Differentiation of Plasmodium male gametocytes is initiated by the recruitment of a chromatin remodeler to a male-specific cis-element

Izumi Kaneko; Tsubasa Nishi; Shiroh Iwanaga; Masao Yuda

doi:10.1073/pnas.2303432120

. 2023 May 8;120(20):e2303432120. doi: 10.1073/pnas.2303432120

Differentiation of Plasmodium male gametocytes is initiated by the recruitment of a chromatin remodeler to a male-specific cis-element

Izumi Kaneko ^a, Tsubasa Nishi ^a, Shiroh Iwanaga ^b, Masao Yuda ^a,¹

PMCID: PMC10193995 PMID: 37155862

Significance

Plasmodium gametocytes, precursors of gametes, play a central role in parasite transmission to the mosquito vector. Male gametocytes transform into multiple motile male gametes in the blood meal of a mosquito and promptly fuse with female gametes. The transcriptome of male gametocytes consists of genes for genome replication and flagella-based motility, reflecting their roles in generating multiple male gametes. Here, we report that a gametocyte-specific chromatin remodeler plays a critical role in establishing this unique gene expression repertoire by creating nucleosome-free regions upstream of these genes. We report that chromatin remodeling is an initial step in stage conversion in the Plasmodium lifecycle.

Keywords: Plasmodium, chromatin remodeler, SNF2 family ATPase, male gametocyte

Abstract

Plasmodium parasites, the causative agents of malaria, possess a complex lifecycle; however, the mechanisms of gene regulation involved in the cell-type changes remain unknown. Here, we report that gametocyte sucrose nonfermentable 2 (gSNF2), an SNF2-like chromatin remodeling ATPase, plays an essential role in the differentiation of male gametocytes. Upon disruption of gSNF2, male gametocytes lost the capacity to develop into gametes. ChIP-seq analyses revealed that gSNF2 is widely recruited upstream of male-specific genes through a five-base, male-specific cis-acting element. In gSNF2-disrupted parasites, expression of over a hundred target genes was significantly decreased. ATAC-seq analysis demonstrated that decreased expression of these genes correlated with a decrease of the nucleosome-free region upstream of these genes. These results suggest that global changes induced in the chromatin landscape by gSNF2 are the initial step in male differentiation from early gametocytes. This study provides the possibility that chromatin remodeling is responsible for cell-type changes in the Plasmodium lifecycle.

Plasmodium parasites have a complex life cycle; they alternate between mosquito and vertebrate hosts and pass through multiple life stages in these hosts. Regulation of gene expression is critical for invasion of specific target cells or organs and adapting to the new environment. Elucidation of mechanisms of stage-specific gene regulation is important to understand the molecular basis of parasitism and will help us to develop strategies to prevent infections.

Gametocytes, precursors of gametes, develop from asexual erythrocytic stages in the circulation and mediate parasite transmission to the mosquito (1). They are a target of malaria transmission-blocking strategies that aim to prevent the prevalence of the disease among humans. Gametocytes generation, or gametocytogenesis, involves two steps: differentiation from replicative asexual stage into nonreplicative early gametocytes and differentiation from early gametocytes into female and male gametocytes. Multiple AP2 family transcription factors (TFs) are involved in these steps. In the initial step of the gametocytogenesis, AP2-G plays essential roles (2, 3). AP2-G is responsible for the onset of the gametocytogenesis as a master TF; upon disruption of this, TF parasites lose the ability to produce gametocytes, and by conditional overexpression of AP2-G gametocytes can be induced in asexual erythrocytic stage parasites (4). Our recent study on Plasmodium berghei demonstrated that this TF induces multiple TFs essential for gametocyte development and generates a driving force for gametocytogenesis (5). AP2-G2 is one of the target genes of AP2-G and supports transition of asexual stage parasites to early gametocytes as a repressor (6, 7). In the second step of differentiation of early gametocytes into female gametocytes, at least two targets, AP2-FG and AP2-O3, play a critical role (7, 8). AP2-FG plays essential roles in female maturation. It binds the ten-base motif TGYAYRTRCA, on the promoter of target genes. AP2-O3 plays a role in development as a transcriptional repressor, repressing several male-specific genes, and supports female differentiation.

Differentiation of males, which starts at the same time as that of females, also contains two steps. The first step is common between the two sexes because disruption of AP2-G2 results in developmental arrest of both sexes (6). Male-specific development begins in the second step and male-specific repertoire of gene expression is established during this period. Sex-specific proteome and transcriptome analyses revealed that repertoire of genes expressed in males contains genes for genome replication and flagella-based motility, which are required for the generation of multiple motile male gametes (9–11). However, in contrast to females, transcriptional regulation involved in the second step of male-specific differentiation, remains poorly understood. It remains unclear whether a master TF like AP2-FG is also present in males and which sequences in the promoter act as male-specific cis-acting elements.

In eukaryotes, gene regulation by a sequence-specific TF involves several steps (12). The series of events require a nucleosome-free region (NFR) which allows TFs, coregulators, and basic TFs to access genomic DNA. Thus, the regulatory region sometimes has to be altered to NFR beforehand by removing or repositioning nucleosomes (13). Mechanisms involving ATP-dependent chromatin remodeling complexes (remodelers) play important roles in this process; they slide or evict nucleosomes in ATP-dependent manner and make the regulatory DNA region accessible. Genome-wide change of chromatin structures by a remodeler can govern gene expression repertoire and contribute to establishment of a cell-type-specific gene expression program.

ATPase subunit is a core subunit of the remodeler, and it constitutes the complex with multiple subunits associated with it. Chromatin remodelers are classified into four major families based on the domain structure of the ATPase subunit: SWI/SNF, SWI, INO80, and CHD families (14). These families are involved in various cellular processes dependent on chromatin remodeling, such as transcription and genome replication and repair. SWI/SNF plays an important role in transcriptional regulation. In the budding yeast Saccharomyces cerevisiae, one of the member RSC (essential remodeling complex) is recruited to the promoter of the target gene, slides nucleosomes, and creates NFR limited by stably positioned +1 and −1 nucleosomes, initiating transcription (15).

We analyzed the target genes of AP2-G and found a gene encoding an ATPase subunit of the SWI/SNF family (SNF2 ATPase) in the targets (Gene ID: PBANKA031270). Based on its specific expression in gametocytes, we designated this gene as gSNF2 and investigated its role in the gametocyte, expecting its possible involvement in gametocytogenesis.

Results

Targets of AP2-G Contain an ATPase Subunit of the SWI/SNF Family.

The gSNF2 gene, PBANKA031270, is one of the targets of AP2-G, harboring a ChIP-seq peak of AP2-G approximately 0.8 kbp upstream (Fig. 1A) (5). This gene is included in 30 genes that are up-regulated by the conditional expression of AP2-G in P. berghei in the earliest period (within 6 h after the induction of AP2-G), which is consistent with our ChIP-seq result showing that the gSNF2 gene is a direct target of AP2-G (4). Currently, this gene is classified as a DEAD-box RNA helicase, because it possess DEXDc and HELICc domains (16). However, phylogenetic analyses using its helicase-related region found these to be more similar to ATPase subunit of remodelers, particularly to the ATPase subunit of the SWI/SNF family (SNF2 ATPase) (Fig. 1B). Therefore, we investigated additional domains that indicated that it is an ATPase subunit of the remodeler complex. Analysis with SMART (http://smart.embl.de/smart/set_mode.cgi) and Pfam domain (http://pfam.xfam.org) databases using amino acid regions conserved in Plasmodium orthologues, identified two putative domains, a HSA (helicase/SANT-associated) and SnAC (SNF2 ATPase-coupling) (Fig. 1C and SI Appendix, Tables S1 and S2). HSA domains are involved in binding to actin-related proteins (17). SnAC domains are specific to the ATPase subunit of the SWI/SNF2 family chromatin remodeler (18) and participates in attaching to the nucleosome (17). Alignment of its amino acid sequence with SnAC domains of human and model organisms are shown in Fig. 1D. On the other hand, bromodomains, which have the capacity to bind acetylated lysine residues of proteins and are located typically in the C-terminal portion of the SNF2 ATPase of these model organisms including S. cerevisiae (Fig. 1C), were not identified. These results strongly suggested that this gene is an ATPase subunit of the SWI/SNF family that possess a unique domain structure. We further investigated whether this protein is unique to the Plasmodium species or is observed in other organisms. BLASTP search (https://blast.ncbi.nlm.nih.gov/Blast.cgi) identified several proteins with high sequence similarity to the helicase-related region of this protein in the apicomplexan parasites and their photosynthetic relative, Vitrella brassicaformis (Fig. 1C). All of them possessed the putative HAS and SnAC domains (Fig. 1C) but not the bromodomains. Many of these genes have been annotated as SNF2-related genes due to sequence similarities of their helicase-related regions to SNF2 proteins, indicating that these SNF2 ATPases are widely present in apicomplexan parasites.

To study the expression profile of this gene, we developed parasites that express it as a protein fused with mNeonGreen, the brightest green-yellow monomeric fluorescent protein reported thus far (gSNF2::mNeon parasites, SI Appendix, Figs. S1A and S1B) (19). In these parasites, clear signals were observed in the nucleus of both male and female gametocytes but not in the asexual stages (Fig. 1E). To investigate the expression profile of this gene, we prepared rats synchronously infected with gSNF2::mNeon parasites by intravenously injecting them with cultured mature schizonts. Time-course study using these rats showed that expression of the gene starts at 20 hpi (hours post erythrocyte infection), a few hours before manifestation of sexual dimorphism in the early gametocytes, and continued until they developed into mature gametocytes (Fig. 1F). The expression started a few hours after the expression of AP2-G, consistent with being a target of AP2-G.

gSNF2 Is Essential for the Development of Male Gametocytes.

Two gSNF2-disrupted [gSNF2(−)] parasite clones were developed independently (SI Appendix, Fig. S1C). Both clones showed normal proliferation in mice (Fig. 2A) and produced male and female gametocytes that appeared mature in Giemsa-stained blood smear, that is, completely filling the host erythrocyte and displaying sex-specific features including shape of the nucleus and staining of the cytoplasm (Fig. 2B). However, they completely lacked the capacity of exflagellation, suggesting impaired male development. In mosquito transmission experiments, no oocysts were formed on the midgut of mosquitoes that fed on mice infected with these parasites (Fig. 2C). To examine whether the development of both female and male gametocytes was impaired, parasites were subjected to cross-fertilization experiments (Fig. 2D). Ookinetes were produced by fertilization between their females and normal males, but no ookinetes were generated by fertilization between their male and normal females, demonstrating that only male gametocytes were functionally impaired by the disruption.

Fig. 2. — Phenotype of *gSFNF*2-disrupted parasites. (A) Proliferation of asexual erythrocytic stage-parasites were compared between *gSNF2*-disrupted [*gSNF2*(−)-1 and *gSNF2*(−)-2] and wild-type parasites. Three mice were infected with parasites, and averages of parasitemia were determined in Giemsa-stained blood smear. (B) Giemsa-stained views of a mature female (*Left*) and male (*Right*) gametocyte of gSNF2-disrupted parasites. Observation of a Giemsa-stained blood smear identified wild-type mature female gametocytes as mononuclear cells with a bluish purple cytoplasm and a compact nucleus and mature male gametocytes as mononuclear cells with a faintly stained cytoplasm and a diffused nucleus. (C) Mice infected with *gSNF2*-disrupted or wild-type parasites were subjected to mosquito bites and oocysts formed in the mosquito midguts were counted after 14 d. Values are average of 20 mosquitoes. (D) Cross-fertilization experiments were performed between *gSNF2*-disrupted parasites and parasites without fertile males [*p48/45*(−)] or parasites without fertile females [*p47*(−)]. Results are shown as conversion rates of female gametocytes to ookinetes. (E) *gSNF2*-disrupted parasites were prepared from transgenic 820cl1m1cl1 parasites that express RFP in female and GFP in male gametocytes. Flow cytometric analyses were performed in wild-type, 820cl1m1cl1, and *gSNF2*-disrupted parasites. GFP- and RFP-positive cells are shown in green and red, respectively. (F) Volcano plot comparing gene expression between the gametocytes of wild-type and *gSNF2*-disrupted parasites. Three biologically independent RNA-seq analyses were performed using gametocyte-enriched blood. The data were analyzed using DEseq2 software. Points corresponding to male-and female-enriched genes are colored in blue and red, respectively. padj, adjusted P value.

gSNF2-disrupted parasites were also prepared using transgenic parasites that express green and red fluorescent protein genes in females and males (P. berghei ANKA 820cl1m1cl1, here called strain 820) (20) (SI Appendix, Fig. S1B). FACS analysis of these parasites showed that green fluorescence in these parasites was significantly weaker than that in the original parasites (approximately fivefold) (Fig. 2E), suggesting that males were produced but their gene expression was affected by the disruption. These results suggested the possibility that chromatin remodeling plays an important role in gene regulation of male gametocytes. The phenotype described above was different from that reported by Kent et al. (4), which reported that parasites with this gene disruption lacked the capacity to produce male gametocytes.

Expression of Male-Specific Genes Decreased in gSNF2-Disrupted Parasites.

To investigate the effects of the disruption of gSNF2 on gene expression in detail, RNA-seq analysis was performed. Mice were pretreated with phenyl-hydrazine to increase reticulocytes. They were infected with gSNF2(−)-1 parasites and treated with sulfadiazine to kill asexual stage parasites, and gametocyte-enriched whole blood was harvested. The rates of female to male gametocytes were similar between the wild-type and gSNF2(−) parasites (SI Appendix, Table S3). In the mutants, expression of 167 genes was significantly decreased compared to the wild-type parasites [log2 (Fold Change) < −1, padj (adjusted P value) < 0.01] (Fig. 2F and Dataset S1). According to sex-specific RNA-seq data (Dataset S2) (11), majority of the significantly decreased genes (78%, 131 genes) were classified into male-enriched genes (fourfold enrichment compared to both the females and asexual stages), and only one gene was classified into female-enriched genes (four-fold enriched compared to both males and asexual stages). On the other hand, no relations to sex were observed in 53 genes whose expression was increased in the gSNF2(−) parasites. These results suggested that gene expression in males was significantly affected by the disruption, which was consistent with the observation that abnormal phenotype was observed only in males.

gSNF2 Is Recruited Upstream of Female- and Male-Specific Genes.

ChIP-seq analysis was performed to investigate the genomic regions of gSNF2 recruitment. For ChIP with anti-GFP antibody, parasites that expressed GFP-fused gSNF2 were prepared (gSNF2::GFP parasites, SI Appendix, Fig. S1C). Tagged protein was observed in the nucleus of both male and female gametocytes as in gSNF2::mNeon parasites using fluorescent microscopy. Gametocyte-enriched blood was prepared for these parasites and subjected to ChIP-seq analysis. Two biologically independent experiments (experiment 1 and experiment 2) were performed, and a total of 1,536 and 1,562 peaks were identified, respectively (Datasets S3 and S4). Approximately 90% of the peaks (1,303 peaks) were shared between these two experiments (Fig. 3 A and B), and these common peaks were used for the following analyses (Dataset S5). Majority of these gSNF2 peaks were located in the intergenic regions of the genome (Fig. 3C), coincided with the role of this remodeler family, that is, creating NFR in the regulatory regions.

Fig. 3. — ChIP-seq analysis of gSNF2. (A) Graphical views of ChIP-seq peaks of two experiments. Peaks on the chromosome 14 are shown. (B) Common peaks between two experiments. Peaks identified by independent experiments were judged as common when they were within 150 bp. The plot shows that nearly 85% of the peaks in experiment 1 had counterparts in experiment 2 within 150 bp. (C) Proportion of gSNF2 peaks identified on intergenic regions and gene bodies are shown in pie graphs. (D) Ten and five-base motifs enriched under gSNF2 peaks. Logos were created using Weblogo (21). The histogram shows distances of peak summits from motifs. (E) gSNF2 peaks upstream of female-specific genes (experiment 1). ChIP-seq peaks of AP2-FG is shown in parallel. Bars indicate the positions of ten-base binding motifs under AP2-FG peaks. (F) Heat maps were created from ChIP-seq data of AP2-FG (*Left*) and gSNF2 (experiment 1, *Right*) by positioning AP2-FG peaks in the center. Peaks were sorted by fold enrichment values calculated using MACS2 software. (G) *gSNF2* peaks upstream of male-specific genes (experiment 1). Bars indicate positions of five-base motifs under gSNF2 peaks are indicated. (H) Five-base motifs enriched under gSNF2 peaks after subtraction of those common with AP2-FG peaks. Logos were created using Weblogo. The histogram shows the distances of peak summits from motifs.

Statistical analysis of gSNF2 peak data showed that two motif sequences were highly enriched around the predicted summits: ten-base sequence TGTAYRTACA and five-base sequence YGTCT (Fig. 3D and Datasets S6 and S7). The ten-base and five-base motifs were clustered around the summit of the peak (96.5% of the peaks harbored either of the two motif sequences and 80.7% of them were within 100 bp from their summits, Fig. 3D). Importantly, the ten-base motif was identical to female-specific cis-acting element, binding motif of a female-specific TF AP2-FG. The peaks with this motif were located upstream of female-specific genes and the peak positions coincided with ChIP-seq peaks of AP2-FG (Fig. 3E and SI Appendix, Fig. S2A). As shown in the heat map, gSNF2 peaks corresponded to AP2-FG peaks throughout the genome (Fig. 3F and SI Appendix, Fig. S2B). These results suggest that a subset of gSNF2 molecules are recruited upstream of female-specific genes by AP2-FG. On the other hand, the five base motifs were observed under gSNF2 peaks that lacked corresponding AP2-FG peaks. Notably, these peaks were located upstream of male-specific genes including genes encoding flagella components and other well-characterized male-specific genes that are important for gametogenesis and fertilization (examples are shown in Fig. 3G and SI Appendix, Fig. S2C) (10, 22–25). This suggested that gSNF2 peaks are composed of two groups: those associated with ten-base motif (thus common with AP2-FG peaks) and present upstream of genes expressed in females and those associated with five-base motif and present upstream of genes expressed in males.

To confirm this, peaks common with AP2-FG peaks were subtracted from gSNF2 peaks. Between gSNF2 and AP2-FG, 575 peaks were common, and 728 peaks remained after subtraction (Datasets S5 and S8). Statistical analysis of the remaining peaks demonstrated that the ten-base motif was absent around the summit of peaks, and instead, five-base motif became most enriched (Dataset S9). Among these peaks, 97.1% contained the five-base motifs within the peak regions, and 85.0% of peaks were within 100 bp from the summits (Fig. 3H). These results showed that the gSNF2 peaks with five-base motifs constituted a separate group relative to the gSNF2 peaks with ten-base motifs. We further determined genes harboring these peaks within 1,200-bp upstream of the first methionine codon and identified 502 genes as putative targets (Dataset S10). According to sex-specific RNA-seq data (Dataset S2), these putative targets constituted 54.6% of genes enriched in males (239 of 438 genes). In contrast, they constituted only 5.6% of genes enriched in females (28/504). Statistical analysis with Fisher’s exact test demonstrated that these peaks are highly enriched in the upstream regions of male-enriched genes (P = 2.96 × 10⁻¹³⁷) and significantly rare in the upstream regions of female-enriched genes (P = 5.87 × 10⁻⁵) compared to other genes. Taken together, these results suggested that gSNF2 molecules are recruited to ten- and five-base motifs, which are present upstream of genes expressed in females and males, respectively.

Five and Ten-Base Motif Sequences Are Essential for the Recruitment of gSNF2.

Next, to examine whether these motifs are essential for recruitment of gSNF2, ChIP–qPCR assay was performed. Parasites with mutated motif sequences under the summits of gSNF2 peaks were prepared using the CRISPR/Cas9 system, and effects of the mutations on binding of gSNF2 to these sites were assessed using ChIP-qPCR with anti-GFP antibodies. For assay of ten-base motif, a motif sequence upstream of the CPW-WPC protein family gene (PBANKA_1346300) was mutated (Fig. 4A) (26). Mice were infected with these parasites, and the gametocyte-enriched blood was subjected to ChIP as described above. Addition of mutations led to a reduction in binding of gSNF2 to the promoter, while binding to promoters of other target genes was not affected (Fig. 4A). For assay of the five-base motif, motifs upstream of the dynein gene (PNANKA_0416100) were mutated. Binding of gSNF2 to the promoter was significantly reduced in the mutant parasites (Fig. 4B). These results demonstrated that the recruitment was dependent solely on the motifs in the promoter and that independent sequence-specific TFs, that is, AP2-FG in females and a sequence-specific TF binding to the five-base motif in male, determined genomic regions of gSNF2 recruitment.

Fig. 4. — gSNF2 is recruited to male- or female-specific cis-acting elements. (A) The diagram shows a gSNF2 peak upstream of the CPW-WPC family protein gene (PBANKA_6346300). Ten-base motif under the peak is indicated. This ten-base motif was mutated using the CRISPR/Cas9 system in parasites expressing GFP-fused gSNF2 by adding three point mutations. Sequence data around the ten-base motif of these parasites are shown on the lower side of the diagram (mutated nucleic acids are highlighted). ChIP-qPCR experiments of gSNF2 were performed in mutated and original parasites (expressing GFP-fused gSNF2), and the percentage of input was compared. *P28*, a target of gSNF2 expressed in females was the positive control and *TRAP*, a sporozoite-specific gene was negative control. The data are mean ± SE of three biologically independent experiments. (B) Two five-base motifs indicated in the diagram under the summit of gSNF2 peak upstream of the dynein heavy chain gene (PBANKA_0416100), were mutated busing the CRISPR/Cas9 system in parasites expressing GFP-fused gSNF by adding one point mutation to each motif. Sequence data around the five-base motifs of these parasites are shown on the lower side of the diagram (mutated nucleic acids are highlighted). ChIP-qPCR experiments of gSNF2 were performed in mutated and original parasites, and the percentage of input was calculated. The gene encoding a dynein light chain (PBANKA_1133600) and HAP2 are targets of gSNF2 expressed in males and used as positive controls. The data are mean ± SE of three biologically independent experiments. (C) Promoter assay was performed in parasites transfected with a centromere plasmid. The promoter of the dynein gene was inserted upstream of the reporter gene (*mCherry*) in the centromere plasmid. The GFP gene in the plasmid was under the control of the *P. berghei* elongation-factor promoter for constitutive expression. Flow cytometric analyses were performed in parasites transfected with a centromere plasmid containing a mutated (*Left*) and an original wild-type (*Right*) promoter. Mutations were added to the same positions as shown in B.

Five-Base Motif Acts as a Male-Specific Cis-Acting Element.

Ten-base motifs act in female gametocytes as a cis-acting element, and expression of many female-specific genes is under the control of this element (8). However, male-specific cis-acting elements remain unknown. We determined the sequences that are enriched in the upstream regulatory region of male-enriched genes by performing Fisher’s exact test in the upstream region (300 to 1.200 bp from the first methionine codon) between male-enriched genes and other genes. Among top 20 six-base sequences with the smallest P value, 16 sequences contained this motif sequence or its reverse complimentary sequence AGACA (Dataset S11), and among the five-base sequences, this motif sequence and its reverse complementary sequence had the smallest P values (P = 2.37 × 10⁻³⁹ and P = 2.39 × 10⁻²⁹, Dataset S12). Together with the results showing that expression of male-specific genes widely decreased in SNF2(−) parasites and that gSNF2 peaks with the five-base motif are highly enriched upstream of male-enriched genes, this result strongly suggested that the five-base motif is a male-specific cis-acting element.

We validated this hypothesis using in vivo promoter assay using a P. berghei centromere plasmid (27). The male-specific dynein gene (PNANKA_0416100) was used for this assay. Male-specific expression of this gene is well established (20), and the promoter of this gene has been used for expression of male-specific maker gene using centromere plasmids(8). The two motif sequences in 1.5 kbp upstream region of this gene were mutated as in ChIP-qPCR assay (Fig. 4B) and inserted upstream of the marker gene red fluorescent protein mCherry in a centromere plasmid. Transfected parasites were selected using a pyrimethamine-resistant marker, and the infected blood was subjected to FACS. As shown in Fig. 4C, intensities of fluorescent signals of the reporter gene decreased approximately five times compared with those in wild-type promoter, supporting the assumption that it was a male-specific cis-acting element.

gSNF2 Broadly Targets Male-Enriched Genes.

Targets of gSNF2 were predicted from the ChIP-seq data. A total of 1,209 and 1,239 genes (Datasets S13 and S14, respectively) were predicted as targets in experiments 1 and 2, respectively, and 90.0% of the targets (1,089/1,209) were shared between them (Dataset S15). These common target genes were classified into several groups based on the annotation in PlasmoDB (Fig. 5A and Dataset S16). They contained genes for zygote/ookinete development, which are expressed in females, genes related to flagella formation and genome replication, which are expressed in males, and genes related to both sexes, such as osmiophilic body protein genes.

Fig. 5. — Target genes of gSNF2 and male-specific targets. (A) Predicted target genes of gSNF2 were classified into functional categories according to functional annotation in PlasmoDB, and the result is shown in a pie graph. The number on the graph is the total number of genes in each group. Morn-repeat proteins were classified into the group “genome replication and cell division” (28). (B) Proportion of gSNF2 targets (after subtraction of AP2-FG targets) in male-enriched genes. Male-enriched genes (438 genes) were sorted according to the expression values compared to that in females (Dataset S18), and the number of gSNF2 targets in every 100 genes is shown in a bar graph. (C) AP2-FG targets were subtracted from the whole gSNF2 targets, and the remaining genes were classified. (D) Series of processes for genome replication from origin recognition to DNA synthesis is illustrated using genes classified into the group genome replication and cell division in C. Genes newly classified to this group in this study from “genes of unknown function” are marked with asterisks.

Next, to select male-specific targets, AP2-FG targets (823 genes) (8) were subtracted from the common gSNF2 targets, which left 637 genes (Datasets S15 and S17). These predicted male-specific targets covered 67.1% (294/438) of the genes enriched in males (P = 3.98 × 10⁻¹⁷⁵). When male-enriched genes were ordered based on the fold enrichment value (compared to their expression in females), the rates of coverage increased along with the fold enrichment value and reached 79.0% (158/200) for the top 200 genes (P = 1.71 × 10⁻¹⁰⁵) (Fig. 5B and Dataset S18). This observation showed that the majority of genes that were highly and uniquely expressed in males are targets of gSNF2. Classification of these genes (Fig. 5C and Dataset S19) showed that they contain genes grouped as “genomic replication” and “flagellar-based motility” as major constituents. These groups contained a comprehensive set of genes for DNA replication (Fig. 5D) and flagella components (SI Appendix, Fig. S3), suggesting that gSNF2 targets male genes directly and comprehensively.

In addition to these two major groups, “gametogenesis and fertilization,” “cell signaling,” and “cell cycle” contained genes related to male functions. The group “gametogenesis and fertilization” contained genes encoding 6-cys family proteins, gamete fusion protein HAP2, and other male-specific proteins (22–25, 29, 30). The group “cell signaling” contained genes involved in calcium signal transduction and cAMP signaling (31–33). The group “cell cycle” contained genes involved in cell-cycle progression, which may be required for regulating cycles of genome replication and cytokinesis to produce multiple microgametes.

We further explored genes belonging to these five groups in genes classified as genes of unknown functions in PlasmoDB by analyzing their amino acid sequences with the Blastp program and the domain search database Smart. Sixteen genes were identified as genes possibly related to male-specific function (Dataset S20).

gSNF2 Disruption Reduced NFR Formation Upstream of Target Genes.

According to the target prediction by ChIP-seq, 123 of 167 genes significantly decreased by the disruption were target genes of gSNF2 (Dataset S1). To demonstrate that the decrease of expression in these target genes was correlated with reduced NFR formation, we performed ATAC-seq (assay for transposase-accessible chromatin with high-throughput sequencing). Two independent sets of ATAC-seq analyses were performed between the wild-type and gSNF2-disrupted parasites (experiments 1 and 2) using gametocyte-enriched blood, which was prepared as described above. The wild-type and gSNF2-disrupted parasites presented rates of female to male gametocytes of 2.8 and 2.3 in experiment 1, respectively, and 3.0 and 2.7 in experiment 2, respectively. Gametocytes were purified based on a density gradient from the gametocyte-enriched blood, and the nuclei of these parasites were incubated with Tn5 transposome. Reads obtained using ATAC-seq were mapped to the genome (Fig. 6A), and the average read coverage was calculated around the summits of gSNF2 peaks upstream of the target genes (within 200 bp from summits of gSNF2 peaks), which were normalized based on the average coverage of the whole intergenic regions (Datasets S21 and S22). Among the 113 and 94 genes with significantly decreased expression after gSNF2 disruption (10 and 29 target genes were excluded because of low coverage around the summits of gSNF2 peaks), decrease of coverage was observed in 86 and 69 genes (in experiments 1 and 2, respectively). Fisher’s exact test carried out with these results indicated that decreased gene expression was closely related to decreased NFR formation (Fig. 6B). Examples of imaginary views of ATAC-seq peaks in experiments 1 and 2 are shown in Fig. 6C and SI Appendix, Fig. S4A, respectively.

We further calculated the correlation coefficients of these data using whole target genes. As shown in scatter plots (Fig. 6D and SI Appendix, Fig. S4B), positive correlations of 0.60 and 0.54 (experiments 1 and 2, the same below) were observed between the decreased gene expression and read coverage. When male- and female-enriched genes were analyzed separately, strong correlations of 0.68 and 0.66 were observed in male-enriched genes. In contrast, no correlation was observed in female-enriched genes. These results indicate that gSNF2 contributes to transcriptional activation of male genes through NFR formation.

Discussion

Plasmodium parasites pass through several stages in their lifecycle and display stage-specific gene expression to adapt to the host environments. Chromatin remodeling could play roles in the gene regulation of these stages, but thus far, clear evidence has not been obtained. The present study revealed that SNF2-like ATPase, a core subunit of the SNF/SWI remodeler, is recruited upstream of genes that constitute a major repertoire of gene expression in males and create NFR, initiating transcription of these genes. The result encourages us to investigate whether chromatin remodelers are involved in the establishment of stage-specific gene expression programs in other life stages.

This study also revealed that gSNF2 is recruited to the five-base motif that acts as a male-specific cis-acting element. This element broadly covers genes expressed in male gametocytes, suggesting that a sequence-specific TF, which regulates the male-specific gene expression, is involved in this recruitment. It is possible that this TF recruits gSNF2 as an initial step of target activation and creates an opened promoter, followed by the recruitment of the mediator complex (12), leading to transcriptional activation of target genes. We speculate that different impacts of gSNF2 disruption on target gene expression would be correlated with the importance of remodeling by gSNF2 between genes in this initial step. Effects of disruption would be smaller in genes that already contain an upstream NFR, while the effects would be more pronounced in the genes that harbor an upstream closed promoter. Identification of the associated TF and elucidation of the complete process of target gene activation are necessary in the future.

gSNF2 is also recruited, presumably by AP2-FG, to female-specific cis-acting elements, which is similar to the proposed initial step of gene activation in males. However, a clear effect of the disruption of gSNF2 was not observed in females. Therefore, it can be concluded that remodeling by gSNF2 is not important for female differentiation. We speculate that two possibilities might explain this difference between sexes. One is that another remodeler is additionally expressed during female development and/or mechanisms such as those involving histone modifying enzymes complement the effects of gSNF2 disruption in females. Another possibility is that the chromatin structure of early gametocytes is similar to that of females and thus would not require major changes during female differentiation, that is, contribution of chromatin structure on female differentiation is insignificant; therefore, detection of effects of gSNF2 disruption on female gametocyte is difficult. The second hypothesis implies that promoters of the female-specific genes are already open without NFR formation by gSNF2, while male gametocytes have to largely change their chromatin structure to accomplish differentiation. This implies that early gametocytes are immature female-like cells, and a subset of them differentiate into males by underdoing epigenetic regulation to form a distinct cell population. To validate this hypothesis, technical improvement is necessary which will allow purification of early gametocytes and comparing their chromatin structure with that of mature female and male gametocytes.

In conclusion, this study revealed that genome-wide changes in chromatin structure by gSNF2 are important mechanisms for sexual differentiation in Plasmodium parasite and provided the possibility that sex-specific master TFs play a role in its recruitment. Future studies to determine whether this mechanism is specific to sexual differentiation or similar mechanisms are employed generally for transcriptional regulation in other stages in the lifecycle (particularly during stage conversion) of the parasite are important. The elucidation of mechanisms that establish stage-specific gene expression programs would enhance our understanding of the parasitic strategies, leading to discovery of novel targets for prevention of malaria.

Materials and Methods

Parasite Preparations.

ANKA strain of P. berghei was maintained in female BALB/c mice (6 to 10 wk old). To examine the oocyst number, infected mice were subjected to Anopheles stephensi mosquitoes. Fully engorged mosquitoes were selected and maintained at 20 °C. The number of oocysts and oocyst sporozoites was counted at 14 d after the infected blood meal. To prepare mature schizont, infected mice blood was cultured in the culture medium (RPMI supplemented with 10% fetal calf serum) for 16 h. Mature schizonts were purified using a density gradient of OptiPrepp^TM (1.077 g/mL). To prepare gametocyte-enriched blood, mice were pretreated with phenyl-hydrazine and infected with P. berghei by intraperitoneal injection of the infected blood. After parasitemia increased over 1%, they were treated with sulfadiazine for 2 d in drinking water (20 mg/L) to kill asexual parasites. After checking exflagellation rates (>250/1 × 10⁵ RBC), whole blood was collected for following experiments.

Preparation for Transgenic Parasites.

gSNF2::mNEON parasites expressing sex-specific marker gene were prepared as reported previously (8). Briefly, gSNF2::mNEON parasites were transfected with the pCen vector containing the expression cassette of the sulfadiazine-resistant P. falciparum DHPS (dihydropteroate synthase) gene as a selectable marker and the mCherry expression cassette for conferring parasites with sex-specific expressions of mCherry. The promoter regions of CCP2 (PBANKA_1319500) and the dynein heavy chain gene (PBANKA_0416100) were used for female- and male-specific expression of mCherry, respectively. Transfected parasites in drinking water (10 mg mL⁻¹) were selected with sulfadiazine.

To prepare gSNF2::mNeon parasites, the mNeonGreen gene was fused to the 3′-portion of the gSNF2 gene using double cross-over homologous recombination as reported previously (5). Briefly, DNA fragments corresponding to 3′-portions of the gSNF2 gene were amplified using PCR using genomic DNA as a template and were subcloned into the vector containing the mNeonGreen gene, 3′-terminal portion of the HSP70 gene, and a selectable maker cassette for expressing the human DHFR gene. Cultured mature merozoites were transfected with the linearized construct and injected intravenously into mice. These mice were treated with pyrimethamine to select for parasites integrated with the construct. Parasites were cloned using limiting dilution to obtain transgenic parasite clones. Parasites expressing GFP-fused gSNF2, gSNF2::GFP, were prepared using the same procedure. For the preparation of gSNF2(−) parasites, targeting construct, DNA fragments for homologous recombination were annealed to each side of DNA fragments containing human DHFR as a selectable marker gene using overlapping PCR. Transfection, drug selection, and cloning of transgenic parasites were performed as described above.

Transgenic parasites for ChIP-qPCR assays, which harbored the mutated sequences upstream of a gSNF2 target gene, were prepared using CRISPR/Cas9 method reported previously (34). Briefly, cultured mature merozoites of CAS9-expressing P. berghei parasites were transfected with a gRNA plasmid vector and a linear DNA template containing mutations in the motif sequences and were injected intravenously into mice. These mice were treated with pyrimethamine for 2 d to eliminate parasites lacking the mutation. These parasites were cloned using limiting dilution. Mutations of the motifs were confirmed using direct sequencing. To prepare parasites containing mutations in five-base motif sequences upstream of the dynein gene, mutations were added in two steps due to lack of appropriate protospacer adjacent motif sequences around these motifs (SI Appendix, Fig. S1E). Briefly, mutants with additional mutations near the locus were prepared using the CRISPR/Cas9 method and thereafter these mutations were returned to wild-type sequences using the same method. The GFP gene was fused to 3′-end of the gSNF2 gene of motif-mutated parasites using double cross-over homologous recombination using a construct containing pyrimethamine-resistant selectable maker as in gSNF2::GFP parasites described above.

Flow Cytometric Analysis.

Flow cytometric analysis was performed using an LSR-II flow cytometer (BD Biosciences) as described previously (6). Briefly, in experiments using 820 parasites, the blood obtained from an asynchronously infected mouse was selected by gating on forward and side scattering, and then, the gated population was analyzed on fluorescence of GFP and RFP (575/26 and 530/30, respectively). In a promoter assay using a centromere plasmid, the blood obtained from an asynchronously infected mouse was selected by gating on forward and side scattering and then on fluorescence of GFP. The selected population was analyzed on fluorescence of mCherry (575/26). The analysis was performed using the DIVER program (BD Biosciences).

ChIP-seq.

Gametocyte-enriched blood was collected from mice, filtered with a Plasmodipur filter to remove white blood cells, and fixed in 1% paraformaldehyde for 30 min at 30 °C. Two mice were used for each experiment. ChIP was performed as previously described (35). Briefly, red blood cells were lysed in 0.84% NH₄Cl and residual cells were lysed in lysis buffer containing 1% SDS. The lysate was sonicated in Bioruptor 2 and centrifuged at 4 °C for 30 min at 14,000 rpm. The supernatant was diluted with dilution buffer and subjected to ChIP with anti-GFP antibody. Library was prepared using hyper prep kit (Kapa biosystems). Sequencing was performed using NextSeq sequencer (Illumina). The data have been deposited in the Gene Expression Omnibus (GEO) database under accession number GSE198588 (36).

Analysis of ChIP-seq Data.

Sequence data were mapped onto P. berghei genome sequence (PlasmoDB, version 3) using BOWTIE2 software with default settings (37). Mapping data were analyzed using MACS2 peak-calling algorithm. Conditions for peak calling were FDR < 0.01 and fold enrichments > 3 (experiment 1) or 2.2 (experiment 2). Option “Callsummit” was used to detect large peaks constituted from multiple sub-peaks, because in default settings, this software sometimes misses large peaks composed of multiple peaks in analysis of ChIP-seq data obtained in this parasite. To identify motif sequences associated with ChIP-seq peaks, sequences concentrated around the predicted summits of ChIP-seq peaks were investigated using Fisher’s exact test carried out between 100-bp regions with summits in the center and 100-bp regions excised from the genome excluding the former regions to cover the entire genome sequence. Genes were determined as targets when their 1.2-kbp upstream regions contained the predicted summits of ChIP-seq peaks. When the upstream intergenic region was less than 1.2 kbp, the whole upstream intergenic region was used for prediction. To search for exceptional target genes that are transcribed from more than 1.2-kbp upstream and harbored predicted summit beyond the 1.2-kbp criterion, upstream regions of genes that were not predicted as targets were manually examined using RNA-seq and ChIP-seq data on the integrative genomics viewer (IGV) (38).

ChIP-qPCR.

ChIP for ChIP-qPCR assays was performed in the same manner as ChIP-seq described above. Briefly, mice were pretreated with phenyl-hydrazine and infected with original or motif-mutant parasites. After sulfadiazine treatment, whole blood was collected and passed through the filter to remove white blood cells. One infected mouse was used for each ChIP experiment, and three independent ChIP experiments were performed. Obtained DNA was subjected to qPCR, and values of % of input DNA (DNA extracted from lysate before ChIP) were compared between the different P. berghei parasites expressing GFP-fused gSNF2 and those with the mutated motifs. Primers used for this assay are listed in Dataset S23.

RNA-seq Analysis.

Gametocyte-enriched blood was subjected to erythrocyte lysis in 0.84% NH₄Cl. Total RNA was extracted from residual cells using Isogen 2 (Nippon Gene, Tokyo, Japan) according to the manufacturer’s protocol. Libraries were prepared using the RNA hyperprep kit (KAPA Biosystems), and sequencing was performed using an Illumina NEXTseq sequencer. The data have been deposited in the GEO database under accession number GSE198588.

Analysis of RNA-seq Data.

Three independent experiments were performed in both wild-type and gSNF2-disrupted parasites. Single read sequences were mapped on P. berghei genome (version 3), and read number on each gene was counted using Feature Counts software (39). TPM (transcripts per million) was calculated in each gene, and only genes in which TPM maximum > 20 in wild-type parasites were used for the following analyses (genes of low expression in wild-type parasites were excluded). Additionally, genes located in the subtelomeric regions were excluded due to variable expression among clones. Ratio of read numbers and padj were calculated for each gene between wild-type and gSNF2-disrupted parasites using DEseq2 software (40). The volcano plots of the obtained results were created using DDplot2 software. The mapping data were visualized using the IGV software.

Male- and female-enriched genes were determined using RNA-seq data of male, female, and asexual stages reported by Witmer et al. (11). Gene expressions were compared using DEseq2 between these stages. Genes with a difference in expression of over fourfolds and padj < 0.001 as compared to the other two stages were designated as male- and female-enriched genes. Genes whose FPKM (fragments per kilo base of transcript per million mapped fragments) values were under 10 and genes located in the subtelomeric regions were excluded from the analyses.

ATAC-seq.

Gametocyte-enriched blood was prepared as described above and uninfected red blood cells were removed using a density gradient of OptiPrepp^TM (1.077 g/mL). Purified cells were washed in the culture medium. ATAC-seq was performed using ATAC-Seq kit (Active Motif) according to the manufacturer’s instructions. Briefly, 1 × 10⁶ infected erythrocytes were washed in PBS and lysed in a cell lysis buffer. Nuclei were collected by centrifugation and incubated with Tn5 transposome for 30 min at 37 °C in tagmentation buffer with constant swirling. Genomic DNA was purified, and libraries were prepared using PCR with primers for Illumina sequencing. The data have been deposited in the GEO database under accession number GSE198588.

Analysis of ATAC-seq Data.

Paired-end read sequences were mapped using Bowtie 2 software (version 3) on the P. berghei genome. Duplicate fragments mapped on the genome were removed using Picard MarkDuplicates application (http://picard.sourceforge.net). Peak calling using MACS2 software was performed using these mapping data with the following settings: —shift −100, —extend 200, —nonmodel, and —keepduplicates all. Generated bedgraph files were converted to vigwig format using Wig/BedGraph-to-bigWig software. Read numbers were counted using Deeptools software (41) in the regions within 200 bp from gSNF2 peak summits upstream of gSNF2 target genes. These data were normalized with total numbers of reads mapped on the intergenic regions, and ratios of read coverages were calculated between wild-type and gSNF2-disrupted parasites.

Supplementary Material

Appendix 01 (PDF)

Click here for additional data file.^{(719.5KB, pdf)}

Dataset S01 (XLSX)

Click here for additional data file.^{(322.2KB, xlsx)}

Dataset S02 (XLSX)

Click here for additional data file.^{(17.6KB, xlsx)}

Dataset S03 (XLSX)

Click here for additional data file.^{(122KB, xlsx)}

Dataset S04 (XLSX)

Click here for additional data file.^{(117KB, xlsx)}

Dataset S05 (XLSX)

Click here for additional data file.^{(102.3KB, xlsx)}

Dataset S06 (XLSX)

Click here for additional data file.^{(9.8KB, xlsx)}

Dataset S07 (XLSX)

Click here for additional data file.^{(9.8KB, xlsx)}

Dataset S08 (XLSX)

Click here for additional data file.^{(59.5KB, xlsx)}

Dataset S09 (XLSX)

Click here for additional data file.^{(9.7KB, xlsx)}

Dataset S10 (XLSX)

Click here for additional data file.^{(22.2KB, xlsx)}

Dataset S11 (XLSX)

Click here for additional data file.^{(8.9KB, xlsx)}

Dataset S12 (XLSX)

Click here for additional data file.^{(8.9KB, xlsx)}

Dataset S13 (XLSX)

Click here for additional data file.^{(44.2KB, xlsx)}

Dataset S14 (XLSX)

Click here for additional data file.^{(45KB, xlsx)}

Dataset S15 (XLSX)

Click here for additional data file.^{(36.9KB, xlsx)}

Dataset S16 (XLSX)

Click here for additional data file.^{(29.6KB, xlsx)}

Dataset S17 (XLSX)

Click here for additional data file.^{(25.2KB, xlsx)}

Dataset S18 (XLSX)

Click here for additional data file.^{(35.7KB, xlsx)}

Dataset S19 (XLSX)

Click here for additional data file.^{(20.8KB, xlsx)}

Dataset S20 (XLSX)

Click here for additional data file.^{(8.9KB, xlsx)}

Dataset S21 (XLSX)

Click here for additional data file.^{(55.1KB, xlsx)}

Dataset S22 (XLSX)

Click here for additional data file.^{(54.4KB, xlsx)}

Dataset S23 (XLSX)

Click here for additional data file.^{(9.1KB, xlsx)}

Acknowledgments

This work was supported by JSPS KAKENHI Grant No. 17H01542 to M.Y.

Author contributions

I.K., T.N., and M.Y. designed research; I.K., T.N., and M.Y. performed research; I.K., T.N., S.I., and M.Y. analyzed data; and I.K. and M.Y. wrote the paper.

Competing interests

The authors declare no competing interest.

Footnotes

This article is a PNAS Direct Submission.

Data, Materials, and Software Availability

NGS data have been deposited in Gene Expression Omnibus (GSE198588) (36). All study data are included in the article and/or supporting information.

Supporting Information

References

1.Ngwa C. J., Rosa T. F. D. A., Pradel G., “The biology of malaria gametocytes” in Current Topics in Malaria (IntechOpen, London, United Kingdom, 2016), pp. 117–144, 10.5772/65464. [DOI] [Google Scholar]
2.Sinha A., et al. , A cascade of DNA-binding proteins for sexual commitment and development in Plasmodium. Nature 507, 253–257 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Kafsack B. F. C., et al. , A transcriptional switch underlies commitment to sexual development in malaria parasites. Nature 507, 248–252 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Kent R. S., et al. , Inducible developmental reprogramming redefines commitment to sexual development in the malaria parasite Plasmodium berghei. Nat. Microbiol. 3, 1206–1213 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Yuda M., Kaneko I., Murata Y., Iwanaga S., Nishi T., Mechanisms of triggering malaria gametocytogenesis by AP2-G. Parasitol. Int. 84, 102403 (2021). [DOI] [PubMed] [Google Scholar]
6.Yuda M., Iwanaga S., Kaneko I., Kato T., Global transcriptional repression: An initial and essential step for Plasmodium sexual development. Proc. Natl. Acad. Sci. U.S.A. 112, 12824–12829 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Li Z., et al. , Plasmodium transcription repressor AP2-O3 regulates sex-specific identity of gene expression in female gametocytes. EMBO Rep. 22, 1–18 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Yuda M., Kaneko I., Iwanaga S., Murata Y., Kato T., Female-specific gene regulation in malaria parasites by an AP2-family transcription factor. Mol. Microbiol. 113, 40–51 (2019), 10.1111/mmi.14334. [DOI] [PubMed] [Google Scholar]
9.Yeoh L. M., Goodman C. D., Mollard V., McFadden G. I., Ralph S. A., Comparative transcriptomics of female and male gametocytes in Plasmodium berghei and the evolution of sex in alveolates. BMC Genomics 18, 734 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Khan S. M., et al. , Proteome analysis of separated male and female gametocytes reveals novel sex-specific Plasmodium biology. Cell 121, 675–687 (2005). [DOI] [PubMed] [Google Scholar]
11.Witmer K., Fraschka S. A., Vlachou D., Bártfai R., Christophides G. K., An epigenetic map of malaria parasite development from host to vector. Sci. Rep. 10, 1–19 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Soutourina J., Transcription regulation by the Mediator complex. Nat. Rev. Mol. Cell Biol. 19, 262–274 (2018). [DOI] [PubMed] [Google Scholar]
13.Klemm S. L., Shipony Z., Greenleaf W. J., Chromatin accessibility and the regulatory epigenome. Nat. Rev. Genet. 20, 207–220 (2019). [DOI] [PubMed] [Google Scholar]
14.Reyes A. A., Marcum R. D., He Y., Structure and Function of Chromatin Remodelers. J. Mol. Biol. 433, 166929 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Lorch Y., Kornberg R. D., Chromatin-remodeling for transcription. Q. Rev. Biophys. 50, 1–15 (2017). [DOI] [PubMed] [Google Scholar]
16.Byrd A. K., Raney K. D., Superfamily 2 helicases. Front. Biosci. 17, 2070–2088 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Wagner F. R., et al. , Structure of SWI/SNF chromatin remodeller RSC bound to a nucleosome. Nature 579, 448–451 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Sen P., Ghosh S., Pugh B. F., Bartholomew B., A new, highly conserved domain in Swi2/Snf2 is required for SWI/SNF remodeling. Nucleic Acids Res. 39, 9155–9166 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Shaner N. C., et al. , A bright monomeric green fluorescent protein derived from Branchiostoma lanceolatum. Nat. Methods 10, 407–409 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Mair G. R., et al. , Universal features of post-transcriptional gene regulation are critical for Plasmodium zygote development. PLoS Pathog. 6, e1000767 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Crooks G. E., Hon G., Chandonia J. M., Brenner S. E., WebLogo: A sequence logo generator. Genome Res. 14, 1188–1190 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Tachibana M., Ishino T., Takashima E., Tsuboi T., Torii M., A male gametocyte osmiophilic body and microgamete surface protein of the rodent malaria parasite Plasmodium yoelii (PyMiGS) plays a critical role in male osmiophilic body formation and exflagellation. Cell. Microbiol. 20, 1–12 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Hirai M., et al. , Male fertility of malaria parasites is determined by GCS1, a plant-type reproduction factor. Curr. Biol. 18, 607–613 (2008). [DOI] [PubMed] [Google Scholar]
24.Liu Y., et al. , The conserved plant sterility gene HAP2 functions after attachment of fusogenic membranes in Chlamydomonas and Plasmodium gametes. Genes Dev. 22, 1051–1068 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Van Dijk M. R., et al. , Three members of the 6-cys protein family of Plasmodium play a role in gamete fertility. PLoS Pathog. 6, e1000853 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Kangwanrangsan N., et al. , A member of the CPW-WPC protein family is expressed in and localized to the surface of developing ookinetes. Malar. J. 12, 129 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Iwanaga S., et al. , Functional identification of the Plasmodium centromere and generation of a Plasmodium artificial chromosome. Cell Host Microbe 7, 245–255 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Gubbels M. J., Vaishnava S., Boot N., Dubremetz J. F., Striepen B., A MORN-repeat protein is a dynamic component of the Toxoplasma gondii cell division apparatus. J. Cell Sci. 119, 2236–2245 (2006). [DOI] [PubMed] [Google Scholar]
29.Ning J., et al. , Comparative genomics in Chlamydomonas and Plasmodium identifies an ancient nuclear envelope protein family essential for sexual reproduction in protists, fungi, plants, and vertebrates. Genes Dev. 27, 1198–1215 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Deligianni E., et al. , Critical role for a stage-specific actin in male exflagellation of the malaria parasite. Cell. Microbiol. 13, 1714–1730 (2011). [DOI] [PubMed] [Google Scholar]
31.Billker O., et al. , Calcium and a calcium-dependent protein kinase regulate gamete formation and mosquito transmission in a malaria parasite. Cell 117, 503–514 (2004). [DOI] [PubMed] [Google Scholar]
32.Tewari R., Dorin D., Moon R., Doerig C., Billker O., An atypical mitogen-activated protein kinase controls cytokinesis and flagellar motility during male gamete formation in a malaria parasite. Mol. Microbiol. 58, 1253–1263 (2005). [DOI] [PubMed] [Google Scholar]
33.Baro Graf C., et al. , Everything you ever wanted to know about PKA regulation and its involvement in mammalian sperm capacitation. Mol. Cell. Endocrinol. 518, 110992 (2020). [DOI] [PubMed] [Google Scholar]
34.Shinzawa N., et al. , Improvement of CRISPR/Cas9 system by transfecting Cas9-expressing Plasmodium berghei with linear donor template. Commun. Biol. 3, 1–13 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Kaneko I., Iwanaga S., Kato T., Kobayashi I., Yuda M., Genome-wide identification of the target genes of AP2-O, a Plasmodium AP2-family transcription factor. PLoS Pathog. 11, e1004905 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Yuda M., Differentiation of malaria male gametocytes is initiated by recruitment of a chromatin remodeler to male-specific cis-acting elements. Gene Expression Omnibus. https://www.ncbi.xyz/geo/query/acc.cgi?acc=GSE198588. Deposited 14 Mar 2022.
37.Langmead B., Salzberg S. L., Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Robinson J. T., et al. , Integrative genome viewer. Nat. Biotechnol. 29, 24–26 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Liao Y., Smyth G. K., Shi W., FeatureCounts: An efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930 (2014). [DOI] [PubMed] [Google Scholar]
40.Love M. I., Huber W., Anders S., Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 1–21 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Ramírez F., et al. , deepTools2: A next generation web server for deep-sequencing data analysis. Nucleic Acids Res. 44, W160–W165 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]