Department of Biochemistry and Molecular Biology, Oregon Health Sciences University, Portland, ... from the Aldrich Chemical Company (Milwaukee, WI).
Am. J. Trop. Med. Hyg., 62(1), 2000, pp. 77–81 Copyright 䉷 2000 by The American Society of Tropical Medicine and Hygiene
XANTHONES AS ANTIMALARIAL AGENTS: STAGE SPECIFICITY MARINA V. IGNATUSHCHENKO, ROLF W. WINTER, AND MICHAEL RISCOE Department of Biochemistry and Molecular Biology, Oregon Health Sciences University, Portland, Oregon; Medical Research Service, Department of Veterans Affairs Medical Center, Portland, Oregon; Interlab, Inc., Lake Oswego, Oregon
Abstract. The erythrocytic development of Plasmodium falciparum is divided into the ring, trophozoite, and schizont stages based on morphologic assessment. Using highly synchronous ring and trophozoite cultures of P. falciparum, we observed considerable differences in their sensitivity to hydroxyxanthones: trophozoites were much more sensitive to the drugs than ring-stage parasites. Trophozoites treated with a prototypic xanthone, the 2,3,4,5,6pentahydroxy derivative (X5), were arrested in their development and became degenerate in appearance within 24 hr of drug exposure. These morphologic changes appeared to reflect the cytotoxic nature of the action of the drug against the parasite, since daughter ring-stage forms were not observed following addition of the drug. That X5 was more active against parasites in the later stages of intraerythrocytic development is consistent with the proposed mode of action, inhibition of heme polymerization. Knowledge of the structure-activity relationships for xanthones as antimalarial agents has also been expanded. Xanthones with a hydroxyl group in the peri-position exhibited decreased antimalarial activity, possibly due to intramolecular hydrogen bonding with the carbonyl and consequent reduced affinity for heme. Paired hydroxyls attached to the lower half of the xanthone greatly enhanced drug potency. Fisher Scientific (Pittsburgh, PA). The xanthones used in this study were synthesized in the Medicinal Chemistry Division of Interlab, Inc. (Lake Oswego, OR). Culture of P. falciparum. The chloroquine-susceptible D6 clone of P. falciparum has been previously described.6,7 The parasites were cultured in group A⫹ human erythrocytes suspended at a 2% hematocrit in RPMI 1640 medium (pH 7.15) that contained 3 g/L of glucose, 50 g/L of gentamicin, and 10% human serum, and maintained at 37⬚C in a gas mixture of 5% O2, 5% CO2, and 90% N2. For the experiments assessing drug efficacy on synchronized cultures, parasites were subjected to 2 cycles of sorbitol treatment separated by 34 hr, yielding tightly synchronized (0–4 hr) ring-stage cultures.8 The ring-stage cultures were incubated for an additional 24–28 hr to yield parasites at the trophozoite stage of development. Light microscopy. Thin blood smears prepared from X5treated and control cultures were air-dried and stained with a modified Wright-Giemsa stain (LeukoStat Stain Kit). The films were examined microscopically under oil immersion at 1,000⫻ magnification. Drug testing. In vitro antimalarial activity of the test compounds was assessed by following incorporation of [3H]-ethanolamine (50 Ci/mmol) into parasite lipids as described by Elabbadi and others with minor modifications.9 Stock solutions of the xanthones were made by dissolving the compounds in dimethylsulfoxide at a concentration of 10 mM. The experiments were set up in duplicate in 96-well plates with various concentrations of each xanthone (10⫺9 to 10⫺4 M) across the plate in a total volume of 200 l and at a final red blood cell concentration of 2% (v/v). An initial parasitemia of 1% was attained by addition of uninfected red blood cells to the stock culture of infected red blood cells (0–4-hr rings or 24–28-hr trophozoites). [3H]-Ethanolamine (50 Ci/ mmol, 1 Ci in 20 l of medium) and the test drug were added to each well, and the experiments were terminated after 24 hr of incubation by collecting the cells onto glass fiber filters with a semiautomated Tomtec (Orange, CT) 96well plate harvester. [3H]-Ethanolamine uptake was quantitated by scintillation counting of the filters using a Wallac
We were led to investigate the antiparasitic action of xanthones by the discovery of a remarkable antimalarial synergy between exifone (a hexahydroxybenzophenone) and two oxidant drugs (rufigallol and ascorbic acid).1–3 We speculated that free radical hydroxylation of exifone followed by cyclodehydration inside parasitized red blood cells results in the formation of 2,3,4,5,6-pentahydroxyxanthone (X5), and that this xanthone is responsible for the enhanced antimalarial effect.3 Synthetic X5 was shown to possess an impressive inhibitory activity in vitro against both chloroquinesensitive and multidrug-resistant strains of Plasmodium falciparum.3 As a direct result of hemoglobin digestion, a vast quantity of heme is liberated in the food vacuole of the parasite.4 To avoid accumulation of toxic heme during this catabolic process, the parasite has evolved a mechanism of heme polymerization that results in the formation of an insoluble substance commonly referred to as hemozoin. We have demonstrated the ability of X5 to form soluble complexes with heme and to prevent the polymerization of heme in vitro.5 Figure 1 shows the results of one such experiment in which heme was incubated under mildly acidic conditions in the presence and absence of X5. These findings support the hypothesis that the antimalarial activity of X5 is due to inhibition of heme polymerization in the parasite digestive vacuole. Structure-activity profiling of a limited number of xanthones pointed to the 4,5-dihydroxy derivative as the minimal structural unit of X5 that retained activity in the heme polymerization assay and against malarial parasites. In this report we describe the stage-specific action of X5 against malaria parasites and provide additional details on the structure-activity relationships of hydroxyxanthones. MATERIALS AND METHODS
Chemicals and reagents. Hemin chloride was obtained from the Aldrich Chemical Company (Milwaukee, WI). [3H]-Ethanolamine was obtained from American Radiolabeled Chemicals, Inc. (St. Louis, MO). The LeukoStat Stain Kit for staining parasitized erythrocytes was obtained from
77
78
IGNATUSHCHENKO AND OTHERS
FIGURE 1. Inhibition of in vitro heme polymerization by hydroxyxanthones. 25 M hematin was incubated in 10 ml of 20 mM phosphate solution, pH 5.2 for 2 hr at 37⬚C in the absence (left) or presence (right) of 25 M 2,3,4,5,6-pentahydroxyxanthone (X5). Note that the solution in the tube on the right is completely clear, but the low resolution digital image makes it appear granular.
(Gaithersburg, MD) 1205 Betaplate counter. The concentration of a drug giving 50% inhibition of label incorporation (IC50) was calculated from a computer-generated semi-logarithmic dose-response curve. The IC50 values presented are the average of at least 2 separate determinations of full doseresponse curves each performed in duplicate. In vitro heme polymerization inhibition assay. The dose-dependent inhibition of heme polymerization in a phosphate solution (pH 5.2, 37⬚C) was evaluated as described previously.5 A 10 mM hematin stock solution was prepared by dissolving hemin chloride in 0.1 M NaOH followed by incubation at 37⬚C for at least 1 hr prior to use. The drugs were dissolved in dimethylformamide at a concentration of 10 mM. The concentration of each drug in the assay was varied in the range of 0–100 M; the initial hematin concentration was 25 M. The IC50 values were determined by non-linear regression analysis of percentage inhibition of heme polymerization versus drug concentration. RESULTS
Morphologic assessment of xanthone action. Synchronous cultures of P. falciparum (D6 strain) were incubated → FIGURE 2. Morphologic effects of 2,3,4,5,6-pentahydroxyxanthone (X5) on Plasmodium falciparum. The parasites were incubated for up to 48 hr beginning at the early trophozoite stage (i.e., 24–28
hr after synchronization). Control cultures contained healthy late trophozoite parasites (A, arrows) at 12 hr of incubation and ring-form parasites (B, arrows) at 24 hr of incubation. Cultures incubated with 10 M of X5 contained degenerate trophozoites (C, arrows) at 24 hr of incubation. (Original magnification ⫻ 1,000.)
79
STAGE-SPECIFIC ANTIMALARIAL ACTION OF XANTHONES
TABLE 1 In vitro potency of hydroxyxanthones on synchronous trophozoites of Plasmodium falciparum Xanthones
Monohydroxyxanthones Dihydroxyxanthones
Trihydroxyxanthones
Tetrahydroxyxanthone
Pentahydroxyxanthones Hexahydroxyxanthones
Compound number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Chemical name
2-hydroxyxanthone 3-hydroxyxanthone 1,3-dihydroxyxanthone 2,5-dihydroxyxanthone 3,6-dihydroxyxanthone 4,5-dihydroxyxanthone 1,3,5-trihydroxyxanthone 2,3,4-trihydroxyxanthone 3,4,5-trihydroxyxanthone 3,4,6-trihydroxyxanthone 2,3,4,5-tetrahydroxyxanthone 2,3,4,6-tetrahydroxyxanthone 1,3,5,6-tetrahydroxyxanthone 3,4,5,6-tetrahydroxyxanthone 2,3,4,5,6-pentahydroxyxanthone 1,3,5,6,7-pentahydroxyxanthone 1,2,3,5,6,7-hexahydroxyxanthone 2,3,4,5,6,7-hexahydroxyxanthone
IC50, M P. falciparum D6, synchronous trophozoites*
⬎50 ⬎100 ⬎60 53 ⫾ 10 ⬎60 28 ⫾ 6 ⬎60 36 ⫾ 7 45 ⫾ 3 35 ⫾ 4 9.0 ⫾ 1.0 30 ⫾ 15 ⬎50 1.3 ⫾ 0.7 0.7 ⫾ 0.5 6.5 ⫾ 0.5 54 0.2 ⫾ 0.1
* Values are the mean ⫾ SD for at least 3 independent determinations. IC50 ⫽ 50% inhibitory concentration.
with or without X5 (10 M) for up to 48 hr beginning at the early trophozoite stage. Light microscopy of Wright-Giemsa–stained parasites at 12, 24, and 48 hr following addition of the drug revealed that there was developmental arrest at the trophozoite stage (Figure 2). Control cultures contained healthy late-stage trophozoites at 12 hr of incubation (Figure 2A) and daughter ring-form parasites at 24 hr of incubation (Figure 2B). Cultures incubated with 10 M X5 contained degenerate trophozoites at 24 hr of incubation
FIGURE 3. Correlation between the inhibitory activities of hydroxyxanthones in the heme polymerization assay and in the Plasmodium falciparum growth inhibition assay (strain D6, synchronous trophozoite cultures). IC50 ⫽ 50% inhibitory concentration.
(Figure 2C). The drug-treated parasites exhibited morphologic abnormalities such as a pale appearance and the presence of eccentrically located, pyknotic masses of irregular shape. By 48 hr of incubation, parasitemias in the drug-treated cultures remained unchanged, with only degenerate parasites being visible—clear evidence of the cytotoxic nature of X5 action. When X5 exposure was initiated at the ring stage, the toxic effects of drug exposure were not apparent until the parasites had progressed into the trophozoite stage, and daughter ring stage forms were not observed on further incubation. Growth inhibition of synchronous cultures by xanthone derivatives. To determine the stage specificity of other hydroxyxanthones, we compared the antimalarial activity for each compound against highly synchronous ring- and trophozoite-stage parasites. In this assay, incubation with radioactive ethanolamine and exposure to the test drug occurred simultaneously. Decreased label incorporation into parasite lipids during the brief 24-hr incubation period indicated the growth inhibitory effect of the drug on either the first or the second half of the erythrocytic developmental cycle. All of the xanthones tested were inactive against parasites in the early ring stage at concentrations up to 60 M. However, when tested against synchronous trophozoites, many of the compounds exhibited IC50 values in the low micromolar and submicromolar range (Table 1), their relative activities showing a good correlation with those obtained in the standard 72-hr assay using asynchronous parasite cultures.5 In general, xanthones bearing hydroxy groups at both 4- and 5-positions, especially when paired with neighboring 3- and 6-position hydroxyls, demonstrated the most potent activity against synchronous trophozoites and were among the most active in the heme polymerization assay (Figure 3). On closer inspection of the data, it was observed that xanthones containing hydroxy groups at the periposition (i.e., 1- or 8-) were less active than corresponding isomers without this substitution pattern. For example, 1,3,5trihydroxyxanthone was completely inactive in the heme polymerization assay (IC50 ⬎ 100 M) and without activity
80
IGNATUSHCHENKO AND OTHERS
against P. falciparum parasites in vitro at 60 M, the highest concentration tested. In contrast, the 2,3,4-derivative was active in both systems (IC50 艐 10 M in the heme polymerization assay and IC50 艐 36 M against P. falciparum). Isomeric comparisons of di-, tetra-, penta-, and hexahydroxyxanthones yielded a similar pattern, reflecting the negative influence of peri-hydroxylation. As before, a higher degree of hydroxylation favored the growth inhibitory activity for the xanthone series against parasites at the trophozoite stage. An example of this structural relationship is the 140-fold increased potency of 2,3,4,5,6,7-hexahydroxyxanthone compared with 4,5-dihydroxyxanthone. Replacement of the ring oxygen by sulfur to produce 4,5dihydroxythioxanthone led to a definite reduction of activity in both heme polymerization (IC50 ⬎ 50 M) and growth inhibition (IC50 艐 60 M) assays compared with the corresponding xanthone homolog. DISCUSSION
Proteolysis of hemoglobin in the acidic food vacuole and detoxification of released heme via polymerization into insoluble hemozoin are key metabolic processes of the Plasmodium parasite.4,10 Although the actual mechanism of heme polymerization has yet to be resolved, it has been suggested that it is a spontaneous process that can be carried out in vitro.11,12 We have recently developed an in vitro heme polymerization assay; the reaction proceeds in the absence of iron-coordinating buffers and yields a product with physical and chemical properties of malarial hemozoin.5 We have used the assay to test a number of compounds as heme polymerization inhibitors and have demonstrated that several hydroxyxanthones prevent heme polymerization in this system. Moreover, the relative abilities of these compounds to inhibit polymerization in the assay showed a good correlation with their antimalarial properties. Certain structural features, such as higher degree of hydroxylation and the presence of hydroxy groups at the 4- and 5-positions, were found to favor the activity of xanthones in both systems, thereby suggesting that xanthones exert their antimalarial action by forming soluble complexes with heme liberated during hemoglobinolysis. Based on these findings, we proposed a model for a xanthone-heme complex.5 Since nearly all the available host hemoglobin is consumed during the trophozoite stage of the Plasmodium intraerythrocytic cycle, it was anticipated that xanthones would exert their greatest activity against parasites in the latter stages of development.13,14 To investigate the possible stage-specific actions of xanthones and their analogs, we streamlined the experimental setup for drug screening. This allowed us to focus on either the first or the second half of the intraerythrocytic developmental cycle of the parasite. We found that all of the xanthones included in our study were inactive against the early (ring) forms of P. falciparum, whereas mature trophozoites proved to be sensitive to xanthone action, some with inhibitory activities in the low micromolar to nanomolar range. The relative order of potency of these drugs against synchronous trophozoites showed a good correlation with the relative order of potency observed in the heme polymerization assay (Figure 3). Microscopic investigation revealed that parasites treated with the proto-
typic hydroxyxanthone, X5, halted their development during the late stages of the intraerythrocytic cycle. Twenty-four hours after drug administration, the parasites appeared morphologically degenerate with lucent cytoplasm, pyknotic masses, and decreased or absent malarial pigment particles. These observations are consistent with the notion that xanthones block the polymerization process, and that accumulation of soluble heme-drug complexes increases the osmotic pressure in the vacuole, causing its lysis. To develop a structure-activity profile for xanthones against trophozoite stage parasites, we compared a number of isomeric hydroxyxanthones and included the sulfur homolog, 4,5-dihydroxythioxanthone. Approximately 50 xanthones (see Table 1 for selected examples) were evaluated for antimalarial activity and for inhibitory activity in our heme polymerization assay. Our findings of the structureactivity relationships of xanthones against trophozoites are consistent with earlier determinations conducted on asynchronous parasite cultures, indicating the critical nature of hydroxylation at the 4- and 5-positions.5 We can now add that isomers bearing hydroxyl substituents at the peri-position are less active than those without this substitution, even if the 4- or 5-position is hydroxylated. The diminished antimalarial activity of peri-hydroxylated xanthones may result from intramolecular hydrogen bonding with the carbonyl moiety, resulting in decreased affinity for the drug target, heme. It is also noteworthy that our studies point to the xanthone nucleus as being superior to the thioxanthone nucleus as a lead structure for development of a novel antimalarial agent. Taken together, these findings demonstrate that xanthones exert their primary antimalarial effect during the second half of the Plasmodium erythrocytic cycle, when the production of free heme reaches its peak. Financial support: This project received financial support from the Merit Review Program of the Department of Veterans Affairs (MKR) and from a grant by the Department of Veterans Affairs to the Portland Environmental Hazards Research Center, a joint project of the Portland Veterans Affairs Medical Center and the Center for Research on Occupational and Environmental Toxicology, Oregon Health Sciences University. We also gratefully acknowledge financial contributions by Interlab, Inc. (Lake Oswego, OR), which has received support for this work from the U.S. Department of Defense Small Business Innovative Research program. No author has an undisclosed conflict of interest. Authors’ addresses: Marina V. Ignatushchenko, Department of Molecular Microbiology and Immunology, Oregon Health Sciences University, 3181 SW Sam Jackson Road, Portland, OR 97201. Rolf W. Winter and Michael Riscoe, Medical Research Service, RD-33, Veterans Affairs Medical Center, 3710 SW U.S. Veterans Hospital Road, Portland, OR 97201. REFERENCES
1. Winter RW, Cornell KA, Johnson LL, Riscoe MK, 1995. Hydroxy-anthraquinones as antimalarial agents. Bioorg Med Chem Lett 5: 1927–1932. 2. Winter RW, Cornell KA, Johnson LL, Ignatushchenko M, Hinrichs DJ, Riscoe MK, 1996. Potentiation of the antimalarial agent rufigallol. Antimicrob Agents Chemother 40: 1408– 1411. 3. Winter RW, Ignatushchenko M, Ogundahunsi OA, Cornell KA, Oduola AM, Hinrichs DJ, Riscoe MK, 1997. Potentiation of
STAGE-SPECIFIC ANTIMALARIAL ACTION OF XANTHONES
4. 5. 6.
7.
8.
an antimalarial oxidant drug. Antimicrob Agents Chemother 41: 1449–1454. Olliaro P, Goldberg D, 1995. The Plasmodium digestive vacuole: metabolic headquarters and choice drug target. Parasitol Today 11: 294–297. Ignatushchenko MV, Winter RW, Bachinger HP, Hinrichs DJ, Riscoe MK, 1997. Xanthones as antimalarial agents: studies of a possible mode of action. FEBS Lett 409: 67–73. Oduola AM, Milhous WK, Salako LA, Walker O, Desjardins RE, 1987. Reduced in-vitro susceptibility to mefloquine in West African isolates of Plasmodium falciparum. Lancet 2: 1304–1305. Oduola AM, Weatherly NF, Bowdre JH, Desjardins RE, 1988. Plasmodium falciparum: cloning by single-erythrocyte micromanipulation and heterogeneity in vitro. Exp Parasitol 66: 86–95. Lambros C, Vanderberg JP, 1979. Synchronization of Plasmodium falciparum erythrocytic stages in culture. J Parasitol 65: 418–420.
81
9. Elabbadi N, Ancelin ML, Vial HJ, 1992. Use of radioactive ethanolamine incorporation into phospholipids to assess in vitro antimalarial activity by the semiautomated microdilution technique. Antimicrob Agents Chemother 36: 50–55. 10. Sherman IW, 1979. Biochemistry of Plasmodium (malarial parasites). Microbiol Rev 43: 453–495. 11. Dorn A, Stoffel R, Matile H, Bubendorf A, Ridley RG, 1995. Malarial haemozoin/-haematin supports haem polymerization in the absence of protein. Nature 374: 269–271. 12. Egan TJ, Ross DC, Adams PA, 1994. Quinoline anti-malarial drugs inhibit spontaneous formation of -haematin (malaria pigment). FEBS Lett 352: 54–57. 13. Pandey AV, Tekwani BL, Pandey VC, 1995. Characterization of hemozoin from liver and spleen of mice infected with Plasmodium yoelii, a rodent malaria parasite. Biomed Res 16: 115–120. 14. Sullivan DJ Jr, Gluzman IY, Russell DG, Goldberg DE, 1996. On the molecular mechanism of chloroquine’s antimalarial action. Proc Natl Acad Sci USA 93: 11865–11870.
