Antimalarial 9-Anilinoacridine Compounds Directed at Hematin

Antimalarial 9-anilinoacridines are potent inhibitors of parasiteDNA topoisomerase II both in vitro and in situ. 3,6-Diaminosubstitution on the acridine ring greatly improves parasiticidalactivity against Plasmodium falciparum by targeting DNA topoisomeraseII. A series of 9-anilinoacridines were investigated for theirabilities to inhibit ß-hematin formation, to formdrug-hematin complexes, and to enhance hematin-induced lysisof red blood cells. Inhibition of ß-hematin formationwas minimal with 3,6-diamino analogs of 9-anilinoacridine andgreatest with analogs with a 3,6-diCl substitution togetherwith an electron-donating group in the 1'-anilino position.On the other hand, the presence of a 1'-N(CH₃)₂ group in theanilino ring produced compounds that strongly inhibited ß-hematinformation but which did not appear to be sensitive to the natureof the substitutions in the acridine nucleus. The derivativesbound hematin, and Job's plots of UV-visible absorbance changesin drug-hematin complexes at various molar ratios indicateda stoichiometric ratio of 1:2. The drugs enhanced hematin-inducedred blood cell lysis at low concentrations (<4 µM).These studies open up the novel possibility of development of9-anilinoacridine antimalarials that target not only DNA topoisomeraseII but also ß-hematin formation, which should helpdelay the rapid onset of resistance to drugs acting at onlya single site.

INTRODUCTION

Top

Introduction

Malaria still presents an important health problem in tropicalregions of the world, with over 275 million new cases annuallyand mortality reaching 2 million, particularly among childrenin Africa (21). Control of this debilitating disease has beenseverely compromised by the development in malaria parasitesof resistance to nearly all antimalarial drugs used for prophylaxisand treatment, particularly in Plasmodium falciparum, the mostvirulent of the four species infecting humans. Thus, there isa compelling and urgent need for new antimalarials with modesof action different from those of existing ones in order toreplace those that are becoming obsolete and to identify newdrug targets.

Antimalarial drugs may be broadly classified into two groups:lysosomotropic quinoline-containing drugs and antimetabolites.The former compounds include chloroquine, quinine, quinacrine,amodiaquine, mefloquine, and related drugs (33). All of thesedrugs act exclusively on the blood stages of P. falciparum.The quinolines, which are the mainstays of treatment and prophylaxis,are excellent parasiticidal agents for susceptible organismsbut are becoming increasingly less efficacious. Chloroquinewas introduced for the control of malaria in the 1940s withgreat optimism because the drug was cheap, highly effective,and well tolerated. Due to its specificity, stability, and safety,chloroquine has been one of the most successful and widely usedantimalarial drugs. However, the mechanisms of action of drugsthat contain a quinoline nucleus have, until recently, remainedobscure.

In P. falciparum, hemoglobin is degraded in an acidic vacuolarcompartment to yield amino acids for parasite protein synthesis,and this occurs predominantly at the trophozoite and early schizontstages (16). In higher eukaryotes, heme oxygenase catabolizesthe porphyrin moiety to carbon monoxide, ferric ion, and biliverdin (26).Although the malaria parasite contains heme oxygenase (29),this metabolic pathway is not used (9), and instead, heme isdetoxified by a number of other mechanisms. These include sequestrationinto an insoluble material termed hemozoin or malaria pigment(27) and degradation by glutathione-dependent (19) and peroxidative (24)processes, but the major mechanism of heme detoxification appearsto be the hemozoin sequestration pathway (11).

Chloroquine has recently been shown to inhibit hemozoin formationwithin the parasite food vacuole (30). This process is alsothought to be the molecular target of other quinoline antimalarials(33). Hemozoin was originally considered to be formed by thepolymerization of heme (28), but it has now been demonstratedto be a crystalline cyclic dimer of ferriprotoporphyrin IX (25).Antimalarials such as chloroquine can be considered crystallizationinhibitors or agents that act to divert heme from participatingin the crystallization process (25), leading to the accumulationof free heme, which is potentially toxic. This mechanism hasalso been attributed to the actions of antimalarial acridinessuch as quinacrine (6) and pyronaridine (8), a 9-anilinoaza-acridinederivative that was originally used to treat chloroquine-resistantfalciparum malaria in China and that is now undergoing clinicaltrials in other regions of the world (23). Thus, hemozoin synthesis,a process unique to the malaria parasite, offers a logical andvaluable potential target for new antimalarial drug development. Newdrugs that attack the same vital target of chloroquine but thatare not subject to the same resistance mechanism would be highly desirable.

We have previously shown that 9-anilinoacridines (Fig. 1) havegood antimalarial activities in vitro (3, 15, 17), being potent inhibitorsof parasite DNA topoisomerase II both in vitro (17) and in situ (1).3,6-Diamino substitution of the acridine ring of 9-anilinoacridinesgreatly improves the potencies of these agents against P. falciparum. Inthis study, a series of 9-anilinoacridines were investigatedfor their abilities to inhibit ß-hematin formationin vitro (a process which closely parallels hemozoin synthesiswithin the parasite food vacuole) (13), to form drug-hematincomplexes, and to enhance hematin-induced lysis of red bloodcells.

View larger version (13K):
[in this window]
[in a new window]
FIG. 1. Structures of pyronaridine, chloroquine, and 9-anilinoacridine.

MATERIALS AND METHODS

Top

Materials and METHODS

Drugs. The drugs used in the study were available from previous workand were prepared as described in the report of that study (17).

In vitro assessment of antimalarial activity. The protocol for in vitro testing of the activities of the drugsagainst P. falciparum-infected erythrocytes was a modificationof the [³H]hypoxanthine incorporation method of Desjardins etal. (7). Sorbitol-synchronized parasite cultures (22) with mostlyring stages of the P. falciparum K1 (chloroquine- and pyrimethamine-resistant)strain (31) and the T9/94 (drug-sensitive) clone (32) were used.The levels of parasitemia were determined from Giemsa-stainedthin blood films of parasite cultures. Packed infected red cellswere adjusted to approximately 1.0% by the addition of 50% (vol/vol) uninfectederythrocytes. An aliquot of 0.3 ml of this suspension (50% hematocrit,1.0% parasitemia) was added to 9.7 ml of complete medium (10%serum in RPMI 1640 culture medium supplemented with 32 mM NaHCO₃and 25 mM HEPES buffer [pH 6.5]). Drugs for testing were initiallydissolved in dimethyl sulfoxide (DMSO) and diluted to the requiredconcentrations with the supplemented RPMI 1640 culture mediumdescribed above. Aliquots of 200 µl of the cell suspensiondescribed above (1.0% parasitemia of early ring stage) wereincubated with 25 µl of the drug-containing medium (thefinal concentration of DMSO was 0.01%, which did not affectparasite growth) in 96-well microtiter plates under candle jarconditions for 24 h at 37°C prior to the addition of 25µl of 0.5 µCi of [³H]hypoxanthine (specific activity,28.0 Ci/mmol; Amersham, Little Chalfont, United Kingdom) (34).After a further incubation for 18 h, the parasites were harvestedfrom each well, placed onto glass fiber filters (grade 934 AH;Whatman), and lysed with distilled water. The filter disks werethen placed in 1.5-ml microcentrifuge tubes and dried at 60°Cfor 1 h. An aliquot of 0.6 ml of liquid scintillation cocktail {0.35%(wt/vol) 2,4-diphenyloxazole and 0.005% (wt/vol) 1,4-bis[2-(5-phenyloxazoyl)]benzenein toluene} was added to each tube. The microcentrifuge tubeswere placed in scintillation vials, and the levels of [³H]hypoxanthineincorporation into the parasite were determined in a liquidscintillation counter (model LS-1801; Beckman). The 50% inhibitoryconcentrations (IC₅₀s; the concentration that resulted in 50%inhibition of radioactivity incorporation compared to that forthe controls) were obtained from the drug dose-response curves.

Assay of ß-hematin formation inhibition. Determination of the inhibition of ß-hematin formationwas based on the method of Baelmans et al. (2), as modifiedfrom that of Egan et al. (10). A 100-µl aliquot of freshlyprepared hematin solution (41.175 mg of hematin [Sigma] in 10ml of 0.2 M NaOH) was mixed with 50 µl of drug solutionor either 50 µl of water or DMSO as controls, prior tothe addition of 200 µl of 3 M sodium acetate and 50 µlof 17.4 M acetic acid (final pH, 4.8). After 24 h of incubationat 37°C, the tubes were centrifuged at 4,000 x g for 15min and the supernatants were removed. The pellets were resuspendedfive times in 200 µl of DMSO to remove unreacted hematinfrom ß-hematin, which is insoluble in this solvent.The pellets were then dissolved in 0.1 M NaOH for spectroscopicquantification at 405 nm. A standard curve of hematin dissolvedin 0.1 M NaOH was used to calculate the amount of hematin present.The percentage of ß-hematin formed was calculatedby using the following formula: (moles of hematin in sample/molesof hematin in control) x 100.

The results of drug testing were expressed as the IC₅₀ of ß-hematinformation, obtained by nonlinear regression analysis of thedrug dose-response curves. Data are presented as the means ofthree independent experiments, each of which was performed induplicate.

Drug-hematin interaction assay. An aqueous DMSO (40%; vol/vol) solution of 10 µM hematin (pH7.4) was freshly prepared by mixing 25 µl of 4 mM hematin (Sigma)in 0.1 M NaOH solution with 4 ml of DMSO and 1 ml of 0.02 M sodiumphosphate buffer (pH 6.0) and making the volume up to 10 mlwith double-distilled deionized water (under these conditions,hematin is monomeric) (14). Solutions of 9-anilinoacridine compoundswere similarly prepared. To examine hematin-drug interactions,a continuous variation technique (Job's plot) was performedto determine the spectral changes (20). For each compound, solutionscontaining drug and hematin combinations at the following 14 molarratios were prepared: 0:1, 1:9, 1:4, 3:7, 2:3, 1:1, 3:2, 5:3, 13:7,27:13, 7:3, 4:1, 9:1, and 1:0. The final combined concentration ofhematin plus drug in the mixtures was 10 µM. Spectra between 240and 700 nm were recorded in a Shimadzu UV-250 IPC spectrophotometer ata speed of 0.5 nm/min. Baseline values were routinely subtracted fromthe spectra.

Drug-hematin-induced red blood cell lysis. Experiments that evaluated the lysis of human red blood cellsby hematin and drug-hematin complexes were conducted by incubating0.03% (vol/vol) cell suspensions in phosphate-buffered saline(pH 7.4) at 37°C for 1 h with various concentrations ofhematin in the absence or the presence of 9-anilinoacridineanalogs and measuring the decrease in absorbance at 700 nm.

RESULTS AND DISCUSSION

Top

Results and DISCUSSION

Fourteen 9-anilinoacridine analogs were tested for their effectsagainst chloroquine-sensitive P. falciparum clone T9/94 andchloroquine-resistant strain K1 (Table 1). Compounds 1, 2, 9,10, 11, and 13, which contained either an unsubstituted acridinenucleus or the 3-NH₂, 3-Cl, or 3,6-diCl substituent, had notpreviously been evaluated for their antimalarial activities.All the compounds tested displayed in vitro activities againstthe K1 strain comparable to that against the T9/94 clone, indicatingthat these 9-anilinoacridine drugs are not subject to the chloroquineresistance mechanism operating in P. falciparum. Consistentwith a previous study (17), the presence of the 3,6-diNH₂ substitutionenhanced the antimalarial activity, supporting our contentionthat this group of 9-anilinoacridine analogs targets parasiteDNA topoisomerase II in situ (1).

View this table:
[in this window]
[in a new window]
TABLE 1. Inhibitory effects of 9-anilinoacidine analogs on growth of P. falciparum K1 (chloroquine resistant) and T9/94 (chloroquine sensitive) and on ß-hematin formation

An in vitro assay was used to assess the abilities of the 9-anilinoacridine compoundsto inhibit ß-hematin formation. In that assay hematin wasallowed to form ß-hematin under acidic conditions.Of the six 3,6-diNH₂-substituted analogs tested, five showedno measurable activity, and only compound 3, with an N(CH₃)₂substituent in the anilino ring, inhibited ß-hematinformation with an IC₅₀ equivalent to that of chloroquine (Table 1).The presence of a 1'-N(CH₃)₂ group in the anilino ring appearedto be favorable for this activity, since all compounds possessingthis group (compounds 1, 2, 3, and 9) showed measurable levelsof inhibition of ß-hematin formation, regardless ofthe nature of the substitutions in the acridine nucleus. Thus,the potent compound 1'-N(CH₃)₂-3,6-diNH₂-9-anilinoacridine (compound3) not only exerts its antimalarial activity by inhibiting DNAtopoisomerase II in situ (1) but may also act by inhibitinghemozoin formation.

In two series (1'-SO₂NH₂ and 1'-NHSO₂Me, where Me is a methylgroup), the 3,6-diCl analogs were more active than the 3,6-diNH₂analogs in inhibiting ß-hematin formation, with 1'-NHSO₂Me-3,6-diCl(compound 13) being the most active analog evaluated (IC₅₀,0.062 mM, compared to a chloroquine IC₅₀ of 0.125 mM). Havingonly one Cl substituent (compound 11) in the acridine ring didnot improve the activity of the analog compared with that ofthe corresponding disubstituted analog. In the case of the 4-aminoquinolines,the related 7-Cl group has been shown to be absolutely requiredfor inhibition of ß-hematin formation (12), but thisis not so for the 9-anilinoacridines (compare the results forcompounds 1, 2, 3, and 9 in Table 1). The nature of the substitutionin the anilino ring also affects this activity (compare theresults for compounds 12, 13, and 14 in Table 1). However, thepoor antimalarial activities of the 3-Cl and 3,6-diCl derivativesof 9-anilinoacridine (compared with those of the corresponding NH₂analogs) can be attributed to the fact that they are not stronglybasic, a property required for accumulation in the malaria parasiteacidic food vacuole in which hemozoin formation takes place (12).

The ability of 9-anilinoacridines to interact with hematin wasinvestigated by the continuous variation technique (Job's plot),as described in Materials and Methods. A solution of hematinat pH 7.4 showed a sharp peak at 400 nm, indicating that monomerichematin predominated under the in vitro conditions used (40%DMSO, sodium phosphate buffer [pH 7.4]). The addition of chloroquineand the 9-anilinoacridine analogs that inhibited ß-hematinformation (compounds 1, 2, 3, 9, 10, 11, 13, and 14) resultedin the reduction in the Soret absorption band of hematin, indicativeof an association of the compounds with hematin. Plots betweenthe difference in the expected hematin absorbance at 400 nmand the experimental value versus the drug:hematin molar ratiowere constructed for the drugs tested. Changes in absorbanceintensity were maximal when the molar fraction of hematin tochloroquine and to compounds 9, 10, and 13 was 2:1 (see Fig. 2for representative plots). On the other hand, compounds 1, 2,3, 11, and 14 produced only minor changes to the hematin Soretband (data not shown), indicating that these 9-anilinoacridineanalogs interacted weakly with monomeric hematin under the experimentalconditions used. Although the numbers of compounds were limited,those compounds with low IC₅₀s for inhibition of ß-hematinformation formed stable complexes with hematin, whereas 9-anilinoacridineanalogs with higher IC₅₀s formed weaker complexes with hematin (althoughthere was considerable overlap between the two groups). There wereno apparent structural features that could explain the stabilities ofthe drug-hematin complexes. As mentioned above, compound 13was a good inhibitor of ß-hematin formation and formeda stable complex with hematin, but it showed poor in vitro antimalarialactivity on account of its lack of a basic side chain to assistwith accumulation in the food vacuole. On the other hand, compound3 showed a poorer ability to inhibit ß-hematin formationand formed a weak complex with hematin, but nevertheless, itwas a better antimalarial drug on account of its additionalproperty of being an inhibitor of DNA topoisomerase II.

View larger version (13K):
[in this window]
[in a new window]
FIG. 2. Job's plot of chloroquine (A) and 1'-NHSO₂CH₃-3,6-diCl-9-anilinoacridine (B) binding to hematin. The total concentration of the two components was 10 µM in 40% aqueous DMSO, with mole fractions varying from 0 to 1. The absorbance at 400 nm was measured at 25°C.

Chou and Fitch (5) have previously demonstrated that a 0.5%suspension of washed, normal mouse erythrocytes is lysed inthe presence of hematin by a colloid-osmotic mechanism. Fiftypercent hemolysis, obtained with 4 µM hematin, is enhancedto completion in the presence of 5 to 25 µM chloroquine(4). In this study, a 0.03% suspension of human red blood cellswas used, and 50% hemolysis was also seen with 4 µM hematin,with enhancement to completion by chloroquine over a concentrationrange of 0.1 to 20 µM. When chloroquine was replaced withthe 9-anilinoacridine analogs that inhibited ß-hematinformation, enhancement of hematin-induced hemolysis could beobserved, but over a more limited concentration range (0.1 to4 µM). With the exception of compounds 1 and 9, whichbehaved similarly to chloroquine, at the higher concentrationsthe drugs inhibited hematin-induced red cell lysis. It has beensuggested that the hematin-chloroquine complex is able to enhancehematin-induced membrane lysis by increasing the efficiencywith which hematin interdigitates into the bilayer (18). Thus,a possible explanation for the inhibitory phenomena of 9-anilinoacridnesis that above a critical (nonphysiological) drug concentrationcomplexation of drug with hematin may limit the access of thelatter to the membrane.

In order to prolong the useful lifetimes of new antimalarials,which not only are expensive to develop but also are few innumber, it has been recommended that new drugs should be usedin the field as drug combinations to delay the rapid onset ofparasite resistance (21). Toward this goal, we have taken anovel approach by demonstrating the possibility of developing9-anilinoacridine derivatives that target two different sitesor functions within the malaria parasite, namely, DNA topoisomeraseII and ß-hematin formation. Although this initialstudy involved only a limited number of compounds, it has providedstructure-activity relationships that are well worth studying further.

ACKNOWLEDGMENTS

S.A. gratefully acknowledges support from the Haugland Foundationin the United States, and P.W. is a Senior Research Scholarof the Thailand Research Fund.

FOOTNOTES

* Corresponding author. Mailing address: Department of Biochemistry, Faculty of Science, Mahidol University, Rama 6 Rd., Bangkok 10400, Thailand. Phone: (662) 245-5195. Fax: (662) 248-0375. E-mail: scpwl{at}mahidol.ac.th.

REFERENCES

Top