Plasmodium berghei Gamete Egress Protein is required for fertility of both genders

Abstract Male and female Plasmodium gametocytes ingested by the Anopheles mosquitoes during a blood meal egress from the red blood cells by rupturing the two surrounding membranes, the parasitophorous vacuole and the red blood cell membranes. Proteins of the so‐called osmiophilic bodies, (OBs), secretory organelles resident in the cytoplasm, are important players in this process. Once gametes emerge, the female is ready to be fertilized while the male develops into motile flagellar gametes. Here, we describe the function(s) of PBANKA_1115200, which we named Gamete Egress Protein (GEP), a protein specific to malaria parasites. GEP is restricted to gametocytes, expressed in gametocytes of both genders and partly localizes to the OBs. A mutant lacking the protein shows aberrant rupture of the two surrounding membranes, while OBs discharge is delayed but not aborted. Moreover, we identified a second function of GEP during exflagellation since the axonemes of the male flagellar gametes were not motile. Genetic crossing experiments reveal that both genders are unable to establish infections in mosquitoes and thus the lack of GEP leads to a complete block in Plasmodium transmission from mice to mosquitoes. The combination of our results reveals essential and pleiotropic functions of GEP in Plasmodium gametogenesis.

cGMP-dependent signaling pathway resulting in Ca 2+ mobilization from internal stores and activation of the calcium-dependent protein kinase 4 (CDPK4). This master regulator is essential for male gamete maturation, which entails three mitotic divisions, and assembly of axonemes to form eight motile gametes (Billker et al., 2004). Blood stage gametocytes develop within a parasite-specific compartment, the so-called parasitophorous vacuole (PV). The egress of activated male and female gametocytes from the host RBC involves the sequential rupture of the PV membrane (PVM) and the RBC membrane (RBCM) (Deligianni et al., 2013;Sologub et al., 2011). Nearly concomitant with the RBCM rupture, the male gamete flagella start beating detaching from the residual body of the cytoplasm (Andreadaki et al., 2018). Female gametogenesis is regulated by translational repression of messenger RNAs; mRNA turnover influences gene expression in a stage-specific manner. This was shown for the DDX6-class RNA helicase, DOZI (development of zygote inhibited), usually found in a complex with mRNA species in cytoplasmic bodies of females . Gene deletion of DOZI led to inhibition of the ribonucleoprotein complexes formation with subsequent degradation of at least 370 transcripts. However, female gametogenesis morphologically comprises, as far as is known, only the egress from the host RBC.
Proteins localized in specialized secretory organelles called osmiophilic bodies (OBs) have been identified as having specific roles in the egress process. In the rodent Plasmodium berghei, mutants lacking Plasmodium male development gene -1 (MDV1) (Ponzi et al., 2009) and gamete egress and sporozoite traversal protein (GEST) (Talman et al., 2011), detected in OBs of both genders, are unable to rupture the PVM, while lack of G377, a female-specific OB factor, cause a slight delay in the egress of female gametes (Olivieri et al., 2015). MTRAP, a member of the thrombospondin-related anonymous protein (TRAP) family, is also required for PVM rupture (Bargieri et al., 2016). The putative pantothenate transporter PAT, a membrane component of the OBs, has a function in the release of OB contents, and gametes lacking pat remain trapped inside the PVM . A P. berghei Ferlin-like protein (FLP) localizes to vesicles distinct from OBs and is also involved in gamete egress (Obrova, Cyrklaff, Frank, Mair, & Mueller, 2019). Gametes lacking FLP remain trapped inside PVM and RBC membranes. A male-specific perforin-like protein, PPLP2, is specifically required for RBCM rupture (Deligianni et al., 2013). Other studies have shown that male gametogenesis is dependent upon the APC/C (Anaphase Promoting Complex/Cyclosome) protein for chromosome condensation and cytokinesis (Wall et al., 2018); in its absence, cytokinesis is abnormal and no gametes are formed, resulting in a similar phenotype to mutants lacking the CDC20 homolog (Guttery et al., 2012). The aspartic protease plasmepsin X (PMX) is also involved in P. berghei gametocyte egress. Inhibition of PMX activity through treatment with the aspartic protease inhibitor 49c causes a 10-fold decrease exflagellation rate and prevents rupture of the RBCM in both female and male gametocytes (Pino et al., 2017).
Flagellar proteins are also required for proper axoneme formation.
PF16 functions in maintaining microtubule structure, and in its absence, flagellar movement is abnormal and fertility is reduced . In parasites lacking the SAS-6 protein, a component of the basal body and centriole, flagellar motility, and nuclear allocation are affected leading to reduced fertility (Marques et al., 2015). A mutant lacking Pbmap-2 (Tewari, Dorin, Moon, Doerig, & Billker, 2005) fails to release flagellated gametes, while the serine/arginine-rich (SR) protein kinase (SRPK) is essential for male gamete formation . Furthermore, the stage-specific actin II is required both for axoneme activation and egress of male gametocytes (Deligianni et al., 2011). A role in axoneme assembly and flagellum formation during male gamete development is played by kinesin-8B (Zeeshan et al., 2019).
In this work, we characterized a P. berghei (PBANKA_1115200), here named Gamete Egress Protein (GEP) due to its essential role(s) in gametogenesis of both male and female gametocytes. GEP is expressed in sexual blood stages and partly localizes to the OBs. When GEP is lacking, gametocytes develop normally but their maturation into gametes is severely affected. In males, axonemes are formed though they do not beat. Egress from the host RBC is also delayed in both genders and transmission to the mosquito is completely abrogated. Overall, our results suggest an important and pleiotropic function of GEP during Plasmodium transmission to the vector.

| Mosquito and parasite strains
Anopheles gambiae mosquitoes were reared as described in Facchinelli et al., 2015. Plasmodium berghei ANKA 8417HP strain was used throughout the study (Janse et al., 1989) and was maintained in CD1 mice. Transfections were performed as described in . Mosquito infections were carried out by offering Anopheles gambiae strain G3 to mice infected with WT and mutant parasites. Mice were matched to have a parasitemia of roughly 5%-10% and a similar exflagellation rate in each experiment. Mosquito midguts were dissected after 13 days. Synchronous infections were established by intravenous injection of purified schizonts (Janse & Waters, 1995). Blood was collected by heart puncture under anesthesia, and leukocytes were removed using Plasmodipur leukocyte filters (Euro Diagnostica). Schizont or gametocyte infected erythrocytes were purified through Nycodenz density gradient centrifugation (Janse & Waters, 1995).

| The plasmid used for transfection and generation of knockout lines
The pBAT plasmid (Kooji, Rauch, & Matuschewski, 2012) was used to prepare the final transfection construct for the generation of gep (-) parasites.
The transgenic parasite line was generated by introducing a 728 bp PCR fragment amplified from the 5'UTR of PBANKA_1115200 (primers L-f and L-r) in the restriction sites SacII-EcoRI upstream to the mCherry coding region. A PCR fragment of 692 bp (primers R-f and R-r, Table A1) amplified in the 3'UTR of PBANKA_1115200 was inserted in restriction sites XhoI-KpnI of the pBAT (Table A1). These two regions provided the targets used for double crossover recombination in the genomic locus. An hDHFR-yFcu cassette, which encodes antifolate resistance, allowed for pyrimethamine selection of transfected parasites. Transfection and selection of transformed parasites were performed using standard genetic modification technologies for P. berghei (Janse, Franke-Fayard, Mair, et al., 2006) using P. berghei HP as the parent parasite line.
After transfection of P. berghei 847HP as parental line, 5 × 10 5 GFP-fluorescent red blood cells were sorted and intravenously injected into two CD1 mice and kept under pyrimethamine selection.
When parasitemia was established, sorting of fluorescent parasite and reinfection under selection was repeated two more times.

| Production of GEP serum and antibodies dilutions
Bacterially expressed recombinant protein was produced using the pGEX-6P-1 vector into which a 453 bp PCR fragment, corresponding to a region within the 6th exon of PBANKA_1115200, coding for amino acids 284-399, was inserted. This fragment was amplified using the primers: REL-10-for and REL-10-rev (Table A1). This recombinant protein was used to produce a specific immune serum in BALB/c mice. About 50 µg of the recombinant protein and complete Freund's adjuvant were mixed to form a stable emulsion and injected intraperitoneally in the stomach area. The immunization procedure was repeated using 25 mg of antigen in incomplete Freund's adjuvant 28, 42, 56, and 70 days after the first injection. At day 84, the immunized mice were bled to obtain immune serum. Before the immunization cycle, blood (100 µl) was collected from the submandibular vein of each mouse to obtain preimmune serum.

| Immunofluorescence assay
The TER-119 antibody was used to label the outer leaflet of the RBCM, and anti-SEP1 antibody was used as a marker for PVM. Gametocytes were activated, and samples were fixed in 4% formaldehyde at different time points after the activation. All steps were carried out at room temperature. For immunolabeling of WT and gep(-) parasites with anti-SEP1 antibody activated for 15 min, samples were permeabilized with 0.1% saponin for 30 s. The TER-119 antibody was added diluted 1:200 in PBS with 2% normal goat serum for 1 hr incubation.
Hoechst 33342 was used to stain DNA. Samples are washed twice with 1 × PBS before mounting in Vectashield (Vector laboratories).

| GEP is conserved within the Plasmodium genus
Gamete Egress Protein is a protein of 1,138 amino acids encoded by a 4,426 bp gene on chromosome 11 having 10 exons residing in a syntenic region (Figure 1a). The protein is well conserved within the Plasmodium genus as revealed by multiple sequence alignment comparing five different Plasmodium species ( Figure A1). BLAST analysis of the protein did not retrieve any similar protein outside the genus or known protein domains or motifs. GEP has previously been detected in the proteome of both male and female gametocytes (Khan et al., 2005) and was shown to be phosphorylated in gametocytes High throughput screening gene deletion analysis revealed that the protein was dispensable for growth of the asexual blood stages (Bushell et al., 2017).

| A gep knockout mutant is strongly impaired in gametocyte egress from the red blood cell
To ( Figure A2). Correct integration of the construct was confirmed by diagnostic PCR ( Figure A2).
Growth of asexual blood stages was not impaired in either cloned lines lacking GEP compared with the WT (Figure 1b). This is consistent with previously reported results (Bushell et al., 2017). In the two independent gep(-) clones, both gametocytemia and morphology of male and female gametocytes were also roughly similar to WT as scored in Giemsa stained blood smears from infected mice (Table A2).
F I G U R E 1 (a) Scheme of the gep locus (PBANKA_1115200) showing the coding region of 1,138 amino acids in 10 exons (red boxes). Violet and pink indicated, respectively, the 5' upstream and the 3' downstream regulatory regions. F I G U R E 2 (a) Western blot analysis on purified HPE parasites, WT gametocytes (2 independent preparations), and gep(-) gametocytes (clones 1 and 2). GEP serum recognizes a specific band at 133 kDa in the WT gametocyte samples, while no signal was detected both in asexual parasites (HPE) and in gametocytes of the gep(-) cloned lines. Samples were normalized using anti-Pb14-3-3 detected in both asexual and sexual stages (Lalle et al., 2011). (b) Immunofluorescence assay using GEP-specific serum. A signal is detected in WT gametocytes but not in mutant parasites. Scale bar 5 μm. (c) Double IFA on WT gametocytes using immune sera against GEP serum and the nuclear protein SET, highly abundant in male gametocytes. Scale bar 5 μm. with either a ruptured PVM and/or RBCM (Figure 1e). This revealed striking differences between the two parasite lines.  (Figure 1d,e).

| GEP is expressed in male and female gametocytes and partly localizes to OBs
To determine the expression pattern and subcellular localization of GEP protein, a specific serum was produced against a recombinant peptide expressed in E. coli.
In a Western blot assay (Figure 2a), a band at the predicted size of 133 kDa was detected in WT gametocytes samples from two independent preparations but no signal was detected in the two gep (-) clones neither in blood samples of the P. berghei line HPE which does not produce gametocytes (Janse, Ramesar, Berg, & Mons, 1992), indicating that GEP is expressed specifically in sexual stages. 14-3-3 was used as a control to normalize sample amount (Lalle et al., 2011) . In IFA, the specific serum detected GEP in a punctated pattern in WT gametocytes while no signal was seen in gep(-) clones ( Figure 2b). GEP serum labels both female and male gametocytes ( Figure 2c) identified by costaining with anti-SET serum, highly abundant in the nuclei of male gametocytes (Pace et al., 2006). Double IFA using anti-GEP and anti-G377 immune sera showed that GEP localizes partly to OBs (Figure 2d).  Figure 5a). In the cytoplasm of activated males, axonemes were visible with the typical ring of nine outer microtubule doublets ( Figure 5a, right panel). However, we did not observe a concomitant nuclear division but an enlarged multilobated nucleus and only in a few cases microtubule organization centers (Figure 5b).

The same preparations of activated WT and gep(-)gametocytes
were also examined in parallel by IFA using anti-tubulin antibodies that stain male gamete flagella (Figure 5c). In the WT, gamete formation was almost completed 10 min postactivation. In gep(-) parasites, male gametocytes were still enclosed in the membranes 20 min postactivation. Distinct nuclei were not visible while axonemes were fully formed consistent with the EM data. Overall, these data suggest that in gep(-) mutant, axoneme formation and nuclear divisions are not coupled.

| Genetic crosses reveal a functional defect in both genders of gep(-)
Our line were unable to produce ookinetes in three independent experiments, while a cross of Δ47 and Δ48/45 done as a control produced fully mature ookinetes.
To check the in vivo phenotype of gep(-) parasites, An. gambiae mosquitoes were fed on mice infected with WT or gep(-) cloned lines.
As expected, no oocysts were detected in 2 independent experiments using the two gep(-) clones, while in both experiments, WT parasites were infective to mosquitoes (Figure 6b).

| D ISCUSS I ON
In this study, we identified the protein GEP (PBANKA_1115200) as a novel factor required for the successful gametogenesis of malaria parasites. GEP is a well-conserved protein within the Plasmodium genus, but without any similarity to known proteins of other organisms pointing to a highly specific function in malaria parasites.
The protein was observed in the cytoplasm of gametocytes partly colocalized with an osmiophilic body (OB) protein. Our functional analysis revealed that mutants lacking gep had a pleiotropic phenotype with defects in gametogenesis resulting in the inability of both genders to establish mosquito infections. Both male and female gametocytes had a severe delay in egress from the red blood cell. At 20 min postactivation, the PVM was still intact in about half of the cells. A more dramatic effect was seen in RBCM rupture with only a few mutant parasites being released from the host cell. Furthermore, during male gametogenesis, no motile flagella were formed, although axonemes were assembled. The ultrastructural analysis confirmed this finding and also revealed that nuclear division did not take place.
Genetic crosses confirmed that GEP was essential for the fertility of both genders.
Gametogenesis is a multi-step process, with many aspects unique to the malaria parasite. In both genders, the coordinate action of several Plasmodium-specific proteins is required for egress of mature gametes from the host RBC. In P. berghei, OB-resident proteins MDV1 (Ponzi et al., 2009) and GEST (Talman et al., 2011) are necessary for membrane rupture and in mutants lacking either of these proteins mature gametes of both genders remain trapped inside the host cell. OBs move to the cell membrane F I G U R E 6 (a) Ookinete conversion from genetic crosses of gep (-) parasites and parasite lines producing sterile male (Δ45/48) and female (Δ47) gametocytes. As a control, Δ45/48 was also crossed to Δ47. Dotted line: average WT ookinete conversion. b) Oocyst formation of WT and gep(-). In mosquitoes infected with gep(-) parasites, no oocysts were detected in dissected mosquito midguts 13 days after blood feeding after activation, and cargo proteins are secreted in the PV lumen 5-8 min upon gametocyte activation. While gep(-) parasites are also severely affected in egress from the host cell, in contrast to MDV and GEST, the protein is mostly retained inside the cytoplasm of the gametocytes. However, we cannot exclude a minor portion of the protein being secreted due to the lack of a suitable marker to investigate this aspect. We observed that OBs also in our mutants moved to the cell membrane, but the release of cargo was delayed though not abolished. This suggests that GEP has a role upstream of OB discharge. Interestingly, a function in membrane fusion and secretion has been shown for the OB-resident small solute transporter PAT (Kehrer, Singer, et al., 2016). Similar to GEP, the phenotype of mutant parasites lacking PAT is a severe reduction in the release of gametes of both genders.
Male gametogenesis is a complex process with three mitotic divisions taking place concomitantly with axoneme assembly.
After cytokinesis, axonemes are activated and eight flagellar gametes, each with a haploid nucleus, exit the cell. In the gep(-), mutant axonemes were formed but not activated and cytokinesis did not take place. Similar phenotypes have been observed previously. Functional studies on the Anaphase Promoting Complex/ Cyclosome (APC/C) revealed a role in cell cycle control during male gametogenesis; a conditional mutant was severely affected in chromosome condensation and cytokinesis (Wall et al., 2018).
A similar phenotype was also observed in P. berghei mutants lacking the CDC20 homolog (Guttery et al., 2012) and the MAP2 kinase (Tewari et al., 2005). The possibility that GEP together with these three proteins is part of a common pathway is an interesting hypothesis.
GEP has also been identified as interacting with P granule proteins ALBA4 (Muñoz et al., 2017), DOZI , and the Sm-like factor CITH (homolog of worm CAR-I and fly Trailer Hitch, Mair et al., 2010) the latter two which are required in females for proper development of the zygote (Mair et al., 2010).
P granules store and protect mRNA and function in post-transcriptional repression of transcripts previously produced in the parasite. Noteworthy is that the phenotype of the mutant lacking GEP is different from mutants lacking DOZI, CITH, and ALBA4.
A mutant lacking ALBA4 showed an increase in male exflagellations but had no apparent function in female gametocytes (Muñoz et al., 2017) while DOZI and CITH mutants were blocked in zygote development, through misregulation of maternal transcripts.
A possible function of GEP in transcript regulation remains to be investigated.
How membrane rupture, flagellar beating, and cytokinesis are coordinated during male gametogenesis is still largely unknown. GEP may represent a valuable tool to further investigate this aspect in the field of gamete egress.

CO N FLI C T O F I NTE R E S T
None declared.

E TH I C S S TATEM ENT
All work was carried out in full conformity with Greek regulations and laws on animal experiments. In Greece, these issues are