Garcinol Inhibits GCN5-Mediated Lysine Acetyltransferase Activity and Prevents Replication of the Parasite Toxoplasma gondii

Abstract

Lysine acetylation is a critical posttranslational modification that influences protein activity, stability, and binding properties. The acetylation of histone proteins in particular is a well-characterized feature of gene expression regulation. In the protozoan parasite Toxoplasma gondii, a number of lysine acetyltransferases (KATs) contribute to gene expression and are essential for parasite viability. The natural product garcinol was recently reported to inhibit enzymatic activities of GCN5 and p300 family KATs in other species. Here we show that garcinol inhibits TgGCN5b, the only nuclear GCN5 family KAT known to be required for Toxoplasma tachyzoite replication. Treatment of tachyzoites with garcinol led to a reduction of global lysine acetylation, particularly on histone H3 and TgGCN5b itself. We also performed transcriptome sequencing (RNA-seq), which revealed increasing aberrant gene expression coincident with increasing concentrations of garcinol. The majority of the genes that were most significantly affected by garcinol were also associated with TgGCN5b in a previously reported chromatin immunoprecipitation assay with microarray technology (ChIP-chip) analysis. The dysregulated gene expression induced by garcinol significantly inhibits Toxoplasma tachyzoite replication, and the concentrations used exhibit no overt toxicity on human host cells. Garcinol also inhibits Plasmodium falciparum asexual replication with a 50% inhibitory concentration (IC₅₀) similar to that for Toxoplasma. Together, these data support that pharmacological inhibition of TgGCN5b leads to a catastrophic failure in gene expression control that prevents parasite replication.

INTRODUCTION

The intracellular protozoan parasite Toxoplasma gondii is the causative agent of toxoplasmosis, which is estimated to infect one-third of the world's population. Toxoplasmosis is usually asymptomatic, but if an infected individual becomes immunocompromised through chemotherapy, organ transplantation, or AIDS infection, latent parasites residing as tissue cysts can reactivate, leading to acute disease and rapid tissue destruction, organ failure, and death (1). The frontline antifolate treatment currently available to treat toxoplasmosis is inadequate, causing toxic side effects and severe allergic reactions in many patients. A better understanding of essential processes in parasite biology is required to develop much-needed new therapies for toxoplasmosis and other related parasitic diseases.

Along with other groups, we have established that lysine acetylation is an essential posttranslational modification (PTM) in Toxoplasma, particularly as it pertains to histone acetylation. Histone acetylation is an important component of the “histone code” that contributes to the regulation of gene expression (2). This reversible PTM is “written” by lysine acetyltransferases (KATs) and “erased” by lysine deacetylases (KDACs); interference with this process has proven detrimental to parasite survival. The target of the antiparasitic compound apicidin was found to be a parasite KDAC homologue (3). FR23522 and its derivatives also display potent activity against Toxoplasma by targeting a KDAC called TgHDAC3 (4, 5). In contrast, there are no reports to date of a Toxoplasma KAT inhibitor despite genetic studies showing that KATs are essential for parasite viability (6–13). The putative KAT inhibitor MC1626 has inhibitory activity against tachyzoites but most likely through an off-target effect (14). However, curcumin and anacardic acid have been reported to inhibit proliferation of the related parasite Plasmodium falciparum (the causative agent of malaria) through inhibition of its GCN5 KAT (15, 16). Unfortunately, these compounds are notoriously nonspecific, also inhibiting microtubules (17), PfATP6 (18), protein disulfide isomerase (19, 20), and xanthine oxidase (21, 22), and can also lead to the generation of reactive oxygen species (16).

The development of specific KAT inhibitors has progressed more slowly than the development KDAC inhibitors. One of the more recent KAT inhibitors that has been identified is garcinol, a polyisoprenylated benzophenone derivative extracted from the dried rind of kokum fruit (Garcinia indica). This compound has been reported to inhibit the enzymatic activities of two distinct classes of mammalian KATs, the GCN5 family member PCAF and P300, with 50% inhibitory concentrations (IC₅₀s) in the range of 5 to 7 μM (23). Here, in this study, we show that garcinol inhibits TgGCN5b KAT activity and reduces global lysine acetylation in vivo in treated parasites, including its preferred substrate, histone H3. Transcriptomic analysis of garcinol-treated parasites revealed altered expression of mRNAs previously linked to TgGCN5b regulation through genome-wide chromatin immunoprecipitation. These effects of garcinol are sufficient to prevent Toxoplasma tachyzoites from replicating without producing overt damage to their human host cells.

MATERIALS AND METHODS

Reagents and parasite culture.

Human foreskin fibroblasts (HFFs) were grown to confluence in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (Atlanta Biologicals). Type I strain RH and type II strain ME49 were maintained in HFFs in DMEM supplemented with 1% heat-inactivated fetal bovine serum. Cultures were kept in humidified incubators at 37°C with 5% CO₂.

Parasite inhibition assays.

The rate of replication of ME49 tachyzoites was determined by a parasite doubling assay (24). Intracellular tachyzoites growing in HFF monolayers were released and purified from host cells by syringe lysis and filtration through a 3.0-μm filter. The tachyzoites were inoculated onto a fresh HFF monolayer and allowed to invade for 4 h, parasites that did not invade host cells were then removed, and fresh medium with increasing concentrations of garcinol or the vehicle (dimethyl sulfoxide [DMSO]) was added. The infected monolayers were incubated for a further 24 h (28 h postinvasion), when they were fixed in 100% methanol and stained with Diff-Quik stain (Siemens Healthcare). The number of parasites per vacuole was counted for at least 100 vacuoles at each concentration of garcinol tested.

The IC₅₀ of garcinol for Toxoplasma tachyzoite replication was determined by a PCR-based B1 assay as described previously (25). Briefly, a 2-fold dilution series of garcinol was prepared in a 24-well plate with HFF monolayers infected with RH tachyzoites, including a vehicle control (DMSO). The infected monolayers were incubated in the presence of the compound for 32 h, at which point they were harvested in lysis buffer for total genomic DNA extraction using the DNeasy blood and tissue DNA kit (Qiagen). Real-time quantitative PCR (qPCR) was performed on the extracted genomic DNA to quantify the extent of parasite proliferation in each sample by using SYBR green master mix (Life Technologies) and 0.5 μM primers “B1 F” and “B1 R,” designed to anneal to the parasite-specific B1 gene (see Table S1 in the supplemental material). qPCR was performed with the Step One Plus real-time PCR system and StepOne software v 2.3 (Life Technologies).

P. falciparum strains HB3 (Honduras) and Dd2 (Indochina) were maintained in O⁺ human red blood cells (Biochemed, Winchester, VA) and RPMI 1640 medium (Gibco) supplemented with 0.5% AlbuMAX II (Gibco), 0.25% sodium bicarbonate (Corning), and 0.01 mg/ml gentamicin (Gibco) under an atmosphere of 90% nitrogen, 5% O₂, and 5% CO₂. Cultures underwent at least two life cycles prior to the initiation of assays to ensure that normal growth was established. Dose-response curves were generated by using a hypoxanthine incorporation assay (26). Briefly, parasite cultures for assays were maintained as asynchronous cultures and required to reach a parasitemia level of no less than 1%, with 70% of parasites in the early ring stage. Sample parasitemia and hematocrit values were then adjusted to 0.2% and 2%, respectively, and the samples were added to test plates containing 2-fold dilutions of garcinol. Parasites were exposed to drug dilutions for 48 h. [³H]hypoxanthine (PerkinElmer) was added to the plates, and the plates were incubated for an additional 20 h before freezing at −80°C for >24 h. Plate contents were harvested, and the number of parasites was counted on a Trilux beta counter. The dose-response curves show means for three independent biological replicates.

Immunoprecipitation and Western blotting.

Intracellular tachyzoites were treated with 4 μM garcinol (catalogue number BML-GR343; Enzo Life Sciences), 1 μM pyrimethamine (catalogue number 46706; Sigma), or the DMSO vehicle (catalogue number D4540; Sigma) for 12 h and then purified from host cells as described above. Purified tachyzoites were lysed in lysis buffer (150 mM NaCl, 50 mM Tris-Cl [pH 7.5], 0.2% NP-40) and sonicated. The amount of parasite protein in the lysate was quantified by using the Bio-Rad DC assay (Bio-Rad). Protein lysates were separated by SDS-PAGE and transferred onto nitrocellulose before blocking in 5% milk–Tris-buffered saline–Tween (TBST), followed by incubation of the following primary antibodies diluted in 5% milk–TBST: anti-acetyl-lysine (AcK) (catalogue number 9441; Cell Signaling) at a 1:1,000 dilution, anti-acetyl H3 (AcH3) (catalogue number 39139; Active Motif) at 1:2,000, anti-histone H3 lysine 9 acetylation (H3K9Ac) (catalogue number 9649; Cell Signaling) at 1:2,000, anti-H3 (catalogue number 39763; Active Motif) at 1:2,000, and anti-P. falciparum aldolase-horseradish peroxidase (HRP) (catalogue number ab38905; Abcam) at 1:2,000. Secondary antibodies conjugated to horseradish peroxidase, including anti-mouse (catalogue number NA931; GE Life Sciences) at a 1:5,000 dilution, anti-rabbit (catalogue number NA934; GE Life Sciences) at 1:2,000, and anti-rat (catalogue number NA935; GE Life Sciences) at 1:2,000, were diluted in 5% milk–TBST. Blots were imaged with a ProteinSimple detection system. Blots were stripped by using Thermo Scientific Restore stripping buffer and reprobed for loading controls.

TgGCN5b was immunoprecipitated from parasites overexpressing an ectopic copy of TgGCN5b with an N-terminal hemagglutinin (HA) tag (^HATgGCN5b) (12) by using an anti-HA affinity matrix (catalogue number 11815016001; Roche). Lysates were mixed with the affinity matrix and rocked at 4°C overnight. The matrix was then washed three times in lysis buffer, and bound protein was eluted from the matrix by boiling for 10 min in 1× Novex SDS sample buffer (Life Technologies) with 5% beta-mercaptoethanol. Eluted proteins were separated by 4 to 12% Bis-Tris SDS-PAGE and probed by Western blotting as described above.

In vitro lysine acetyltransferase assays.

Immunoprecipitated ^HATgGCN5b (see above) or recombinant human P300 (catalogue number 14-418; Millipore) was incubated with a recombinant human histone H3.1 substrate (catalogue number M2503S; New England BioLabs) and acetyl coenzyme A (acetyl-CoA) (catalogue number A2056; Sigma) in KAT buffer (50 mM Tris-Cl [pH 8.0], 5% glycerol, 0.1 mM EDTA, 50 mM KCl, 10 mM sodium butyrate, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride) for 1 h at 37°C. KAT reaction mixtures were then resolved by 4 to 12% Bis-Tris SDS-PAGE (Life Technologies). Quantification of KAT activity was determined by immunoblotting the KAT assay reaction mixture with anti-acetyl H3 antibody (catalogue number 39139; Active Motif) at a 1:2,000 dilution.

RNA sequencing analysis and qRT-PCR.

Intracellular tachyzoites in T-150 flasks were treated with the indicated concentrations of garcinol or the vehicle for 8 h. Tachyzoites were purified from host cells as described above and then washed in phosphate-buffered saline (PBS). Total RNA was extracted in TRIzol (Life Technologies), genomic DNA was degraded by using the DNA-free DNA removal kit (Ambion), and any remaining contaminants in the RNA were removed by using the Qiagen RNeasy kit. RNA was collected from three independent replicates for each concentration of garcinol tested. RNA quality was verified by using a BioAnalyzer. Sequencing libraries were generated by using the Kapa Stranded mRNA-Seq kit (catalogue number KR0960; Kapa Biosystems). RNA samples were subjected to poly(A) purification using magnetic oligo(dT) beads and then fragmented to 200- to 500-nucleotide (nt) fragments in fragmentation buffer. First-strand cDNA synthesis was performed by using random hexamer primers followed by second-strand cDNA synthesis carried out with dUTP replacing dTTP. The double-stranded cDNA was purified with a QIAquick PCR extraction kit (Qiagen), ligated to sequencing adaptors, and purified by gel electrophoresis, followed by fragment enrichment by PCR amplification using primers recognizing the sequencing adaptors. The library was subjected to 100-bp paired-end sequencing with the Illumina HiSeq 2000 system. The reads were aligned to the ME49 reference genome sequence from ToxoDB (version 13.0) with Tophat-2.0.13 (27). Only reads that mapped to a single location in the genome and with ≤2 mismatches were used for further analysis. Read counts were calculated by using bamutils from NGSUtils (28). Differential gene expression analysis was performed with edgeR (29). Genes were categorized as significantly changed with a false-discovery rate (FDR) of <0.5, a P value of <0.05, and a fold change of ±1.5 relative to the vehicle-treated control.

To validate the fold changes of select genes, 1.0 μg of total RNA purified from intracellular parasites was transcribed into cDNA by using Superscript III with oligo(dT) primers according to the manufacturer's protocol (Life Technologies). Reverse transcription-qPCR (qRT-PCR) was performed with SYBR green PCR master mix (Life Technologies), 0.5 μM each forward and reverse primer (see Table S1 in the supplemental material), and a 1:10 dilution of cDNA. Target genes were amplified by using the Step One Plus real-time PCR system and analyzed with relative quantification software (StepOne software v 2.3; Life Technologies).

RESULTS

Garcinol inhibits TgGCN5b in an in vitro KAT assay.

Garcinol is reported to inhibit the mammalian KATs PCAF (a GCN5 family member) and P300 with IC₅₀s of 5 to 7 μM (23). P300 KATs are not present in apicomplexan parasites, but Toxoplasma has two homologues of PCAF, TgGCN5a, which is not essential in tachyzoites, and TgGCN5b, which is essential (8, 12, 30). To evaluate the inhibitory activity of garcinol against the essential KAT TgGCN5b, we purified TgGCN5b from Toxoplasma parasites engineered to ectopically express an HA-tagged version of the protein (31). Following immunoprecipitation with anti-HA antibody, ^HATgGCN5b was incubated in a standard in vitro KAT assay mixture with acetyl-CoA, a histone H3 substrate, and increasing concentrations of garcinol. ^HATgGCN5b KAT activity was quantified by Western blotting using an anti-acetyl H3 (AcH3) antibody. Recombinant P300 was run in parallel assays as a control for KAT activity and garcinol inhibition. As seen for the recombinant human P300 control, garcinol inhibited the enzymatic activity of TgGCN5b in a dose-dependent manner (Fig. 1).

FIG 1.

Garcinol inhibits lysine acetyltransferase activity of TgGCN5b in vitro. (A) Immunoprecipitated ^HATgGCN5b was used in an in vitro KAT assay with a histone H3 substrate. Reaction mixtures were incubated with the indicated concentrations of garcinol or the vehicle (DMSO). KAT activity was monitored by Western blotting using an antibody that recognizes acetylated H3 (AcH3). Equal loading of both the substrate H3 and ^HATgGCN5b was confirmed with anti-H3 and anti-HA antibodies, respectively. (B) An in vitro KAT assay using recombinant human P300 and histone H3 was performed as a control for garcinol inhibitory activity.

Garcinol reduces lysine acetylation levels in Toxoplasma.

To determine if garcinol affects KAT activity in vivo, we examined total lysine acetylation in intracellular tachyzoites treated with 4 μM garcinol versus the vehicle (DMSO) or an unrelated antiparasitic drug (1 μM pyrimethamine). Total lysine acetylation was detected in whole-protein lysates by Western blotting using a panspecific antibody that recognizes acetylated lysine residues. As shown in Fig. 2A, there was a marked reduction in the acetylation level of numerous proteins from parasites exposed to garcinol. In contrast, lysine acetylation levels remained unchanged in parasites exposed to pyrimethamine, suggesting that the reduction in lysine acetylation is specific to garcinol and not a general response to antiparasitic drugs.

FIG 2.

Garcinol treatment reduces lysine acetylation in vivo. Shown is Western blotting of lysates from purified, intracellular tachyzoites following 12 h of treatment with 5 μM garcinol (GAR), 1 μM pyrimethamine (PYR), or the vehicle (DMSO). Blots were analyzed for total lysine acetylation (AcK) (A), total histone H3 acetylation (H3Ac), and acetylation of histone H3 lysine 9 (H3K9Ac) (B). T. gondii aldolase (TgAldolase) and histone H3 (H3) were used as loading controls. The arrow indicates a band corresponding to histone H3. *, P < 0.05. Error bars represent standard deviations of data from three independent experiments. MW, molecular weight (in thousands).

One of the proteins on the pan-acetyl-lysine blot that was reduced by garcinol treatment corresponds to histone H3 (Fig. 2A, arrow), a preferred substrate for GCN5 family KATs, including TgGCN5b (8, 10, 12). To examine acetyl H3 levels more closely, lysates from parasites treated with garcinol or controls were immunoblotted with antibodies specific to total acetylated H3 or H3 lysine 9 (H3K9), a residue of H3 known to be acetylated by TgGCN5b (10, 12) (Fig. 2B and C). Results show that levels of acetylated H3, and specifically acetylated H3K9, are sharply reduced (75% and 86%, respectively) in parasites treated with garcinol.

Garcinol reduces autoacetylation of TgGCN5b.

Proteome-wide acetylome analyses of Toxoplasma tachyzoites identified seven acetylated lysines on TgGCN5b (K811, K857, K941, K989, K997, K1002, and K1027) (32, 33; V. Jeffers and W. J. Sullivan, Jr., unpublished observations). These acetylated lysines are likely due to autoacetylation and may be involved in regulating the enzymatic activity, stability, or protein-protein interactions of TgGCN5b. If TgGCN5b is capable of autoacetylation, then garcinol treatment would be expected to diminish acetylation levels on TgGCN5b. To test this possibility, ^HATgGCN5b was immunoprecipitated from ^HATgGCN5b-expressing parasites following treatment with garcinol or the vehicle, and the extent of acetylation on the purified protein was assessed by Western blotting using a panspecific AcK antibody (Fig. 3). Garcinol treatment resulted in a 65% reduction in the acetylation level of ^HATgGCN5b, most likely as a consequence of inhibition of TgGCN5b autoacetyltransferase activity.

FIG 3.

Garcinol reduces levels of lysine acetylation on TgGCN5b. ^HATgGCN5b was immunoprecipitated from ^HATgGCN5b-expressing parasites after 12 h of treatment with either 5 μM garcinol (GAR) or the vehicle (DMSO). Acetylation levels of TgGCN5b were detected by using a pan-acetyl-lysine antibody. Levels of total TgGCN5b protein were determined with an anti-HA antibody. *, P < 0.05. Error bars represent standard deviations of data from three independent experiments.

Garcinol-mediated changes in the parasite transcriptome.

We previously reported that induction of a dominant negative version of TgGCN5b caused gene dysregulation that arrested replication (12). Given the inhibitory effects of garcinol on TgGCN5b-mediated histone acetylation, we hypothesized that there would be similar disruptions in gene expression patterns in garcinol-treated parasites. We therefore performed transcriptome sequencing (RNA-seq) and gene expression analysis on ME49 parasites that were treated with 0.5 μM, 1 μM, or 2 μM garcinol or the vehicle for 8 h. Approximately 42 million 100-bp paired-end reads were obtained for each replicate, with an average of 75% of the reads mapping to the Toxoplasma ME49 genome (ToxoDB). A dose-dependent effect on parasite gene expression was observed compared to vehicle treatment (see Data Set S1 in the supplemental material). Treatment with 0.5 μM garcinol had no detectable effect on parasite gene expression; however, treatment with 1 μM and 2 μM garcinol resulted in increasing levels of anomalous gene expression relative to that with the vehicle (Table 1). Parasites treated with 1 μM garcinol displayed altered expression of 29 genes (11 upregulated and 18 downregulated) (Table 2), and treatment with 2 μM garcinol led to aberrant expression of 398 genes (see Data Set S1 in the supplemental material). A subset of the differentially expressed genes was validated by independent qRT-PCR, confirming the fidelity of RNA-seq (see Table S2 in the supplemental material).

TABLE 1.

Genes modulated by garcinol treatment

Garcinol concn (μM)	No. of genes upregulated (+1.5-fold)	No. of genes downregulated (−1.5-fold)
0.5	0	0
1	11	18
2	173	226

TABLE 2.

Genes significantly modulated by treatment with 1μM garcinol^a

Accession no.	Product description	Fold change in expression
Upregulated
TGME49_313650	Hypothetical protein	1.85
TGME49_297820	Sperm-associated antigen 6, putative	1.82
TGME49_200230	Microneme protein MIC17C	1.81
TGME49_259205	Hypothetical protein	1.75
TGME49_268360	Hypothetical protein	1.75
TGME49_297280	Hypothetical protein	1.72
TGME49_200240	Microneme protein MIC17B	1.61
TGME49_234200	Hypothetical protein	1.57
TGME49_216053	Hypothetical protein	1.56
TGME49_269380	Hypothetical protein	1.53
Downregulated
TGME49_278090	Toxoplasma gondii family A protein	0.67
TGME49_258570	Hypothetical protein	0.67
TGME49_215185	Hypothetical protein	0.67
TGME49_262715	Hypothetical protein	0.67
TGME49_313000	Toxoplasma gondii family D protein	0.66
TGME49_313660	Hypothetical protein	0.66
TGME49_301170	Surface antigen (SAG)-related sequence SRS19D	0.65
TGME49_299995	Hypothetical protein	0.64
TGME49_224770	SAG-related sequence SRS40D	0.63
TGME49_309300	SAG-related sequence SRS55A	0.62
TGME49_305530	Hypothetical protein	0.62
TGME49_203375	CPW-WPC domain-containing protein	0.59
TGME49_244715	Hypothetical protein	0.58
TGME49_261560	Tat binding protein 1-interacting protein TBPIP	0.56
TGME49_231610	ATP:guanido phosphotransferase, C-terminal catalytic domain-containing protein	0.56
TGME49_310480	Flagellar/basal body protein, PARK2 coregulated (PACRG) family protein	0.55
TGME49_240620	Hypothetical protein	0.54
TGME49_310275	Hypothetical protein	0.48

^aSignificance was defined as an FDR of <0.05 and a P value of <0.005.

The genes that are significantly altered in response to garcinol do not belong to a single cellular pathway; rather, they comprise a wide variety of cellular functions. These findings paralleled the results of a previously reported chromatin immunoprecipitation with microarray technology (ChIP-chip) analysis of TgGCN5b in tachyzoites, which found TgGCN5b to be present at 1,090 genes that carry out diverse functions (12). We therefore compared the genes affected by garcinol treatment with the gene loci found to be associated with TgGCN5b in our ChIP-chip analysis (Fig. 4). Strikingly, 100% of the genes that are significantly altered by treatment with 1 μM garcinol were also shown to be associated with TgGCN5b in the ChIP-chip analysis, strongly suggesting that garcinol inhibits TgGCN5b in vivo. However, only 10% of the genes (41 out of 398) altered by treatment with 2 μM garcinol were associated with TgGCN5b, suggesting that a cascade of additional gene dysregulation ensues following the initial wave of TgGCN5b-controlled genes. Collectively, these findings are consistent with the idea that garcinol induces abnormal gene expression through inhibition of TgGCN5b KAT activity.

FIG 4.

Correlation between genes modulated by garcinol treatment and genome localization of TgGCN5b. Genes that are differentially expressed during treatment with 1 μM or 2 μM garcinol were compared to gene loci associated with TgGCN5b as determined in a previously reported ChIP-chip study (12).

Garcinol inhibits replication of Toxoplasma and Plasmodium falciparum in vitro.

We previously established that the TgGCN5b complex is essential in tachyzoites (12), so pharmacological inhibition of TgGCN5b is expected to arrest parasite replication. We therefore evaluated the effect of garcinol on Toxoplasma in vitro using a doubling assay to monitor parasite replication (Fig. 5). Our findings show that as little as 0.5 μM garcinol inhibits Toxoplasma tachyzoite replication, with near-complete growth inhibition achieved at 4 μM, a concentration that shows no obvious adverse effects on human host cells. We also tested the concentrations of garcinol against HFF cells freshly seeded into culture flasks; garcinol-treated HFF cells showed no defects in growth to confluence unless 10 μM garcinol was used (data not shown). The 50% effective concentration (EC₅₀) for garcinol against Toxoplasma tachyzoite intracellular replication was calculated by a B1 assay to be 1.79 μM (95% confidence interval [CI], 1.557 to 2.050 μM) (see Data Set S1 in the supplemental material). We also assessed the effect of garcinol on P. falciparum erythrocytic asexual replication and determined that garcinol was equally effective at inhibition of both chloroquine-sensitive (HB3) and chloroquine-resistant (Dd2) strains, with IC₅₀s of 1.69 μM (95% CI, 1.510 to 1.892 μM) and 2.05 μM (95% CI, 1.848 to 2.266 μM), respectively (Fig. 5B and C).

FIG 5.

Garcinol inhibits asexual replication of apicomplexan parasites. (A) A parasite counting assay was used to determine the effect of garcinol on T. gondii strain ME49 tachyzoite replication. Stacked bars represent the proportions of vacuoles containing 1, 2, 4, 8, or 16 parasites at each concentration of garcinol tested. (B) Growth inhibition curve of T. gondii RH tachyzoites treated with increasing concentrations of garcinol. The number of parasites was quantified by qPCR of the parasite-specific B1 gene. (C) Growth inhibition curves of P. falciparum strains HB3 and Dd2 treated with increasing concentrations of garcinol. Error bars represent standard deviations of data from at least three independent assays.

DISCUSSION

Our study shows that garcinol diminishes global lysine acetylation and dysregulates gene expression in Toxoplasma tachyzoites, resulting in the cessation of parasite proliferation in vitro. Garcinol also inhibits P. falciparum asexual-stage replication with similar inhibitory concentrations that affect Toxoplasma replication.

We show that garcinol inhibits the essential Toxoplasma KAT TgGCN5b in vitro, which is consistent with the decreased H3 acetylation observed in vivo when Toxoplasma is exposed to garcinol. Our findings further indicate that lysine acetylation on TgGCN5b itself can result from an autoacetylation mechanism that is also inhibited by garcinol. The biological role of these acetylated lysines on TgGCN5b is currently unknown, but garcinol may prove to be a useful tool to investigate this PTM.

Given that histone H3 is a well-known, preferred substrate of GCN5 family KATs, it is not surprising to see dramatically reduced acetylation of H3 in garcinol-treated parasites (Fig. 2). RNA-seq analysis of tachyzoites subjected to treatment with 1 μM garcinol revealed that 29 genes were aberrantly expressed, all of which were previously implicated as being under TgGCN5b control based on ChIP-chip studies (12). A concentration of 2 μM caused a much larger array of genes to be differentially expressed, many independent of any known association with TgGCN5b. These data suggest that a catastrophic cascade of misregulated gene expression befalls parasites following initial inhibition of TgGCN5b by garcinol. It is also possible that the higher 2 μM dose of garcinol might produce off-target effects that further contributed to the dysregulation of gene expression. Genes linked to TgGCN5b control may also be influenced by other factors that were not impaired by garcinol; thus, their expression was not significantly altered. Histone acetylation is generally associated with gene activation, so genes that are regulated by TgGCN5b are expected to be downregulated in the presence of a KAT inhibitor. While most of the genes in garcinol-treated tachyzoites are downregulated, a subset of differentially expressed genes is upregulated. This result likely reflects the complexity of gene regulation, which involves dozens of other factors working in concert. It is possible that some genes rely on TgGCN5b to acetylate nonhistone proteins that function in gene repression. Consistent with our findings, garcinol treatment of HeLa cells (100 μM for 24 h) repressed the expression of most genes but upregulated others as well (23).

We previously established that hundreds of proteins in all compartments of the parasite are subject to lysine acetylation (32, 33). Thus, there could be nonhistone substrates affected by the inhibition of TgGCN5b that also contribute to the demise of the garcinol-treated parasites. Probing of global lysine acetylation levels in parasites exposed to garcinol lends support this idea (Fig. 2). Toxoplasma possesses a second GCN5 KAT (TgGCN5a), but since it is not essential for tachyzoite replication and we cannot purify sufficient amounts of intact protein, it was not investigated here (8).

We detected no obvious detrimental effects on human host cells at the concentrations of garcinol used to kill Toxoplasma, but it remains possible that there may be subtle changes at the molecular level that make the HFFs inhospitable to the parasites. To address this possibility, we preincubated host cells with up to 10 μM garcinol for 8 h prior to infecting them with Toxoplasma in fresh medium without garcinol. Tachyzoites were able to invade and replicate normally in these garcinol-treated host cells (data not shown), lending support to the idea that garcinol is directly toxic to the parasites.

We also examined the utility of garcinol in an in vivo model of toxoplasmosis by treating mice infected with the hypervirulent type I RH strain. (Infection of mice was performed based on an approved protocol [10852] from the Institutional Animal Care and Use Committee [IACUC] of the University of Indiana School of Medicine [IUSM]. The IUSM is accredited by the International Association for Assessment and Accreditation of Laboratory Animal Care.) Garcinol (at 10 mg/kg of body weight or 20 mg/kg by intraperitoneal administration) did not confer protection from acute infection (data not shown), possibly because garcinol exhibits high-affinity binding to albumin (34), which limits its distribution to tissues. Considering the inhibitory effects of garcinol on both Toxoplasma and P. falciparum in vitro, the efficacy against parasitic infections in vivo should be revisited when newer derivatives of garcinol with improved bioavailability are developed.

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.