Drug inhibition of HDAC3 and epigenetic control of differentiation in Apicomplexa parasites

Abstract

Plasmodium and Toxoplasma are parasites of major medical importance that belong to the Apicomplexa phylum of protozoa. These parasites transform into various stages during their life cycle and express a specific set of proteins at each stage. Although little is yet known of how gene expression is controlled in Apicomplexa, histone modifications, particularly acetylation, are emerging as key regulators of parasite differentiation and stage conversion. We investigated the anti-Apicomplexa effect of FR235222, a histone deacetylase inhibitor (HDACi). We show that FR235222 is active against a variety of Apicomplexa genera, including Plasmodium and Toxoplasma, and is more potent than other HDACi's such as trichostatin A and the clinically relevant compound pyrimethamine. We identify T. gondii HDAC3 (TgHDAC3) as the target of FR235222 in Toxoplasma tachyzoites and demonstrate the crucial role of the conserved and Apicomplexa HDAC-specific residue TgHDAC3 T99 in the inhibitory activity of the drug. We also show that FR235222 induces differentiation of the tachyzoite (replicative) into the bradyzoite (nonreplicative) stage. Additionally, via its anti-TgHDAC3 activity, FR235222 influences the expression of approximately 370 genes, a third of which are stage-specifically expressed. These results identify FR235222 as a potent HDACi of Apicomplexa, and establish HDAC3 as a central regulator of gene expression and stage conversion in Toxoplasma and, likely, other Apicomplexa.

Figure 1.

In vitro antiprotozoal activity of FR235222 and other HDACi's. In vitro inhibitory concentrations for FR235222 and other compounds against Apicomplexan parasite growth were determined by measuring the incorporation of [³H]uracil by intracellular T. gondii tachyzoites and of [³H]thymidine by Plasmodium species, as described in Materials and methods. (A) Effect of FR235222 and other HDACi's on T. gondii RH strain growth in HFF monolayer. Pyrimethamine (non-HDACi compound) was used as a clinically relevant control. Means ± SD of parasite growth (percentages) are shown (n = 3 experiments). (B) Effect of FR235222 on N. caninum; T. gondii types I (RH), II (Pru), and III (CTG) wild-type strains; and the type I FR235222-resistant mutant M190D4. The data are plotted as the percentage of the control in the absence of drug. Means ± SD of parasite growth (percentages) are shown (n = 3 experiments). (C) Effects of FR235222 on P. falciparum clones 3D7 and Dd2. Effectiveness of FR235222 was compared with Chloroquine. EC₅₀ values are plotted. Means ± SD of EC₅₀ (nanomolar) are shown (n = 4 experiments for each set of data). (D) Effect of FR235222 on P. berghei blood stage development. Intraerythrocyte P. berghei GFP@hsp70 ANKA parasites were synchronized as described in Materials and methods. Synchronized in vitro cultures were treated with 54 nM FR235222 after 5, 11, and 22 h of intracellular growth. The development of the parasites was assessed at the indicated time points by Giemsa staining. Bars, 8 µm. (E) Effects of FR235222 on histone H4 acetylation in intracellular T. gondii parasites. Confluent monolayers of HFF cells were infected with T. gondii RH WT and R20D9 (TgHDAC3^T99A) strains in the presence of 40 nM FR235222 and 0.1% DMSO as a control. As a control, parasites were treated with 1 µM pyrimethamine. After 24 h of growth, cells were fixed and stained for AcH4 (red) and IMC1 (green). The arrowhead indicates aberrant progeny. A representative set of data is shown. Bars, 5 µm. (F) Extracellular T. gondii parasites (RH WT, R20D9 [TgHDAC3^T99A], and M3135E11 [TgHDAC3^T99I]) were treated with the indicated concentrations of FR235222 for 4 h and lysed. Total cell lysates were analyzed by immunoblot with anti-AcH4, anti-H4, anti-TgHDAC3, and anti–α-tubulin antibodies as indicated.

Figure 2.

FR235222 induces acetylation of histones in T. gondii. (A) Sequence alignment of HDAC3 homologues in Apicomplexan parasites and other organisms. The region (from amino acids 122–141 of TgHDAC3) surrounding the point mutation identified in TgHDAC3-resistant mutants is shown. Rpd3 and HDLP are the HDAC homologues in Saccharomyces cerevisiae and the hyperthermophilic bacterium Aquifex aeolicus, respectively. Point mutations identified in the T. gondii FR235222-resistant mutants are shown at the bottom. The multiple sequence alignment was generated by the Clustal alignment method using the BLOSUM scoring matrix (CLC Free Workbench, available at http://www.clcbio.com). Regions identical to the three proteins (light blue) and amino acids conserved exclusively in Apicomplexan parasites (red) are shown. Cp, Cryptosporidium parvum; Hs, Homo sapiens; Nc, N. caninum; Pb, P. berghei; Pf, P. falciparum; Tg, T. gondii. (B) Schematic representation of the exon (yellow boxes)–intron (dashed lines) structure of TgHDAC3. DNA fragments encompassing the point mutations T99A and T99I were PCR amplified using the primers indicated (arrows), and genomic DNA from the resistant mutants M190D4 and M3135C3 as DNA template, respectively. T. gondii RH strain was transformed with the resulting PCR-DNA fragments, and FR235222 was used to select transformants harboring the allelic replacement of TgHDAC3 on the T. gondii genome. The mutated residues T99A and T99I (red) are shown. The corresponding chromatograms are shown at the bottom. (C) Comparison of T. gondii WT and FR235222-resistant mutant line growth in the presence of different concentrations of FR235222. T. gondii RH WT strain resistant lines from the mutagenesis screen carrying the HDAC3^T99A (M190D4) and the HDAC3^T99I (M3135C3) mutations, and the reconstructed mutants by allelic replacement for HDAC3^T99A (R20D9) and HDAC3^T99I (R01E11) were grown in HFF monolayer until numerous parasitophorous vacuoles were observed with the resistant parasites (∼6 d). Means ± SD of cpm are shown (n = 2 experiments). (D) FR235222 is a direct inhibitor of TgHDAC3. Partially purified TgHDAC3-HA-FLAG (tag in the C-terminal position) from T. gondii parasites was assayed for HDAC activity in vitro as described in Materials and methods. Equivalent amounts of immunoprecipitated protein were incubated with [³H]-labeled acetylated histones to determine enzymatic activity. Enzymatic activity is shown as a percentage of HDAC activity in the absence of inhibitor (set at 100%). Means ± SD are shown (n = 2 experiments).

Figure 3.

Genome-wide analysis of the effect of FR235222 treatment on histone H4 acetylation in T. gondii tachyzoites. (A) Statistical analyses of the AcH4 ChIP signal repartition across the T. gondii genome (see Materials and methods). The acetylated mark is more frequently enriched in both proximal and intergenic regions than within an internal gene region. (B) Diagram of the distribution of 369 FR235222 target genes (the up group) over Toxoplasma life-cycle stages, according to EST expression (tachyzoite, bradyzoite, and sporozoite). Numbers represent the percentage of genes that are either shared between two or three stages, or that are detected in a single stage. (C and D) The graphs indicate the fold enrichment of AcH4 over several FR235222 target genes. The signal from multiple closely spaced probes will often form a peak at or near the binding site. For each peak, the signals (selected cutoff = 1.5) were summed (Σratio [log₂]) over the distance of the peak (1,000 bp upstream the start codon). (E) A representative view of Toxoplasma ChIP-on-chip analysis. High resolution mapping of AcH4 location on chromosome IV (SRS9 locus) after FR235222 treatment versus mock DMSO. Each vertical bar represents the log₂ signal ratio of the test sample signal divided by the input control signal. The x axis denotes the genomic position of each probe. A threshold of 1.5 was applied to each experiment. Green (intron) and blue (exon) boxes indicate ToxoDB 4.3 gene annotation. The data are visualized by Genobrowser software. (F and G) FR235222 treatment induces T. gondii Prugniaud strain differentiation. Intracellular parasites (12h after infection) were incubated in the presence of either alkaline growth media to induce in vitro differentiation or 40 nM FR235222 for 2–3 d. As a control, parasites were treated with 1 µM pyrimethamine. Bradyzoite-differentiated parasites were identified by IFA using an anti-P36 antibody. Percentages of SRS9/P36-positive parasitophorous vacuoles in each condition are shown. Means ± SD of P36⁺ vacuoles are shown (n = 2 experiments). P36 (SRS9) expression was analyzed by IFAs in intracellular tachyzoites (pH 7.2) and in bradyzoite-differentiated parasites in vitro (pH 8.2). Tachyzoites were grown in the presence of 0.1% DMSO or 0.01% ethanol for 24 h (pH 7.2) or for 2–3 d (pH 8.2), 40 nM FR235222 for 3 d, or 1 µM pyrimethamine for 2–3 d. Bars, 10 µm.

Figure 4.

TgHDAC3 is a regulator of T. gondii differentiation. (A) High resolution ChIP-on-chip mapping of AcH4 location on chromosome VIIa (20.m00351 locus) after FR235222 treatment versus mock DMSO, as described in Fig. 3 E. (B) High resolution ChIP-on-chip mapping of AcH4 location on chromosome XII (DHFR locus) after FR235222 treatment versus mock DMSO, as described in Fig. 3 E. (C) FR235222 treatment induced the expression of the gene 20.m00351. mRNA levels of 20.m00351 were analyzed by RT-PCR in intracellular tachyzoites (pH 7.2) and in bradyzoite-differentiated parasites in vitro (pH 8.2). Tachyzoites were grown in the presence of 0.1% DMSO for 24 h (pH 7.2) or for 2–3 d (pH 8.2), or 40 nM FR235222 for 3 d. Lanes 1–4 and lanes 5–8 represent fourfold serial dilutions of the first-strand cDNA from parasites treated with 0.1% DMSO (pH 7.2 and 8.2) and 40 nM FR235222. α-Tubulin mRNA are expressed equally in each set of experiments and are presented as a control (n = 3 experiments). (D) Scanning ChIP experiments showing the effects of FR235222 on AcH4 levels in the presence of the TgHDAC3 WT, TgHDAC3^T99A, and TgHDAC3^T99I alleles at the promoter region of the bradyzoite-specific gene 20.m00351. ChIP analysis and quantification were performed as described in Materials and methods. The ratios of 20.m00351 and control DHFR signals present in input samples were used to calculate the relative precipitated fold enrichment shown below each lane (n = 3 experiments). A representative set of data is shown.