al., 2004;Sinden et al., 1976;Sinden et al., 2010). This assembly mechanism is rapid and presumably an adaptation to facilitate efficient transition through this life cycle stage.
We show that Plasmodium kinesin-8B has a specific role in building the male gamete flagellum, and in its absence parasite transmission from insect vector to mammalian host is blocked. The severe defect in exflagellation exhibited by Δkinesin-8B parasites is consistent with the location of protein we describe, and with its presence in the male proteome (Khan et al., 2005;Talman et al., 2014). Furthermore, a specific role for kinesin-8B at this life cycle stage is consistent with it being dispensable during blood stages (this study and (Bushell et al., 2017)).
Male gamete development in the malaria parasite involves three rounds of replication and mitotic division followed by chromosome condensation and cytokinesis, resulting in eight gametes Guttery et al., 2012a). DNA content analysis of the Δkinesin-8B mutant showed that the normal three rounds of replication occurred since octaploid nuclei (8N) were observed at 8-10 mpa. The distinct separation of kinesin-8B from the kinetochore marker Ndc80, and the normal nuclear ultrastructure, further substantiate the idea that this kinesin is not part of the mitotic spindle within the nucleus. The experiments using gametocyte-specific inhibitors (Delves et al., 2018;Gamo et al., 2010), provide further discrimination between MT-dependent events in the nucleuswhere these compounds have no effect on Ndc80 localisation and therefore presumably no effect on mitotic spindle MTsand the cytoplasm, where they have considerable effect on cytoplasmic axoneme assembly and kinesin-8B localisation.
Global transcript analysis and qRT-PCR examination of genes coding for proteins shown to affect cytokinesis and chromosome condensation, like CDPK4, CDC20, GEST and MAP2 (Billker et al., 2004;Guttery et al., 2012;Talman et al., 2011;Tewari et al., 2005) revealed no differentially expressed genes in the mutant, in defective Δkinesin-8B parasite male gametogenesis which were deemed important. This suggests that the observed phenotype is due to the absence of kinesin-8B alone.
The ultrastructure studies provide a very striking comparison between the assembled axonemes of wild type parasites and the lack of proper axoneme assembly in the absence of kinesin-8B. Examination of two different time points (8 and 15 mpa) revealed that single and doublet MTs were formed and elongated in the mutant parasites, but they were not organised into the classical 9+2 symmetry. Furthermore, although the amorphous nature of the basal body precluded identification of structural abnormalities in the absence of kinesin-8B, there appeared to be at least partial loss of the tight association between the basal body and the nuclear pole. This phenotype is very different to that of many other mutants with exflagellation defects (Billker et al., 2004;Ferguson DJP, 2013;Wall et al., 2018), but there is some similarity to that of other mutants affecting the basal body (SAS6) or involved in assembly of the central pair of MTs (PF16) in Plasmodium axonemes (Marques et al., 2015;Straschil et al., 2010). These data suggest that the kinesin-8B deletion causes a defect in both the basal body and axonemes.
Using live cell imaging, we show the dynamic localisation of kinesin-8B from the earliest stages of male gamete development (summarized in Fig 5). Kinesin-8B is a cytoplasmic protein exhibiting a basal body-like pattern in the early stages of gamete developmentthis highlights that basal body replication occurs within 2 minutes of activation and that associated axoneme-like structures are assembled later. The association of kinesin-8B with axonemes is supported by co-localisation with cytoplasmic MT. The pattern of axoneme development revealed by tracking kinesin-8B is consistent with earlier ultrastructural studies (Sinden et al 1976(Sinden et al , 1978. The formation of basal bodies from a single MTOC-like structure was evident at a very early time in gamete development (Sinden et al., 1978;Sinden et al., 1976). Here we have displayed the real-time dynamics of their duplication and localization as distinct structures. These features are consistent with the basal body defects resulting in disorganised axoneme assembly that we observe with the deletion of kinesin8B.
Kinesin motors participate in diverse functions in many processes including cell division, intracellular transport, intraflagellar transport and controlling MT dynamics ((Cross and McAinsh, 2014;Dawson et al., 2007;Scholey, 2008;Verhey and Hammond, 2009). The lack of association of kinesin-8B with mitotic assembly within the nucleus -and with kinetochores in particular -is distinct from what is seen for kinesin-8As in most eukaryotes (Savoian et al., 2004;Savoian and Glover, 2010;Wang et al., 2016). The metazoan kinesin-8B, Kif19, is involved in ciliary length regulation, but the role of kinesin-8B in axonemal assembly appears to be more profound in Plasmodium than controlling length alone. Kinesin-8B is consistently present in those Apicomplexa with flagellated gametes like Plasmodium, Toxoplasma and Eimeria and absent in other genera like Theileria, Babesia and Cryptosporidium which lack flagellated gametes (Zeeshan M, 2019). The properties of kinesin-8B contrast with those of Plasmodium kinesin-8X, which is located on the mitotic spindle (Zeeshan M, 2019).
In conclusion, this is the first study exploring the real-time dynamics and functional role of the kinesin-8B molecular motor in flagellum formation during Plasmodium male gamete development. It has a key role in establishing basal body formation and axoneme structure and assembly, thereby regulating male gamete development, which is an essential stage in parasite transmission.

Ethics statement
The animal work performed in the UK passed an ethical review process and was

Generation of transgenic parasites
The gene-deletion targeting vector for kinesin-8B (PBANKA_020270) 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 previously (Saini et al., 2017). PCR primers N1261 and N1262 were used to generate a 1000 bp fragment of kinesin-8B 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 1240 bp fragment generated with primers N1263 and N1264 from the 3′ flanking region of kinesin-8B was then inserted downstream of the dhfr/ts cassette using EcoRI and XbaI restriction sites. The linear targeting sequence was released using ApaI/XbaI. A schematic representation of the endogenous Pbkinesin-8B locus the constructs and the recombined kinesin-8B locus can be found in S1 Fig.
The C-terminus of kinesin-8B was tagged with GFP by single crossover homologous recombination in the parasite. To generate the kinesin-8B-GFP line, a region of the kinesin-8 gene downstream of the ATG start codon was amplified using primers T1991 and T1992, ligated to p277 vector, and transfected as described previously (Guttery et al, 2012 expressing GFP (for gene deletion) were transfected by electroporation (Janse et al., 2006).

Parasite genotype analyses
For the gene knockout parasites, diagnostic PCR was used with primer 1 (IntN126) and primer 2 (ol248) to confirm integration of the targeting construct, and primer 3 (N102 KO1) and primer 4 (N102 KO2) were used to confirm deletion of the kinesin-8B gene (S1 Fig). For the parasites expressing a C-terminal GFP-tagged kinesin-8B protein, diagnostic PCR was used with primer 1 (IntT199) and primer 2 (ol492) to confirm integration of the GFP targeting construct (S2 Fig).

Purification of gametocytes
The purification of gametocytes was achieved using a protocol described previously (Beetsma et al., 1998)  The gametocytes were harvested from the interface and washed.

Live cell-and -time lapse imaging
Purified gametocytes were examined for GFP expression and localization at different time points (0, 1-15 min) after activation in ookinete medium (Guttery et al 2014).
Images were captured using a 63x oil immersion objective on a Zeiss Axio Imager M2 microscope fitted with an AxioCam ICc1 digital camera (Carl Zeiss, Inc). Time-lapse videos (1 frame every 5 sec for 15-20 cycles) were taken with a 63X objective lens on the same microscope and analysed with the AxioVision 4.8.2 software.

Fixed immunofluorescence assay and measurements
The To measure nuclear DNA content of activated microgametocytes by direct immunofluorescence, images of parasites fixed (0 mpa and 8-10 mpa) and stained as above were analyzed using the ImageJ software (version 1.44) (National Institute of Health) as previously described (Tewari et al., 2005).

Generation of dual tagged parasite lines
The kinesin-8B-GFP parasites were mixed with Ndc80-cherry parasites in equal numbers and injected into a mouse. Mosquitoes were fed on this mouse 4 to 5 days after infection when gametocyte parasitemia was high. These mosquitoes were checked for oocyst development and sporozoite formation at day 14 and day 21 after feeding. Infected mosquitoes were then allowed to feed on naïve mice and after 4 -5 days these mice were examined for blood stage parasitemia by microscopy with Giemsa-stained blood smears. In this way, some parasites expressed both kinesin-8B-GFP and Ndc80-cherry in the resultant gametocytes, and these were purified and fluorescence microscopy images were collected as described above.

Inhibitor studies
Gametocytes were purified as above and treated with two antimalarial molecules (TCMDC-123880 and DDD01028076) (Delves et al., 2018;Gamo et al., 2010) at 4 mpa and then fixed with 4% PFA at 8 min after activation. Dimethyl sulfoxide (DMSO) was used as a control treatment. These fixed gametocytes were then examined on a Zeiss AxioImager M2 microscope fitted with an AxioCam ICc1 digital camera (Carl Zeiss, Inc).

Parasite phenotype analyses
Blood containing approximately 50,000 parasites of the Δkinesin-8B line was injected intraperitoneally (i.p) into mice to initiate infections. Asexual stages and gametocyte production were monitored by microscopy on Giemsa stained thin smears. Four to five days post infection, exflagellation and ookinete conversion were examined as described previously (Guttery et al, 2012) with a Zeiss AxioImager M2 microscope (Carl Zeiss, Inc) fitted with an AxioCam ICc1 digital camera. To analyse mosquito transmission, 30-50 Anopheles stephensi SD 500 mosquitoes were allowed to feed for 20 min on anaesthetized, infected mice with an asexual parasitemia of 15% and a 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 14 post feeding, and oocysts were counted on an AxioCam ICc1 digital camera fitted to a Zeiss AxioImager M2 microscope using a 63x oil immersion objective. On day 21 post-feeding, another 20 mosquitoes were dissected, and their guts crushed in a loosely fitting homogenizer to release sporozoites, which were then quantified using a haemocytometer or used for imaging. Mosquito bite back experiments were performed 21 days post-feeding using naive mice, and blood smears were examined after 3-4 days.

Electron microscopy
Gametocytes activated for 8 min and 15 min were fixed in 4% glutaraldehyde in 0.1 M phosphate buffer and processed for electron microscopy as previously described (Ferguson et al., 2005). Briefly, 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).

Quantitative Real Time PCR (qRT-PCR) analyses
RNA was isolated from 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 a SYBR green fast master mix kit (Applied Biosystems). All the primers were designed using the primer3 software (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 three 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 S2 Table. RNAseq analysis Libraries were prepared from lyophilized total RNA using the KAPA Library

Statistical analysis
All statistical analyses were performed using GraphPad Prism 7 (GraphPad Software). For qRT-PCR, an unpaired t-test was used to examine significant differences between wild-type and mutant strains.

Data Availability. Sequence reads have been deposited in the NCBI Sequence
Read Archive with accession number: PRJNA549466

Author Contributions
Conceived    showing the resulting phenotype of compound addition at 4 mpa. In contrast to the effect on kinesin-8B-GFP distribution no significant effect was seen on Ndc80-RFP.
Inhibitors were added at 4 mpa and parasites were fixed at 8 mpa. Scale bar = 5 µm. One to two min post activation (mpa) it accumulates at one end of the nucleus and forms four foci reminiscent of the 'tetrad of basal body' defined by electron microscopy (Sinden et al, 1976). These tetrad foci are duplicated within 2 to 4 mpa, and by 4 to 6 mpa fibre-like structures, representing axonemal growth decorated with kinesin-8B-GFP, extend from these tetrads to make a basket-like structure around the nucleus, which is completed by 8 to 10 mpa. Following exflagellation kinesin-8B-GFP is located along the length of the flagellum in the free male gamete (15 mpa, inset). For each time point a cartoon guide is presented. B. Still images (at every 5 seconds) of tetradfoci duplication and the start of axonemal growth decorated with kinesin-8B-GFP at 2 to 3 mpa (Fig 3B.1; Supplementary videos SV1) and 3 to 4 mpa (Fig 3B.2; SV2). Scale bar = 5 µm. Supplementary Tables S1 Table. List of differentially expressed genes between ∆kinesin-8B and WT activated gametocytes S2