Ex Vivo Activity of Histone Deacetylase Inhibitors against Multidrug-Resistant Clinical Isolates of Plasmodium falciparum and P. vivax

Abstract

Histone acetylation plays an important role in regulating gene transcription and silencing in Plasmodium falciparum. Histone deacetylase (HDAC) inhibitors, particularly those of the hydroxamate class, have been shown to have potent in vitro activity against drug-resistant and -sensitive laboratory strains of P. falciparum, raising their potential as a new class of antimalarial compounds. In the current study, stage-specific ex vivo susceptibility profiles of representative hydroxamate-based HDAC inhibitors suberoylanilide hydroxamic acid (SAHA), 2-ASA-9, and 2-ASA-14 (2-ASA-9 and 2-ASA-14 are 2-aminosuberic acid-based HDAC inhibitors) were assessed in multidrug-resistant clinical isolates of P. falciparum (n = 24) and P. vivax (n = 25) from Papua, Indonesia, using a modified schizont maturation assay. Submicromolar concentrations of SAHA, 2-ASA-9, and 2-ASA-14 inhibited the growth of both P. falciparum (median 50% inhibitory concentrations [IC₅₀s] of 310, 533, and 266 nM) and P. vivax (median IC₅₀s of 170, 503, and 278 nM). Inverse correlation patterns between HDAC inhibitors and chloroquine for P. falciparum and mefloquine for P. vivax indicate species-specific susceptibility profiles for HDAC inhibitors. These HDAC inhibitors were also found to be potent ex vivo against P. vivax schizont maturation, comparable to that in P. falciparum, suggesting that HDAC inhibitors may be promising candidates for antimalarial therapy in geographical locations where both species are endemic. Further studies optimizing the selectivity and in vivo efficacy of HDAC inhibitors in Plasmodium spp. and defining drug interaction with common antimalarial compounds are warranted to investigate the role of HDAC inhibitors in antimalarial therapy.

Malaria is a major public health issue in more than 100 countries with 40% of the world's population at risk of acquiring the disease (21). Early diagnosis, followed by prompt and effective treatment, remains a cornerstone of malaria control programs (47). However, the effectiveness of this strategy is compromised by the emergence and spread of drug-resistant malaria. Most antimalarial drug research has focused on the treatment of Plasmodium falciparum, the dominant species in sub-Saharan Africa. However, outside Africa, P. falciparum and Plasmodium vivax are sympatric, with the latter species often accounting for half of malaria episodes. In Papua, Indonesia, P. vivax has become highly resistant to chloroquine and sulfadoxine-pyrimethamine (36, 41), with evidence now emerging of declining chloroquine efficacy throughout most areas where vivax malaria is endemic (11). Reports from the island of New Guinea, where chloroquine resistance is highly prevalent, have demonstrated an association between vivax malaria and severe disease and death (5, 17, 35, 45). In this and other regions where drug resistance has emerged, new treatment strategies that focus on different plasmodial targets are urgently needed for both P. falciparum and P. vivax.

In eukaryotes, histone deacetylase (HDAC) and histone acetyltransferase (HAT) enzymes control the acetylation of both histones and nonhistone proteins and are thus important for a variety of vital cellular functions, such as transcription, DNA replication and repair, cell signaling, as well as cell cycle regulation and differentiation (49, 50). HDACs have been recognized as therapeutic targets in cancer cells for over a decade, with one hydroxamate-based HDAC inhibitor suberoylanilide hydroxamic acid (SAHA) (Vorinostat), approved for the treatment of persistent or refractory T-cell lymphoma (18). More recently, HDACs have been identified as important regulators of transcription in P. falciparum (7, 14, 16, 22). HDAC inhibitors have been shown to have activity against apicomplexan parasites, raising their potential as a new class of antimalarial compounds (3, 23, 33).

The principal aims of this study were to investigate the ex vivo susceptibility profiles of antimalarial HDAC inhibitors against multidrug-resistant clinical isolates of P. falciparum and P. vivax in Papua, Indonesia, and to examine the stage-specific actions of these compounds. The three HDAC inhibitors compared in this study were the clinically approved antitumor compound SAHA, which has in vitro 50% inhibitory concentrations (IC₅₀s) against P. falciparum laboratory lines of ∼100 to 300 nM (12), and two other structurally related hydroxamate-based compounds with greater in vitro activity against laboratory lines (2-ASA-9 with an IC₅₀ of 15 to 39 nM and 2-ASA-14 with an IC₅₀ of 13 to 33 nM [2-ASA-9 and 2-ASA-14 are 2-aminosuberic acid-based HDAC inhibitors]; Fig. 1). These compounds contain a hydroxamic acid zinc binding group and are currently the most promising class of HDAC inhibitors under investigation for use against malaria.

FIG. 1.

Chemical structures of suberoylanilide hydroxamic acid (SAHA) (WR308364) and 2-aminosuberic acid-based HDAC inhibitors, 2-ASA-9 and 2-ASA-14.

MATERIALS AND METHODS

Compounds.

The synthesis, purification, and characterization of 2-aminosuberic acid-based HDAC inhibitors (2-ASA-9 and 2-ASA-14; Fig. 1) have been previously described (24). Suberoylanilide hydroxamic acid (SAHA) was purchased from Sigma. All compounds were prepared as 10 mM stock solutions in dimethyl sulfoxide (DMSO). Drug plates were then predosed by diluting the compounds in 50% ethanol followed by lyophilization and storage at 4°C.

Field location and sample collection.

Between April and September 2008, Plasmodium isolates were collected from patients attending malaria clinics in Timika, Papua province, Indonesia, a region where multidrug-resistant strains of both P. vivax and P. falciparum are endemic (25, 36, 38). Patients with symptomatic malaria presenting to an outpatient facility were recruited into the study if they presented with single-species infections of either P. falciparum or P. vivax. Parasite counts were determined on Giemsa-stained thick films as the number of parasites per 200 white blood cells (WBC), and peripheral parasitemia was calculated assuming a white cell count of 7,300 μl⁻¹. Only isolates with a parasitemia of between 2,000 μl⁻¹ and 80,000 μl⁻¹ were selected for ex vivo analysis. Since previous studies have highlighted a major difference in the stage-specific activities of chloroquine and amodiaquine, isolates of P. vivax were processed only if the majority of asexual forms were at the ring stage (38). Patients treated with antimalarial drugs in the month prior to sampling were excluded from the study. Venous blood (5 ml) was collected by venipuncture, and after removal of host white blood cells using CF 11 cellulose (37), packed infected red blood cells (IRBC) were used for the ex vivo drug susceptibility assay.

Ex vivo drug susceptibility assay.

Drug susceptibility of P. vivax and P. falciparum isolates was measured using a protocol modified from the WHO microtest as described previously (38, 39). Two hundred microliters of a 2% hematocrit blood medium mixture (BMM), consisting of RPMI 1640 medium plus 10% AB⁺ human serum (P. falciparum) or McCoy's 5A medium plus 20% AB⁺ human serum (P. vivax) was added to each well of predosed drug plates containing 11 serial concentrations (2-fold dilutions) of the following antimalarial drugs (maximum concentration shown in parentheses): chloroquine (2,992 nM), amodiaquine (80 nM), piperaquine (769 nM), mefloquine (338 nM), and artesunate (25 nM) and the three HDAC inhibitors SAHA (2,000 nM), 2-ASA-9 (1,000 nM), and 2-ASA-14 (1,000 nM). A candle jar was used to mature the parasites at 37.5°C for 24 to 56 h. Incubation was stopped when >40% of ring-stage parasites had reached the mature schizont stage in the drug-free control.

Thick blood films made from each well were stained with 5% Giemsa solution for 30 min and examined microscopically. Differential counts of 200 asexual parasites in both the preincubation and test slides were classified into ring stage, mature trophozoites, and schizonts. To ensure optimal maturity, ease of parasite identification, and reduction of parasite classification error between microscopists, only schizonts with at least five well-defined chromatin dots were classified as schizonts at harvest. Free merozoites and gametocytes were not included in the count.

The number of schizonts per 200 asexual-stage parasites was determined for each drug concentration and normalized to that of the control well. The dose-response data were analyzed using nonlinear regression analysis (WinNonLin 4.1; Pharsight Corporation), and the IC₅₀s were derived using an inhibitory sigmoid E_max model. IC₅₀ ex vivo data were used only from predicted curves where the maximum effect attributable to the drug (E_max) and the basal effect, corresponding to the response when the dose of the drug is zero (E₀) were within 15% of 100 or 0, respectively.

Data analysis.

Data analysis was performed using STATA (version 10.1; Stata Corp., College Station, TX) and GraphPad Prism (version 5) software. The Mann-Whitney U test, Wilcoxon signed-rank test, and Spearman rank correlation were used for nonparametric comparisons and correlations. In order to examine the robustness of our observations, we used restricted and stratified analysis to control for three major confounding factors: (i) the percentage of ring stages at the start of the assay, (ii) the duration of the assay, and (iii) the initial parasitemia. Robust regression analysis was performed on log-transformed data.

Ethical approval.

Ethical approval for this study was obtained from the ethics committees of the National Institute of Health Research and Development, Ministry of Health, Indonesia, and the Menzies School of Health Research, Casuarina, Darwin, Australia.

RESULTS

Between April and September 2008, ex vivo susceptibility of the HDAC inhibitors SAHA, 2-ASA-9, and 2-ASA-14 was tested in clinical isolates from 49 patients presenting with single-species infections of either P. vivax (n = 25) or P. falciparum (n = 24). Susceptibility profiles on the same isolates were also tested against chloroquine, amodiaquine, piperaquine, mefloquine, and artesunate. Adequate growth for harvest (i.e., maturation from the ring or trophozoite stage to >40% at the schizont stage) was successfully achieved in 83% (20/24) of P. falciparum isolates and 80% (20/25) of P. vivax isolates. Baseline characteristics of the isolates processed are presented in Table 1, and the geometric mean IC₅₀s for the two species, in comparison with published IC₅₀s for laboratory lines, are shown in Table 2 and Fig. 2. A difference in drug susceptibility in the two species was observed for amodiaquine, which showed a higher median IC₅₀ (9.1 versus 5.0 nM) for P. vivax (P = 0.005), and SAHA, which had a higher median IC₅₀ for P. falciparum (301 versus 170 nM; P < 0.001).

TABLE 1.

Baseline characteristics of isolates for which ex vivo assay was accomplished

Baseline characteristic	Value for species:
	P. falciparum	P. vivax
Total no. of isolates reaching harvest	20	20
Median (range) delay from venipuncture to start of culture (h)	2.4 (1.6-4.9)	2.5 (1.6-3.7)
Median (range) duration of assay (h)	46 (41-50)	48 (43-53)
Geometric mean (95% CI)a parasitemia (no. of asexual parasites/μl)	22,622 (11,900-43,001)	11,675 (7,527-18,110)
Median initial % (range) of parasites at ring stage	100b	94 (54-100)
Mean (95% CI) schizont count at harvest	46 (39-52)	40 (35-44)

^a95% CI, 95% confidence interval.

^bNo range given (all values were 100%).

TABLE 2.

Overall ex vivo drug susceptibility for each drug according to the species tested

Antimalarial	Mean IC50 (nM) for P. falciparum lab linesa		P. falciparum		P. vivax
	CQs	CQr	n (%)b	Median IC50(nM) (range)	n (%)	Median IC50(nM) (range)
Chloroquine	12	240	20 (100)	30.2 (6.4-97.1)	20 (100)	40.4 (8.8-81.4)
Amodiaquine	2	8	20 (100)	5.0 (1.0-28.6)	19 (95)	9.1 (2.8-16.0)
Piperaquine	3	11	20 (100)	10.8 (4.4-46.9)	20 (100)	13.2 (3.4-38.4)
Mefloquine	11	8	19 (95)	6.3 (1.7-21.3)	18 (90)	7.7 (1.5-23.2)
Artesunate	2	1	20 (100)	0.92 (0.31-4.1)	20 (100)	0.74 (0.09-1.42)
SAHA	247	161	19 (95)	301 (120-484)	20 (100)	170 (67-281)
2-ASA-9	15	39	15 (75)c	533 (199-964)	19 (95)	503 (203-766)
2-ASA-14	13	33	18 (90)d	266 (87-704)	20 (100)	278 (121-781)

^aMean IC₅₀s (in vitro growth quantified by [³H]hypoxanthine incorporation) as previously published for chloroquine (8, 9), amodiaquine (1, 19), piperaquine (1, 30), mefloquine (1, 10), artesunate (1, 10) (the chloroquine-sensitive and -resistant lines were 3D7 and K1, respectively), SAHA (12), 2-ASA-9 (2), and 2-ASA-14 (2) (the chloroquine-sensitive and -resistant laboratory lines, respectively, were D6 and W2 for SAHA and 3D7 and Dd2 for 2-ASA-9 and 2-ASA-14). CQ^s, chloroquine-sensitive laboratory strain; CQ^r, chloroquine-resistant laboratory strain.

^bn, total number of assays with acceptable IC₅₀; %, total number of assays with acceptable IC₅₀/number of assays with adequate growth harvested.

^cNo IC₅₀ estimates of 2-ASA-9 for four P. falciparum isolates and one P. vivax isolate (MIC > 1,000 nM).

^dNo IC₅₀ estimate of 2-ASA-14 for one P. falciparum isolate (MIC > 1,000 nM).

FIG. 2.

Ex vivo sensitivity (IC₅₀) of clinical field isolates of P. falciparum (closed circles) and P. vivax (open circles) from Papua, Indonesia, to SAHA, 2-ASA-9, and 2-ASA-14. Each symbol represents the IC₅₀ for one patient. Each horizontal bar represents the median IC₅₀ for the group of patients.

Stage-specific antimalarial activity.

To investigate the stage-specific drug susceptibility, isolates with greater than 90% rings were set up in culture in the presence of drug directly and again after ex vivo growth to 90% trophozoites in the absence of drug. For P. falciparum, isolates added to the assay at the trophozoite stage had significantly higher median IC₅₀s for SAHA (520 nM versus 329 nM) and 2-ASA-14 (467 nM versus 238 nM) than the same isolates exposed to drug at the ring stage (P = 0.036 and P = 0.028, respectively; Table 3). The derived IC₅₀s were also significantly higher for chloroquine, amodiaquine, piperaquine, and mefloquine, but not for artesunate. For P. vivax, a stage-specific difference in drug response was observed only for chloroquine (median IC₅₀ of 841 nM versus 53 nM; P = 0.008), and a less-significant difference was observed for amodiaquine (median IC₅₀ of 21.2 nM versus 10.1 nM; P = 0.036; Table 3).

TABLE 3.

Ex vivo drug susceptibility for paired isolates tested at the ring (>90% before culture) and trophozoite (>90% after culture) stage

Antimalarial	P. falciparumassay				P. vivaxassay
	na	Median IC50(range) (nM) for:		Pb	n	Median IC50(range) (nM) for:		P
		Rings	Trophozoites			Rings	Trophozoites
Chloroquine	8	30.2 (14.0-97.1)	63.2 (21.0-673.3)	0.036	9	53.0 (8.8-81.4)	841.4 (133-1930)	0.008
Amodiaquine	8	5.7 (1.0-28.6)	9.6 (4.5-40.5)	0.025	8	10.1 (3.9-15.5)	21.1 (4.0-36.3)	0.036
Piperaquine	8	11.5 (4.4-46.9)	87.6 (30.4-692.7)	0.012	9	15.5 (3.7-34.8)	15.9 (1.32-42.5)	0.515
Mefloquine	6	6.4 (2.7-12.7)	19.9 (6.4-22.5)	0.046	7	9.2 (1.5-23.2)	10.2 (2.3-27.2)	0.612
Artesunate	8	0.88 (0.31-4.11)	1.19 (0.59-4.18)	0.400	8	0.75 (0.09-1.03)	0.39 (0.2-1.09)	0.484
SAHA	8	329 (232-482)	520 (346-1016)	0.036	9	159 (67-223)	147 (40-229)	0.767
2-ASA-9	4	492 (463-533)	519 (362-988)	1.000	9	545 (203-710)	438 (149-2491)	0.767
2-ASA-14	6	238 (209-704)	467 (273-1043)	0.028	9	327 (121-502)	214 (145-588)	0.110

^an is the number of paired isolates.

^bThe Pvalues compare the median IC₅₀ values for rings and trophozoites by the Wilcoxon rank sum test. The statistically significant values are shown in boldface type.

Correlation in antimalarial susceptibility.

The drug responses to the different HDAC inhibitors were positively correlated in both P. falciparum and P. vivax isolates (Spearman rank correlation coefficients [r_s] of 0.515 to 0.818; Table 4). In P. falciparum isolates, inhibition by mefloquine (IC₅₀) correlated positively with SAHA (r_s = 0.635 and P = 0.004), but not 2-ASA-9 or 2-ASA-14. Conversely, in P. vivax, 2-ASA-9 and 2-ASA-14 (IC₅₀), but not SAHA, positively correlated with mefloquine (r_s = 0.527 and P = 0.030 and r_s = 0.610 and P = 0.007, respectively). In addition, P. vivax susceptibility to 2-ASA-14 was positively correlated with chloroquine (r_s = 0.490 and P = 0.028).

TABLE 4.

Correlation coefficients (r_s) for ex vivo antimalarial susceptibilities in P. falciparum and P. vivax

Antimalarial combination	All P. falciparumisolates			All P. vivaxisolates
	Correlationa	Pb	dfc	Correlation	P	df
SAHA-chloroquine	0.367	0.123	18	−0.012	0.960	19
SAHA-amodiaquine	0.167	0.495	18	0.363	0.127	18
SAHA-piperaquine	0.158	0.519	18	0.239	0.310	19
SAHA-mefloquine	0.635	0.004	18	0.191	0.448	17
SAHA-artesunate	0.156	0.523	18	0.300	0.212	18
SAHA-2-ASA-9	0.818	<0.001	14	0.519	0.023	18
SAHA-2-ASA-14	0.515	0.029	17	0.523	0.018	18
2-ASA-9-chloroquine	0.418	0.121	14	0.307	0.201	18
2-ASA-9-amodiaquine	0.104	0.713	14	0.401	0.099	19
2-ASA-9-piperaquine	0.282	0.308	14	0.209	0.391	18
2-ASA-9-mefloquine	0.450	0.092	14	0.527	0.030	16
2-ASA-9-artesunate	−0.057	0.840	14	0.174	0.489	17
2-ASA-9-2-ASA-14	0.662	0.010	13	0.542	0.017	18
2-ASA-14-chloroquine	0.123	0.627	17	0.490	0.028	19
2-ASA-14-amodiaquine	−0.108	0.669	17	0.321	0.180	18
2-ASA-14-piperaquine	0.389	0.111	17	0.087	0.715	19
2-ASA-14-mefloquine	0.123	0.627	17	0.610	0.007	17
2-ASA-14-artesunate	−0.113	0.657	17	0.233	0.336	18

^aSpearman rank correlation coefficients (r_s) are shown. The statistically significant values are shown in boldface type.

^bThe P values show the significance of the correlation between each pair of drugs, with those reaching statistical significance shown in boldface type.

^cdf, degrees of freedom.

DISCUSSION

Inhibitors of histone deacetylase (HDAC) enzymes represent a promising new class of antimalarial compounds (3). Suberoylanilide hydroxamic acid (SAHA) (Vorinostat) was licensed for use in humans for the treatment of advanced cutaneous lymphoma in 2006, and knowledge about the safety and pharmacokinetics of HDAC inhibitors is accruing (6, 44). In vitro drug susceptibility testing has demonstrated that several hydroxamate-based HDAC inhibitors are potent inhibitors (IC₅₀s in low nanomolar concentrations) of laboratory strains of P. falciparum and, importantly, display promising selectivity for malaria parasites versus mammalian cell lines (>100 fold), depending on the inhibitor and cell lines used (2, 12).

In the present study, we compared the ex vivo activities of three hydroxamate-based HDAC inhibitors against multidrug-resistant clinical isolates of P. falciparum and P. vivax. This is the first report of antimalarial activity of this class of compounds directly against clinical isolates and against P. vivax parasites. The derived potencies (IC₅₀) ranged from 87 to 964 nM against P. falciparum and 67 to 766 nM for P. vivax. This level of drug activity was within the range of well-tolerated plasma drug concentrations observed in phase 1 clinical drug trials in oncology patients (26, 27). However, the ex vivo susceptibility of clinical isolates of P. falciparum to HDAC inhibitors was significantly lower than previously reported in culture-adapted P. falciparum laboratory strains (Table 2). Although some of this difference may be explained by the level of multidrug resistance in Papua, Indonesia, the IC₅₀s for chloroquine were not excessive (median of 30 nM) and significantly lower than that reported for the chloroquine-resistant laboratory strain W2 (400 nM; Table 2). An alternative explanation is the difference in assay methodology. Previous studies employed proliferation-based growth inhibition assays of culture-adapted laboratory strains with in vitro growth quantified by [³H]hypoxanthine incorporation (4), whereas in the present study, inhibitory effects were determined in isolates fresh from the patient with quantification of ex vivo schizont maturation assessed by microscopy (38). This phenomenon (i.e., microscopic assessment of schizont maturation producing higher IC₅₀s than those derived from radioisotopic incorporation) has been described previously (46). In addition, even by applying the same methodology (i.e., [³H]hypoxanthine incorporation) but using different assay durations, IC₅₀ estimates derived by a one-cycle assay (42 h) can be 2- to 6-fold higher than the estimates derived from the standard assay (72 h) (29). In our study, IC₅₀s in chloroquine-sensitive and -resistant laboratory strains (K1 and 3D7, respectively) assessed by using the schizont maturation assay were indeed higher (700 to 1,000 nM) than the reported values derived by the isotopic assay (15 to 300 nM).

Our study also presents data for the first time on the ex vivo activity of HDAC inhibitors against the second most significant malaria parasite, P. vivax. On the basis of sequence homology, three class I/II HDAC homologues have been identified in P. falciparum—the class I HDAC homologue (P. falciparum HDAC 1 [PfHDAC1]) and two less-conserved putative HDACs (PfHDAC2 and PfHDAC3) about which little is known. Homologues for all three of these genes are present in other Plasmodium species, including P. vivax. The other HDAC homologues cannot be ruled out as possible targets of HDAC inhibitors in P. falciparum, and furthermore, a correlation between HDAC inhibitors with selective antimalarial action on parasite HDACs versus mammalian cell HDACs has not yet been demonstrated. However, recombinant PfHDAC1 has recently been generated, and the activity of this enzyme has been shown to be inhibited by hydroxamate-based HDAC inhibitors (32). PfHDAC1 and the P. vivax homologue (PlasmoDB gene identification [ID] PVX_099700) share ∼95% sequence identity at the amino acid level, and preliminary homology modeling using the same alignment as for an in silico PfHDAC1 homology model previously generated by us (2) revealed no substituted residues near the active site that were different (D. P. Fairlie and A. J. Lucke, unpublished results). There is currently no homology model available for the other PfHDACs due to limited sequence similarity with other eukaryotic HDAC enzymes.

In view of the confounding effects of stage specificity and duration of assay for drug susceptibility testing in P. vivax (38, 40), we have adopted more-stringent criteria for our P. vivax assay. In the present study, only isolates with more than 70% of the parasites at the ring stage were selected for testing and with culture maintained for a minimum of 30 h prior to harvest. Our results for P. vivax isolates tested at both the ring and trophozoite stages confirm the marked stage specificity of action of chloroquine, with trophozoites being almost completely resistant to the effect of the drug (38, 40). In contrast, the stage-specific action in P. falciparum was modest, with IC₅₀s less than 2-fold higher for trophozoite stages compared to ring stages. The results with HDAC inhibitors are consistent with a previous study in which the effect of continual exposure of P. falciparum line Dd2 to 2-ASA-9 and 2-ASA-14 was examined at the ring stage (∼18 h after invasion), trophozoite stage (∼26 h after invasion), and schizont stage (∼38 h after invasion). Microscopic examination showed that all stages were inhibited, with very small numbers of invasive merozoites being released from schizonts (2).

The ex vivo activity between the HDAC inhibitors was strongly correlated (r_s = 0.515 to 0.818; Table 4). Whereas the drug susceptibility of SAHA was correlated with mefloquine in P. falciparum (r_s = 0.635), this did not reach statistical significance in P. vivax. Conversely, the activities of 2-ASA-9 and 2-ASA-14 were highly correlated with mefloquine in P. vivax, but not in P. falciparum. The statistical significance of our observations may simply reflect the small sample size of the study and intrinsic variance of our assay. Alternatively, ex vivo drug efficacy may not be determined solely by structural differences in the HDAC inhibitors and their plasmodial targets. While cross-susceptibility between HDAC inhibitors and chloroquine or mefloquine has not been observed in sensitive and resistant laboratory strains (2, 12), the differences in correlation profiles for P. falciparum and P. vivax may indicate different susceptibility profiles in these species. Molecular mechanisms underlying chloroquine resistance are likely to differ significantly between P. vivax and P. falciparum (31, 43). However, this is not the case for mefloquine susceptibility, which has been correlated with increased copy number of the mdr1 gene (P-glycoprotein) in both species (34, 42). There is limited data on reduced susceptibility to HDAC inhibitors in eukaryotic cells, but mdr1-mediated tolerance has been reported for the non-hydroxamate-based compounds apicidin and depsipeptide (FK-228) in several cancer cell lines (28, 48). However, there is also evidence for mdr1-independent mechanisms (15), and a more recent study has shown that HDAC inhibitors were able to promote cell line-specific and reversible induction of multiple drug transporter genes (20). Antimalarial resistance is postulated to be a multigenic process (13), and it is possible that reduced Plasmodium susceptibility to HDAC inhibitors will occur. However, since HDAC inhibitors have an effect on multiple vital eukaryotic cell functions (49) and potentially act on both histone and non-histone-related pathways, it is difficult to predict exactly how Plasmodium will develop resistance to this drug class.

In our study, hydroxamate-based HDAC inhibitors SAHA, 2-ASA-9, and 2-ASA-14 showed potent ex vivo efficacy against multidrug-resistant Plasmodium field isolates. Importantly, ex vivo activity of HDAC inhibitors against P. vivax was comparable to that in P. falciparum, suggesting that HDAC inhibitors may be promising candidates for antimalarial therapy in geographical areas where both species are endemic. Elucidation of the species-specific structure of HDAC analogues in Plasmodium should facilitate molecular design aimed at optimizing the selectivity and efficacy of HDAC inhibitors. Studies assessing the comparative clinical efficacy in Plasmodium species and defining drug interaction with common antimalarial compounds will reveal further the potential of HDAC inhibitors in antimalarial therapy.