Distinct Histone Post-translational Modifications during Plasmodium falciparum Gametocyte Development

Abstract

Histones are the building units of nucleosomes, which constitute chromatin. Histone post-translational modifications (PTMs) play an essential role in epigenetic gene regulation. The Plasmodium falciparum genome encodes canonical and variant histones and a collection of conserved enzymes for histone PTMs and chromatin remodeling. Herein, we profiled the P. falciparum histone PTMs during the development of gametocytes, the obligatory stage for parasite transmission. Mass spectrometric analysis of histones extracted from the early, middle, and late stages of gametocytes identified 457 unique histone peptides with 90 PTMs, of which 50% were novel. The gametocyte histone PTMs display distinct patterns from asexual stages, with many new methylation sites in histones H3 and H3.3 (e.g., K14, K18, and K37). Quantitative analyses revealed a high abundance of acetylation in H3 and H4, mono-methylation of H3/H3.3 K37, and ubiquitination of H3BK112, suggesting that these PTMs play critical roles in gametocytes. Gametocyte histones also showed extensive and unique combinations of PTMs. These data indicate that the parasite harbors distinct transcription regulation mechanisms during gametocyte development and lay the foundation for further characterization of epigenetic regulation in the life cycle of the malaria parasite.

Graphical Abstract

INTRODUCTION

Histone proteins are the core units of nucleosomes and the fundamental building blocks of the chromatin. Each nucleosome consists of DNA wrapped in 1.75 superhelical turns around an octamer composed of two monomers of each histone (H2A, H2B, H3, and H4).^1,2 Histones are small, conserved proteins with a common structure containing two flexible domains (N- and C-terminal tails) and a globular domain. Histone tails are exposed outside of nucleosomes and are targets (mainly the N-terminus) of post-translational modifications (PTMs), which may alter the chromatin structure by influencing the interaction between histones and DNA, and the recruitment of proteins recognizing the histone PTMs, the “histone readers”.^3–5 The most common PTMs include acetylation on lysine residues, methylation on lysine or arginine, phosphorylation on serine or threonine, ubiquitination of lysine, and sumoylation of lysine.^6–8 Different types and levels (e.g., mono-, di-, or tri-methylation) of PTMs and their combinations act as an epigenetic marking system, a “histone code”, to dictate the transcriptional status of the genes.^9,10 The histone PTMs are deposited and removed by the reversible actions of histone-modifying enzymes such as histone acetyltransferases, histone deacetylases, histone methyltransferases, and histone demethylases.¹¹

Malaria caused over half a million deaths in 2020 alone, primarily due to Plasmodium falciparum, the most virulent species of the five human malaria parasites.¹² The complex lifecycle of the malaria parasite alternates between the human host and an Anopheles mosquito vector. Infection of the human host starts with the injection of sporozoites from the mosquito bite, which migrate to the liver and undergo schizogony in hepatocytes. The resulting merozoites will invade erythrocytes to initiate the blood-stage cycle. A small portion of the blood-stage parasites will develop into gametocytes, the obligatory stage for transmission to the mosquito. The P. falciparum gametocyte development takes a uniquely long period of ~10 days, undergoing five morphologically distinct stages.^13,14 The Plasmodium lifecycle is orchestrated by a sophisticated transcriptional program,^15–18 where epigenetic regulation plays an essential role.¹⁸

The P. falciparum genome encodes four core histones H2A, H2B, H3, and H4, as well as four variant histones H2AZ, H2BZ, H3.3, and CenH3, where only H3/H3.3 and H4 N-terminus are highly conserved as compared to model organisms.^19–21 Previous studies detected a wide array of conserved and novel histone PTMs during the asexual intraerythrocytic developmental cycle (IDC) of P. falciparum.^20,22,23 Some of the histone PTMs such as H3K9ac, H3K18ac, H3K27ac, H4K8ac, and H3K4me3 are conserved marks for euchromatin,^24–27 whereas H3K9me3 and H3K36me2/3 demarcate the heterochromatin.^24,28–30 The P. falciparum epigenome during the IDC is predominantly euchromatic, while the heterochromatin islands control antigenic variation, drug sensitivity, and gametocytogenesis.^{24,29,31–38} Despite increasing understanding of epigenetic regulation during the IDC, the histone PTMs and chromatin organization during gametocyte development are much less understood. To adapt to the biology of sexual development, the parasite substantially reorganizes its chromatin structure in gametocytes.^38,39 A recent study revealed distinct histone PTMs in gametocytes with abundant repressive marks (H3K20me3, H3K27me3, and H3K36me3) at early stages and active marks (H3K4me1, H3K18ac, H3K27me1, H3K36me1) at the late stage.²³ While this finding suggests divergent epigenetic regulation during gametocyte development, some marks were contradictory and remained to be validated. Herein, we revisited the gametocyte histone PTMs and profiled histone PTMs at three stages during gametocyte development using highly accurate mass spectrometry (MS). We identified a total of 90 PTMs on both canonical and variant histones, of which 45 were found for the first time in the malaria parasite. Quantitative and combination assays confirmed the distinct pattern of PTMs during gametocyte development, suggesting that P. falciparum gametocytes use distinct epigenetic mechanisms of gene regulation.

MATERIALS AND METHODS

Parasite Culture and Gametocytogenesis

P. falciparum clone 3D7 was maintained as described⁴⁰ and synchronized by sorbitol treatment.⁴¹ A modified method was used for gametocyte induction with spent culture media.⁴² Synchronized trophozoite stages were set up and stressed by using conditioning media at the ring stage followed by daily replenishment with fresh media. Heparin (10 units/mL) was added to the gametocyte culture from day 1 for at least three days to clear asexual parasites.⁴³

Extraction and Purification of Histones

Synchronized gametocytes on day 4 (stage II), day 8 (stage III-IV), and day 12 (stage V) were released from erythrocytes by saponin treatment and washed three times with cold phosphate-buffered saline (PBS, pH 7.4). The gametocytes were centrifuged at 500g for 5 min, and the pellet was resuspended in a Triton extraction buffer [TEB, PBS containing 0.5% Triton-X 100, 2 mM phenylmethylsulfonyl fluoride (PMSF), and 0.02% NaN₃] and lysed on ice for 10 min with gentle stirring. After centrifugation at 1000g for 10 min and one wash with TEB, the pellet was resuspended in 0.25 M HCl. After overnight incubation at 4 °C on a rotor, the acid-soluble fractions were pooled and precipitated with an equal volume of 20% trichloroacetic acid, followed by washing with chilled acetone. Then, the air-dried pellet was resuspended in PBS containing a protease inhibitor cocktail for subsequent analyses. At least three biological replicates of the histone extracts for each time point were purified for analysis.

Mass Spectrometry Analysis

Purified histones were analyzed by liquid chromatography coupled to tandem mass spectrometry (LC/MS/MS). Briefly, histone proteins were separated in a 15% SDS-PAGE gel and stained with Coomassie brilliant blue. The entire histone region was excised as three gel slices. In-gel digestion with trypsin, chymotrypsin, or elastase was performed using a robot system (Progest, Digilab). The digests were separated using a Waters NanoAcquity HPLC system and introduced into the Orbitrap mass spectrometer at 7000 and 17,500 FWHM resolution, respectively. The data were used to search the human Plasmodium_20121001 database containing 51436 entries using Mascot version 2.5 and processed using Scaffold version 4.4.5 (Proteome Software, Portland, OR) for validation, filtering, and creation of a non-redundant list per sample. The target decoy option of Mascot was enabled, with peptide mass tolerance and fragment mass tolerance set to 10 ppm and 0.02 Da, respectively. The search parameters were set as follows: (1) fixed modification: carbamidomethylation on cysteine residues; (2) variable modifications: acetylation on lysine and N-terminus; methylation on lysine and arginine residues; di-methylation on lysine; tri-methylation on lysine; ubiquitination on lysine and phosphorylation on serine and threonine; (3) digestive enzyme: non-specific (chymotrypsin and elastase) and strict trypsin with up to two missed cleavage sites. Search results were imported into Scaffold, and the target decoy option was enabled as 90% minimum protein threshold with at least two peptides, 80% peptide threshold. The false discovery rate (FDR) was set as 0.1% for peptide and 0.9% for protein. Scaffold results were exported as mzIdentML and imported into Scaffold PTM to assign site localization probabilities using A-Score.⁴⁴ Data were validated with Scaffold PTM using visual inspection of all MS/MS spectra based on mass accuracy of precursor and fragment ions, presence of a, b, y, and immonium ions, and absence of undetermined peaks. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE⁴⁵ partner repository with the data set identifier PXD031817 and 10.6019/PXD031817.

Quantitative PTM and Combinatorial Probabilities of PTM Co-occurrence Analysis

The modification levels were calculated based on local amino acid spectral counts within Plasmodium histones by NSAF7.^20,46

$$\text{PTM level in AAx}=\frac{\text{AAx PTM spectral counts}}{\text{AAx total spectral counts}}$$

Any PTM in which the corresponding peptide was not discovered in any replicate will be removed, and the PTM levels from at least two-thirds of replicates in each time point were combined to form the final PTM values.

The probabilities of combinatorial associations between modified residues were calculated based on the established method.^20,47 Frequencies of co-occurrence were calculated between individual modified residues such as probability for combinatorial associations of a and b in the condition a was calculated as the spectral counts of peptides bearing both a and b modified residues divided by spectral counts of peptides bearing modified residue a.

Western Blot

Western blot was performed to detect several histone PTMs during gametocyte development. Histone proteins in the parasites at the asexual and different gametocyte stages (≥10⁹) were separated in a 15% SDS-PAGE gel and transferred to a nitrocellulose membrane, which was then probed with primary antibodies against anti-dimethyl-histone H3K14 (A5278, ABclonal), monomethyl-histone H3R26 (A3163, ABclonal), monomethyl-histone H3K27 (A2361, ABclonal), and anti-acetyl histone H4 (catalog no. 06–598) at 1:1000 dilution; anti-monomethyl histone H3K18 (catalog no. A68374–050, abcam), anti-monomethyl histone H3K37 (catalog no. A68320–050, abcam), and anti-dimethyl histone H3K37 (catalog no. 600–401-I91, Rockland antibodies) at 1:500 dilution, and anti-dimethyl-Histone H3K18 (NB21–1142, Novus biologicals) at 1:2000 dilution followed by detection with the secondary anti-rabbit HRP-conjugated antibody at 1:5000 (Sigma-Aldrich). For all western blot experiments, anti-histone H3 (catalog no. 06–755, Millipore) at 1:1000 dilution was used as a loading control and detected with anti-rabbit HRP-conjugated antibody as above. The detected proteins were visualized using an enhanced chemiluminescence kit (Invitrogen).

Data Assembly and Statistical Analysis

Mass spectrometry results from three different digestions were merged for each replicate of the three gametocyte stages. A pairwise Spearman correlation and K-means clustering of the quantitative histone PTM profiles at three gametocyte developmental stages were performed using R. The results of Spearman correlation were plotted with ggplot2⁴⁸ and the output of the hierarchical clustering with PTM values (heatmap) was plotted using the package “ComplexHeatmap”.⁴⁹

RESULTS

High-Coverage MS Analysis of Histones from P. falciparum Gametocytes

Total histone proteins were extracted using strong acid from three developmental stages of gametocytes, representing early (day 4, stage II), middle (day 8, stage III–IV), and late (day 12, stage V) stages. The histone peptides were identified by high-accuracy LC–MS/MS. For each gametocyte stage, three biological replicates were included. All canonical histones (H2A, H2B, H3, and H4) and three variant histones (H2A.Z, H2B.Z, and H3.3) were identified in each gametocyte stage. In previous studies, the centromeric histone CenH3 was the least abundant in asexual parasites and was only detected at low coverage from stage II and IV gametocytes.^20,24,50 We did not detect this protein in the gametocyte histone extracts. The histone variants H2A.Z and H3.3 were identified in the gametocyte histone extracts at high abundance (Figure 1A, Table S1), contrasting the findings from a previous study.²³ In agreement with an early study,²³ H2A showed decreased abundance in the late gametocyte stage (Figure 1A, Table S1). To enhance the coverage of histone sequence, especially for H2A, we performed three more histone extractions from late-stage gametocytes, resulting in >90% coverage for almost all histones and 99% for H2A (Figure 1B, Table S2). Collectively, we have achieved an overall sequence coverage of >80% for all the seven histones in all three gametocyte stages (Figure 1B, Table S2).

Figure 1.

P. falciparum histone prevalence during gametocyte development. (A). Relative protein abundance of the seven P. falciparum histones over the sexual development by comparing the spectral counts in each purified histone sample. The relative abundance (%) of the core and variant histones are shown at early (day 4, stage II), middle (day 8, stage III–IV), and late (day 12, stage V) stages of gametocytes. (B) Histone sequence coverage (%) for the seven P. falciparum histones identified from mass spectrometry analysis of each purified histone sample is shown for each gametocyte’s developmental time points.

Distinct Histone PTMs in P. falciparum Gametocytes

Our study identified 457 unique histone peptides, with 82 (18%) containing modifications. After manual validation of the spectra, 90 PTMs were identified in these peptides based on the A-Score probabilities⁴⁴ (Figure 2, Tables 1, S3). In some peptides, especially at the H3 and H3.3 N-termini, the same residues were found to have different types and levels of modification. Figure 3 shows a prominent example of the spectra of the histone H3.3 peptide KSTGGKAPR, carrying acetylation and mono-, di-, and tri-methylation at the K9 residue. The high accuracy of the MS results enabled confident identification of the mass increase of a methyl group by 14 Da, dimethyl group of 28 Da, and trimethyl group of 42.047 Da, as well as accurate discrimination between the trimethyl and acetyl group (42.01 Da).

Figure 2.

P. falciparum histone PTM landscape in gametocyte. 90 histone PTMs have been identified in P. falciparum gametocyte. Five types of histone PTMs were identified including lysine acetylation (blue ●○), lysine methylation (mono-light green ■□, di-green ■□, and tri-dark green ■□), arginine methylation (mono-light green ■□, di-green ■□), phosphorylation (violet ⧫), and ubiquitination (red ▲). The filled and empty symbols indicate the identified PTMs in this study and unidentified PTMs discovered in an earlier study, respectively. Amino acid residues in red, black, and blue indicate the N-terminal tail, core domain, and C-terminal tail of the histone, respectively.

Table 1.

Overview of Histone PTMs Identified in P. falciparum Gametocytes^a

Histone	PTMs	PTM number	Novel PTM number
H2A	N-term ac, K3ac, K5ac, K41me3, R42me, K75me, R88me, K11me, K123ac	9	6
H2A.Z	N-term ac, K11ac, K15ac, K19ac, K25ac, K28ac, K30ac, K35ac	8	0
H2B	N term ac, K3ac, K4me3, K38ac, R48me. K49me3, K64ac, R91me, K108ac, K112me2, K112me3, K112ub	12	12
H2B.Z	N term ac, K3ac, K8ac, K13ac, K14ac, K18ac, R68me, K104me, K104me3, K116ub	10	4
H3	K9ac, K9me, K9me3, K14ac, K14me, K14me2, R17me, K18ac, K18me, K18me2, K18me3, K23ac, R26me, K27ac, K36me, K36me2, K37me, K37me2, K37me3, R40me, R42me	21	9
H3.3	K9ac, K9me, K9me2, K9me3, K14ac, K14me, K14me2, R17me, K18ac, K18me, K18me2, K18me3, K23ac, R26me, K27ac, K36me, K36me2, K37me, K37me2, R40me, R42me	21	12
H4	N-term ac, R3me, K5ac, K8ac, K12ac, K16ac, K20me3, T30ph, K31ub	9	2
Total		90	45

^aPTMs in bold (45) are novel in gametocytes compared to published histone PTMs in malaria parasites. PTMs in red (13) are newly identified compared to published histone PTMs in model organisms and Toxoplasma. Among them, the italicized PTMs indicate that the methyl or acetyl group is added to six residues which are unique in Apicomplexa while the underlined PTM indicates that a methyl group is added to a residue which is unique in P. falciparum.

Figure 3.

Presentative transitions of peptides by mass spectrometry. Four presentative transitions of H3.3 N-terminal peptides (amino acid 9–17) bearing lysine acetylation and methylation are displayed showing the successful detection of acetylation on K9 and K14, and mono-, di-, and trimethylation on K9.

Methylation and acetylation were the major types of PTMs, while only three ubiquitination sites and one phosphorylation site were identified. Previously, these minor types of modifications were identified after enrichment.^51,52 Compared to histone PTMs identified in earlier studies of both asexual- and sexual-stage parasites,^19–21,23 45 (50%) PTMs at 34 amino acid residues identified in this study were not discovered before and are novel PTMs in P. falciparum (Figure 2, Tables 1, S3). The majority (28/34) of these residues with novel PTMs are conserved in model organisms. In contrast, five residues (H2AK41, H2AK123, H2BK3, H2BK4, and H2BK64) are unique in Apicomplexa, while one (H2BR48) is unique in P. falciparum (Figure 2, Tables 1, S3). Regarding the 45 new PTMs, 32 were also identified in model organisms, whereas 13 were found for the first time in this study, including the six PTMs at those six unique residues (Figure 2, Tables 1, S3). Altogether, this in-depth analysis revealed a distinct repertoire of histone PTMs during gametocyte development.

Extensive Modifications of the Histone H3 and H3.3 N-Terminal Tails

The histone H3 N-terminal tail was most extensively modified with 20 PTMs at 11 sites (Figure 2, Tables 1, S3). Previously identified H3 acetylation at K9, K14, K18, K23, and K27 and methylation at K9me1/3, K14me1, K18me1, and K36me1/2 were also detected in this study. In contrast, previously identified K4me1–3, K9me2, K23me1, K27me1–3, and K36me3 in the N-terminal tail and K56ac, K56me1–3, K79ac, K79me1–3, and K112ac in the core domain were not identified. Notably, several methylation sites, K14me2, K18me2–3, K37me1–3, R26me, R40me, and R42me, were identified in this study for the first time in P. falciparum histones. H3 and H3.3 are 94% identical in P. falciparum and differ only at eight positions.¹⁹ PfH3.3 had similar levels of modifications, reflected in the almost identical PTMs as in H3 except for H3K9me2 and H3K37me3 (Figures 1 and 2, Tables 1, S3).

Unique Modifications in H2A, H2B, Their Variants, and H4

Lysine acetylation was exclusively detected at the N-terminal tails of histone H2A, H2A.Z, and H2B.Z, in addition to H2A.Z K28me1 and a few phosphorylation sites identified recently in P. falciparum gametocytes.²³ Our study, however, revealed many more PTMs in these histones (Figure 2, Tables 1, S3). Besides known acetylation at K3 and K5, we detected six new PTMs (K41me3, R42me, K75me, R88me, K118me, and K123ac) in histone H2A. We also detected all eight previously identified acetylation sites (K11, K15, K19, K25, K28, K30, K35, and K37) but not K28me1 in histone H2A.Z. Twelve PTMs (acetylation and methylation) were detected in H2B, including five types of acetylation (N-term ac, K3ac, K38ac, K64ac, and K108ac), six types of methylation (K4me3, R48me, K49me3, R91me, and K112me2/3), and one ubiquitination (K112ub). In comparison, four novel PTMs (R68me, K104me, K104me3, and K116ub) were detected in H2B.Z. Among the 22 new PTMs identified in H2A, H2B, and their variants, nine (K123ac, K41me3 in H2A, K3ac, K4me3, R48me, K49me3, and K112me3 in H2B, R68me, and K104me3 in H2B.Z) were detected for the first time, while the remaining 13 have been identified in model organisms^20,53–55 and most of these were not functionally studied. Among the 19 amino acid residues modified by these 22 new PTMs, the residue H2BR48 is Plasmodium-specific, while the other five residues (H2AK123, H2AK41, H2BK3, H2BK4, and H2BK64) were unique in apicomplexan.²⁰

We discovered almost all known PTMs in H4 except H4R3me2 and H4K20me1/2²³ (Figure 2, Tables 1, S3). In vitro studies showed that PfMYST catalyzed the acetylation of the four lysine residues in the H4 tail.⁵⁶ These modifications were present in both asexual- and sexual-stage parasites.^19–21,50 In addition, we identified two new PTMs, T30ph and K31ub. Although these two residues are conserved in model organisms and apicomplexan parasites, no modification was identified in T30, while only K31 acetylation and methylation were found in the other organisms.^53,57

Dynamic Histones PTMs during P. falciparum Gametocyte Development

In the course of gametocyte development in P. falciparum, different patterns of histone PTMs were observed (Table 2). Acetylation of lysine residues in canonical and variant histones except H2B was consistently identified from early-to late-stage gametocytes. In histones H3, H3.3, and H2B.Z, methylation sites and types increased during gametocyte development (Table 2). To gain more insight into the global histone PTM patterns, we quantified the levels of the 90 PTMs identified in this study using a label-free quantification method (Table S4). Among them, the levels of 58 PTMs were successfully calculated in all three time points (Figure 4, Table S4). The similarity of PTMs during the gametocyte development was measured by Spearman correlation analysis (Figure 4A). The PTM landscape in early gametocytes is more similar to that in the middle stage (0.78) than in the late stage (0.53), indicating a gradual change in PTM levels during gametocyte development.

Table 2.

Stage Specific PTMs Identified in Gametocyte^a

Histone	Gametocyte stages
	Day 4	Day 8	Day 12
H2A	N-term ac, K3ac, K5ac	N-term ac, K3ac, K5ac, K75me, K119me	N-term ac, K3ac, K5ac, K41me3, R42me, K75me, R88me, K123ac
H2A.Z	K11ac, K15ac, K19ac, K25ac, K28ac, K30ac, K35ac	N-term ac, K11ac, K15ac, K19ac, K25ac, K28ac, K30ac, K35ac	N-term ac, K11ac, K15ac, K19ac, K25ac, K28ac, K30ac, K35ac
H2B	N-term ac, K3ac, K38ac, K64ac, K112ub	K38ac, K112ub	K3ac, K4me3, R48me, K49me3, R91me, K108ac, K112me2, K112me3, K112ub
H2B.Z	N-term ac, K3ac, K8ac, K14ac, K18ac, K116ub	N-term ac, K3ac, K8ac, K14ac, K18ac	N-term ac, K3ac, K8ac, K13ac, K14ac, K18ac R68me, K104me, K104me3
H3	K9ac, K14ac, K18ac, K18me, K18me2, K18me3, K23ac, K27ac, K36me, K37me	K9ac, K14ac, K14me2, K18ac, K18me, K18me2, K18me3, K23ac, K27ac, K36me, K36me2, K37me, K37me2, K37me3	K9ac, K9me, K9me3, K14ac, K14me, R17me, K18ac, K18me2, K18me3, K23ac, R26me, K27ac, K36me, K36me2, K37me, K37me2, K37me3, R40me,_R42me
H3.3	K9ac, K14ac, K18ac, K18me, K18me2, K18me3, K23ac, K27ac, K37me	K9ac, K9me, K9me2, K9me3, K14ac, K18ac, K18me, K18me2, K18me3, K23ac, K27ac, K36me, K37me	K9ac, K9me1, K9me2, K9me3, K14ac, K14m K14me2, R17me, K18ac, K18me, K18me2, K18me3, K23ac, R26me, K27ac, K36me, K36me2, K37me, K37me2, R40me, R42me
H4	K5ac, K8ac, K12ac, K16ac, K20me3	K5ac, K8ac, K12ac, K16ac, K20me3, T30ph, K31ub	N-term ac, R3me, K5ac, K8ac, K12ac, K16ac, K20me3

^aPTMs in bold are novel in gametocytes compared to published histone PTMs in malaria parasites. PTMs in red are newly identified compared to published histone PTMs in model organisms and Toxoplasma. Among them, the italicized PTMs indicate that the methyl or acetyl group is added to the residues which are unique in Apicomplexa, while the underlined PTM indicates that a methyl group is added to a residue which is unique in P. falciparum.

Figure 4.

Dynamic landscape of PTM levels during gametocyte development. (A) Correlation of the overall histone PTM landscape between the three P. falciparum gametocyte developmental stages (Spearman correlation). (B) Histone PTM relative abundance (0–100%) over the gametocyte development was clustered hierarchically. Two major clusters and six sub-clusters are labeled by numbers. (C) Validation of histone acetylation in H4 and methylation in H3 by western blotting with specific antibodies. Anti-H3 antibodies were used for equal histone loading control.

Hierarchical clustering of the histone PTM profiles revealed two major clusters and at least six subclusters exhibiting stagespecific patterns during gametocyte development (Figure 4B). Cluster 1 contains 9 PTMs at medium to high abundance (from ~30–50 to 70–100%), including 6 acetylated lysine residues, 2 methylated lysine residues, and 1 ubiquitination. This major cluster is divided into two subclusters. Subcluster 1–1 includes 5 lysine residues with relatively consistent acetylation at high abundance (60–100%) during gametocyte development. Subcluster 1–2 contains 4 abundant PTMs (H3/H3.3K37me1, H4K5ac, and H2B K112ub) with a decrease at the late gametocyte stage, especially for H2BK112ub, which showed the most substantial reduction in abundance in day 12 gametocytes.

The remaining 49 PTMs detected from medium to low or very low levels (from ~30–50 to 10–30% or below 10%) are grouped in Cluster 2, which is further divided into four subclusters (Figure 4B). Subcluster 2–1 consists of five acetylated lysine residues (H3/K3.3K27ac, H3.3K9ac, H4K5ac, and H4K16ac), most of which are at relatively consistent levels during development except H3.3K9ac showing a sharp reduction at the middle stage. Subcluster 2–2, including 2 acetylated lysine residues (K3/H3.3K18ac) and 6 methylated residues (K9me1/2, K18me3, H3K36me1 in H3.3, and R17me, R40me in H3), showed an increase in abundance from ~0–20 to 15–38% during gametocyte development. Conversely, PTM levels in Subcluster 2–3 harboring H3K18me1/2, H3.3K18me, H3K36me, H3K38ac, and H2B.ZK116ub decreased from 10–30 to 0–10%. Lastly, subcluster 2–4 includes 30 PTMs, and most of them (26/30) are methylated residues, sporadically presenting at very low levels (below 10%) in the three gametocyte stages.

To validate our quantitative PTM analysis, we evaluated the abundance of eight histone marks during development by western blot (WB) analysis (Figure 4B, Figure S1, Table S4). These selected marks include two known PTMs (H4acs and H3K18me1), five novel PTMs (H3K14me2, H3K18me2, H3R26me1, H3K37me1, and H3K37me2), and H3K27me1 which was not detected in this study but was identified at high abundance in the late-stage gametocytes by an earlier study.²³ The anti-acetyl H4 antibodies against H4acs (K5 + K8 + K12 + K16) detected high levels of H4 acetylation in asexual parasites and throughout gametocyte development, while H3K18me1 was found at a relatively high level at the middle-stage gametocyte and the asexual parasites. These results were consistent with our quantitative PTM data and the asexual proteomic data from a previous study.²³ H3K14me2 was detected at a low level in both middle- and late-stage gametocytes, inconsistent with our quantitative PTM data showing that H3K14me2 was at low abundance in the middle-stage gametocytes. However, our quantitative PTM data indicated that H3.3K14me2 was at low abundance in the late-stage gametocytes. Given that H3 and H3.3 are highly conserved in amino acid sequence, anti-H3K14me2 antibodies could also detect H3.3K14me2 in WB. Furthermore, H3K18me2 and H3R26me1 were detected at relatively low levels in the early- and late-stage gametocytes, respectively, and H3K37me1 was low in asexual parasites but was highly abundant in gametocytes, with a significant reduction at the late gametocyte stage, contrasting H3K37me2 at significantly lower levels in both asexual- and sexual-stage parasites. These WBs were congruent with our quantitative H3 and H3.3 PTM data. Additionally, consistent with our proteomic data, H3K27me1 was not detected by WB. Taken together, our quantitative PTM analysis revealed that gametocyte histones harbor distinct, highly dynamic PTMs.

Unique Combinatorial Associations between Histone Modifications

To investigate potential cross-regulation between histone modifications, we analyzed the probabilities of co-occurrence of PTMs on the neighboring residues using established methods (Table S5). The co-occurrence of acetylation was common on gametocyte histones. There were high co-occurrence probabilities between acetylation on K3–K5 of H2A, K11–K15, and K25–K28–K30–K35 (tetra-acetylation) in H2A.Z, K3–K8 and K14–K18 in H2B.Z, K9–K14 and K18–K23 in H3/H3.3, and K5–K8–K12–K16 (tetra-acetylation) in H4. The co-occurrences of lysine and arginine methylation were identified on H3 (K9me3-K14me and K37me3-R40me), H3.3 (K14me1/2-R17me), and H3/H3.3 (K36me1/2-K37me1/2, R17me-K18me2, and K36me-R40me). While the co-occurrences of the same type of PTMs suggest synergistic or cooperative interactions between active or repressive PTMs, several marks of acetylation and methylation were also found to co-occur with high probabilities, including K14ac-K18me1/2 and K9ac-K18me1/2 in H3/H3.3 and K14ac-K9me1/2/3 in H3.3.

DISCUSSION

We studied the repertoire and dynamics of histone PTMs during P. falciparum gametocyte development by qualitative and quantitative MS analyses. Since only a small fraction of the asexual-stage parasites commit gametocytogenesis, isolating histones from gametocytes is challenging. Especially, late-stage gametocytes become lighter and may be lost during changes of the culture medium. To overcome this problem, we used large gametocyte cultures. Additionally, in agreement with an early publication,²³ H2A was at a lower abundance at the late gametocyte stage (Figure 1A). To obtain a high coverage of H2A, we performed three more replicates of histone isolation for day 12 gametocytes. This allowed us to obtain >80% sequence coverage for all histones except CenH3, which was either undetected or at an extremely low level in the previous studies.^20,23 This study identified 90 histone PTMs, of which 45 are novel PTMs not identified previously in P. falciparum. Of the new PTMs, 32 were reported in model organisms, whereas 13 were found for the first time. Quantitative analysis revealed highly abundant acetylation and a wide range of methylation, with many undergoing developmental changes, stressing the importance of epigenetic regulation during gametocyte development. The detection of many co-occurrences of PTMs on histone peptides suggests crosstalk between active or repressive PTMs in gametocytes.

This study identified extensive modifications in H3 and H3.3, with almost the same PTMs (21) being present on the H3 and H3.3 tails (Figure 2, Table 1). Besides several acetyl-lysine residues, most (76%, 16/21) are methyl-lysines and methylarginines, including several novel H3/H3.3 methylation sites (K14me2, K18me1–3, K37me1–3, R26me, R40me, and R42me). These PTMs were missed in an earlier publication probably due to low sequence coverage²³ but recently confirmed from a quantitative middle-down proteomic analysis of the N-terminal tails (amino acid 1–49) of H3 and H3.3.⁵⁸ Another intriguing discrepancy is H3K27me1–3, which were identified earlier as abundant histone marks in gametocytes,²³ but only H3K27me1 was found at a very low level.⁵⁸ Similarly, the H3/H3.3 K79me1–3, considered at high levels in the earlier study,²³ was not detected in several earlier studies.^19–21,59 Our extensive analysis of H3/H3.3 PTMs failed to identify the H3K27 and H3K79 methylation, more consistent with those earlier studies. H3K27 and H3K79 methylation in other organisms is catalyzed by the E(z) subfamily members of SET proteins and Dot1, respectively. Given the discrepant findings discussed above and the lack of the homologs of these methylases in P. falciparum, we argue that P. falciparum does not have the H3K27 and H3K79 methyl marks, or they do not play major roles in Plasmodium epigenetics.^{19–21,23,59,60}

The functional significance of most of the H3/H3.3 PTMs in P. falciparum gametocytes was not known. H3K36me2 and H3K36me3 were recently found to also serve as repression marks at the early gametocyte stage,³⁰ similar to their functions in asexual stages. Some of these methyl marks have been studied in other organisms. For example, H3R40me1 is required for gametogenesis in yeast.⁶¹ Methylation of H3K14 and K18 was recently recognized as repressive marks in mammals and the apicomplexan parasite Theileria, respectively,^62,63 while H3K37 methylation was required for gametogenesis in Schizosaccharomyces pome.⁶⁴ H3K37me1 combined with H3K36me1 may also have a specific function in replication control because it was found at the replication origin.⁶⁵ Intriguingly, H3 K18me1/2 and K37me3 were exclusively found in mouse and human male germ cells (sperm).⁵⁷

Compared to the relative scarcity of PTMs identified in histones H2A and H2B,²³ our analysis identified 22 novel PTMs in H2A, H2B, and their variants during gametocyte development (Figure 2, Table 1). Nine PTMs (K123ac, K41me3 in H2A, K3ac, K4me3, R48me, K49me3 and K112me3 in H2B, R68me and K104me3 in H2B.Z) were reported for the first time, while 13 were also identified in model organisms.^53–55 H2BK38ac, H2BR91me, and H2B.ZK104me3 are conserved in yeast (K49/R102/K111) and human (K46/R99/K108).⁶⁶ Mutational studies in yeast indicated these PTMs were involved in DNA repair and gene silencing.⁶⁷ H2BK112ub/H2B.ZK116ub are conserved in yeast (K123) and human (K120). The ubiquitination of the corresponding site in human cells was involved in transcriptional activation, whereas this PTM in yeast was involved in telomeric silencing.^53,68,69 Interestingly, H2B K112me2 was also found in the mouse sperm, suggesting this modification is specifically involved in sexual development.⁵⁷

Quantitative PTM analyses discovered several highly abundant PTMs, including several acetyl-lysines in H3/H3.3 and H4, K37me1 in H3/H3.3, and H2BK112ub, suggesting that these marks play critical roles in gene regulation (Figure 4). It will be of special interest to determine the role of the novel PTM, H3/H3.3 K37me, in gametocytogenesis. The co-occurrence of H3/H3.3 K36me1/2-K37me1/2 indicates K37me might function as a repression mark. Our quantitative analysis that revealed many PTMs, including some putative repressive marks, was relatively deficient but with upward or downward trends during gametocyte development. Although WBs confirmed four of them (H3K14me2, H3K18me2, H3R26me1, and H3K37me2) were consistent with the quantitative data, the verification of other PTMs is needed given that some might be not discovered by a sufficient depth of coverage in this study. These results indicate that P. falciparum harbors canonical active PTMs and many distinct repressive PTMs to coordinate gene expression during sexual development.

The combination of PTMs is a key aspect of the “histone code” and has been increasingly identified.^3,9,70–73 Our analyses of the co-occurrence of PTMs in gametocytes revealed many types of conserved crosstalk between acetylated and methylated residues, and also some unique combinations, especially between acetylated and methylated residues. Co-occurrence of methylation of H3K9 with acetylated PTMs and the combination of H3K36me and H3K37me were also identified in mammalian cells.^65,72,73 The combination of H3.3 K14ac and K9me1/2/3 suggests that the K9 methylation may be conferred by the methyltransferase PfSET7 since it is preferred to methylate H3K9 in the presence of H3K14ac.⁷⁴ Future in-depth studies are needed to dissect the interactions between active and silent PTMs in balancing gene activation and repression during gametocyte development.

CONCLUSIONS

In this study, a total of 90 PTMs on both canonical and variant histones were identified in the gametocyte at three developmental stages. Notably, half of these PTMs were found for the first time in the malaria parasite and 13 PTMs were identified for the first time compared to all known PTMs in model organisms. Quantitative PTM profiles revealed that several histone acetyl-lysine residues in H3/H3.3 and H4, mono-methylation of H3K37, and ubiquitination of H2BK112 were abundant, while many other residues were methylated at relatively lower levels during gametocyte development. Additionally, many types of combinations of PTMs were also identified. Taken together, these analyses revealed a distinct repertoire of histone PTMs in P. falciparum gametocytes, suggesting different epigenetic mechanisms of gene activation and repression in gametocyte development. This study provides a start point for further functional studies of the novel PTMs, hoping to identify vulnerable therapeutic targets for inhibiting gametocytogenesis and disrupting parasite transmission.