Antiplasmodial Chalcones Inhibit Sorbitol-Induced Hemolysis of Plasmodium falciparum-Infected Erythrocytes

Abstract

A series of alkoxylated and hydroxylated chalcones previously reported to have antiplasmodial activities in vitro were investigated for their effects on the new permeation pathways induced by the malaria parasite in the host erythrocyte membrane. Of 21 compounds with good antiplasmodial activities (50% inhibitory concentrations [IC₅₀s], ≤20 μM), 8 members were found to inhibit sorbitol-induced lysis of parasitized erythrocytes to a significant extent (≤40% of control values) at a concentration (10 μM) that was close to their antiplasmodial IC₅₀s. Qualitative structure-activity analysis suggested that activity was governed to a greater extent by a substitution on ring B than on ring A of the chalcone template. Most of the active compounds had methoxy or dimethoxy groups on ring B. Considerable variety was permitted on ring A in terms of the electron-donating or -withdrawing property. Lipophilicity did not appear to be an important determinant for activity. Although they are not exceptionally potent as inhibitors (lowest IC₅₀, 1.9 μM), the chalcones compare favorably with other more potent inhibitors in terms of their selective toxicities against plasmodia and their neutral character.

It is widely recognized that the intracellular malarial parasite induces in the host erythrocyte membrane new permeation pathways that are absent from the membrane of the uninfected erythrocytes (8, 9, 11, 13). These pathways have the characteristics of anion-selective channels (6, 11) and are permeable to a wide range of chemically different solutes, amino acids (3), nucleosides (22), sugars (10), and inorganic and organic ions (13, 21), many of which are essential nutrients for the survival of the parasite. In view of their selective presence in parasitized cells and their likely role in nutrient acquisition for the intracellular parasite, the parasite-induced channels are considered an attractive chemotherapeutic target (6, 8, 12). A range of structurally diverse compounds have been found to inhibit these parasite-induced channels, as revealed by their inhibition of the transport of small solutes (choline, sorbitol, chloride ion, threonine) into parasitized erythrocytes. These include the bioflavonoid glycoside phlorizin (15), sulfonyl ureas (glibenclamide, tolbutamide) (14), several arylaminobenzoates (12, 20), and cinnamic acid derivatives (10). Many of these compounds have antiplasmodial activity, and for some, like phlorizin, the 50% inhibitory concentrations (IC₅₀) for inhibition of parasite growth (16 ± 7 μM) and uptake of solutes (17 ± 2 μM) are closely aligned (15). For others, like sulfonylureas and arylaminobenzoates, there is a disparity between the concentrations required for antiplasmodial activity and those required for inhibition of solute uptake (12, 14, 20). This has been attributed to the anionic character of these acidic compounds, which promotes interaction with the serum proteins present in the parasite growth medium used for in vitro antiplasmodial activity evaluation. The general consensus is that the antimalarial potential of compounds that selectively inhibit parasite-induced channels can be realized only if they do not bind concurrently to serum components to any significant extent (8, 12).

Interest in the antimalarial activities of chalcones was prompted by the discovery of the antiplasmodial activity of licochalcone A, an oxygenated chalcone found in the roots of the Chinese licorice during routine screening (4). Licochalcone A was subsequently found to possess antileishmanial activity (24), possibly through the selective inhibition of fumarate reductase in the respiratory chain of the parasite (5). At about the same time, a separate computational approach identified chalcones as potential plasmodial cysteine protease inhibitors (16). However, subsequent investigations showed that antiplasmodial activity was not necessarily correlated with the inhibition of malarial cysteine protease (7). Phloretin, the aglycone of the bioflavonoid glycoside phlorizin, was almost as effective as phlorizin in inhibiting sorbitol transport in Plasmodium-infected cells under certain experimental conditions (15). Given their structural resemblance to phloretin (a dihydrochalcone), it was considered of interest to explore the effects of chalcones on the parasite-induced permeation pathways. In the study described here, we investigated the inhibition of the pathways by a series of oxygenated chalcones that have antiplasmodial activities (18, 19)

MATERIALS AND METHODS

The synthesis and antiplasmodial activities of the chalcones investigated in this study have been reported previously (18, 19). The following chemicals were purchased from Sigma Chemical Co. (St. Louis, Mo.): chloroquine diphosphate, furosemide (sodium salt), and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT). Other reagents were of analytical grade.

Effects of compounds on sorbitol-induced hemolysis of P. falciparum-infected erythrocytes.

The effects of the compounds of interest on the parasite-induced permeation pathways were investigated by a sorbitol hemolysis method similar to that described previously (12). Chloroquine-resistant P. falciparum FAF-6 was cultured; and trophozoite-stage infected erythrocytes (approximately 36 to 44 h postinvasion, 20% parasitemia) were harvested by centrifugation (600 × g, 5 min, 25°C), washed twice with a solution of NaCl (150 mM)-HEPES (20 mM) (pH 7.4; 304 mosM), and suspended in the same solution to give a hematocrit of about 50%. In initial experiments, a single concentration (10 μM) of each compound was tested for its ability to inhibit sorbitol-induced hemolysis. Test compounds were dissolved in dimethyl sulfoxide (DMSO) to give 10 mM solutions, which were diluted further with a solution comprising 300 mM sorbitol-20 mM HEPES (pH 7.4; 345 mosM) to give stock solutions of 0.1 mM. A total of 100 μl of this solution was combined with an additional 800 μl of the same sorbitol-HEPES buffer, together with 100 μl of the infected cell suspension in NaCl-HEPES. The mixture was incubated at 37°C for 15 min and centrifuged, and aliquots of the supernatant (100 μl) were dispensed into 96-well plates. The absorbances at 540 nm were read to estimate the amount of hemoglobin released and, hence, the degree of sorbitol-induced hemolysis. The absorbances of the control wells containing only test compounds in sorbitol-HEPES solution were determined concurrently, and corrections to the absorbances measured in the test wells were made when necessary.

The IC₅₀ of the test compounds that at the initial test concentration of 10 μM inhibited sorbitol-induced hemolysis of infected erythrocytes by ≥80% were determined in the following manner. The hemolysis of infected cells in sorbitol-HEPES in the absence of test compound (control) was determined at four time intervals (2.5, 5, 10, and 15 min). A larger volume (∼4.5 ml) of the incubation mixture was used, and aliquots (1 ml) were withdrawn at the prescribed times for determination of the absorbance. In most experiments, absorbance (reflecting the extent of hemolysis) reached a maximum at between 10 and 12 min and remained constant thereafter. Thus, the absorbance determined at the 15-min time point was arbitrarily taken as the maximum, and the absorbance determined at the other time intervals was expressed as a percentage of this maximum value to give the percent hemolysis. A plot of the percent hemolysis against time gave a sigmoidal curve, from which the gradient of the straight-line portion of the curve was determined and taken to represent the rate of hemolysis of the infected cells under control conditions. At the same time as the control determinations were made, similar experiments were carried out with fixed concentrations of the test compound. Absorbances were determined at the same time points and in each case were expressed as a percentage of the maximum absorbance (at the 15-min time point) in the control experiment. Percent hemolysis was plotted against time for each concentration of test compound, and the straight-line portion of the curve was measured to give the rate of hemolysis for that concentration of test compound. The rate was then expressed as a percentage of the control rate (i.e., for infected cells in sorbitol-HEPES alone) determined concurrently. It was not possible to determine the rates of lysis at different concentrations of the same test compound (at least three concentrations were required for IC₅₀ determinations, with no less than duplicate determinations for each concentration) on the same day, and expression of the rates as a percentage of the control rate was one way to accommodate the variations in the readings. A plot of the percentage of the control rate observed at different concentrations of the test compound was then made, with a value of 100% being equivalent to the control rate. A straight line with a negative gradient was observed, and the concentration of test compound required to reduce the control gradient by 50% was designated the IC₅₀.

Evaluation of selectivity ratio.

The selectivity ratio compares the effect of the chalcone of interest on the viability of a mammalian cell line as opposed to its effect on Plasmodium. It is obtained from the following expression: IC₅₀ for cytotoxicity against mammalian cells/IC₅₀ for inhibition of hypoxanthine uptake into Plasmodium.

IC₅₀ for cytotoxicity were evaluated with a human epidermoid cancer cell line (KB3-1 cells) by the microculture tetrazolium assay described by Alley et al. (1), with some modifications. The drug-sensitive cell line KB3-1 has intermediate sensitivity to cytotoxic agents and is recommended for use in the evaluation of the selectivities of antiplasmodial drugs (2). IC₅₀ for inhibition of hypoxanthine uptake were reported previously (18, 19).

KB3-1 cells (a gift from Caroline Lee, Department of Biochemistry, National University of Singapore) were cultured to 60 to 70% confluence in tissue culture flasks containing Dulbecco's modified Eagle's medium (supplemented with 10% fetal bovine serum) at 37°C in a humidified atmosphere with 5% CO₂. On the attainment of confluence, the cells were trypsinized and aliquots (100 μl) of cells in growth medium were transferred to 96-well tissue culture plates at densities of 2 × 10⁴ to 5 × 10⁴ cells/ml. After 24 h of incubation, the cells were attached to the wells and could be used for test procedures. The cells remained in the exponential growth phase during this period. Stock solutions of test compounds were prepared in DMSO (2 to 4 mM), and serial dilutions were made so that the final sample concentration in each well (total volume, 200 μl; maximum amount of DMSO, 1% [vol/vol]) was at least 10-fold higher or lower than the expected IC₅₀. The test sample was incubated with the cells for 15 h under the usual conditions, with triplicate determinations for each concentration. Wells without cells and those with cells under control conditions (i.e., in culture medium alone) were examined in parallel. At the end of the incubation period, the medium was decanted and the cells were washed carefully with phosphate buffer solution, after which an aliquot of MTT (100 μl at 0.5 mg/ml in Hanks balanced salt solution-HEPES buffer [pH 7.4]) was added to each well. The cells were incubated for a further 3 h, after which the medium was removed from each well by careful pipetting. DMSO (150 μl) was added to lyse the cells and dissolve the purple formazan crystals. After 10 min the absorbance of the formazan product at 590 nm was measured with a microtiter plate reader. The absorbance values obtained at each concentration were averaged, adjusted by subtraction of the values for the blank wells, and expressed as a percentage of the average absorbance obtained from control incubations (in the absence of sample). IC₅₀ were determined from logarithmic plots of the percent absorbance versus concentration. Each compound was assayed twice, with at least three determinations made for each concentration.

RESULTS AND DISCUSSION

The series of chalcones used in the present study were divided into two groups, depending on the substitution pattern of ring B, namely, those with alkoxy substituents (2,3,4-trimethoxy, 2,4-dimethoxy, 4-methoxy, and 4-ethoxy) and those with hydroxyl substituents (2,4-dihydroxy, 4-hydroxy, and 2-hydroxy). Compounds with the highest levels of antiplasmodial activity were found in the alkoxy series, with inhibition of hypoxanthine uptake into infected erythrocytes observed at IC₅₀ of ≤10 μM for the most active members (19). The hydroxychalcones were about twofold less potent. In this study, 32 chalcones were tested for their effects on the altered permeability of the infected erythrocytes by a sorbitol-induced lysis method described previously (12). Included in this sample were the active antiplasmodial alkoxychalcones (n = 12), with IC₅₀ ≤10 μM, and hydroxychalcones (n = 9), with IC₅₀ ≤20 μM (19).

The erythrocyte membrane is normally impermeant to sorbitol, but the channels induced by the malaria parasite in the infected cell membrane allow the passage of sorbitol and when mature parasitized erythrocytes are suspended in an isosmotic sorbitol solution, there is a net uptake of sorbitol and water into the erythrocyte, resulting in cell swelling and hemolysis (13). Hemolysis was monitored spectrophotometrically by measuring the absorbance of hemoglobin at 540 nm. The rate of hemolysis gives a semiquantitative estimate of the net rate of influx of sorbitol, albeit under nonphysiological conditions. Hemolysis is reduced in the presence of channel inhibitors (lower absorbance), and monitoring of the rates of hemolysis in the presence of various inhibitors is a convenient means of evaluating the relative potencies of these substances.

Table 1 lists the chalcones that were tested for their abilities to inhibit the parasite-induced pathways. When these compounds were initially tested at 100 μM (data not shown), almost all of them (in particular, the 21 active compounds) impeded sorbitol-induced hemolysis by more than 40% relative to the control, which made it difficult to distinguish their relative potencies. Therefore, the experiments were repeated with the drugs at 10 μM, a concentration that was closer to the antiplasmodial IC₅₀ of the active compounds. Furosemide, an effective inhibitor of the parasite-induced channels (13), was used as a positive control, and at 10 μM it reduced the sorbitol-induced hemolysis to 56% of that observed in the absence of inhibitor. No established antimalarial agents (quinine, chloroquine) inhibit the channels in the therapeutic concentration range.

TABLE 1.

Selectivity ratios, lipophilicities, and percent inhibition of sorbitol-induced hemolysis of P. falciparum-infected erythrocytes

^aCompound numbers in boldface are active antiplasmodial agents (IC₅₀s, <10 μM for alkoxylated chalcones and ≤20 μM for hydroxylated chalcones). Compound numbers that are italicized are good inhibitors of sorbitol-induced lysis (≤40% of that for the control). Compounds numbers that are boldface and italicized have good antiplasmodial and transport-inhibiting activities.

^bRing A is a heterocylic or naphthalene ring.

^cInhibition of [³H]hypoxanthine incorporation by P. falciparum K1-infected erythrocytes (18,19).

^dCell viability of 122% of that for the control for compound 113 at 20 μM and 96% of that for the control for compound 207 at 40 μM.

^eExpressed in terms of log k, the capacity factor of the compound at pH 7.0, as determined by reversed-phase high-performance liquid chromatography (19). The log of the capacity factor increases as lipophilicity increases.

^fValues are means ± standard deviation of three or more determinations. The parasitized cells were exposed to the sorbitol solution (with or without the compounds of interest) for 15 min before sampling. Compounds associated with ≤40% hemolysis were considered good inhibitors of the parasite-induced channels.

^gIC₅₀ of 5 μM for the inhibition of induced choline flux in parasitized cells (13).

^hSelectivity ratio. = IC₅₀ for KB3-1 cells/IC₉₀ for Plasmodium.

ⁱND, not done.

Chalcones that reduced sorbitol-induced hemolysis to ≤40% of the control value were arbitrarily classified as “good” inhibitors, and 10 compounds were classified in this way. It is significant that eight of these good inhibitors were concurrently identified as active antiplasmodial agents in their respective classes, on the basis of their IC₅₀. The exceptions were dimethoxychalcone (compound 102; 17% hemolysis at 10 μM versus an IC₅₀ of 129 μM) and 4-hydroxychalcone (compound 226; 15% hemolysis at 10 μM versus an IC₅₀ of 21.7 μM). If compound 226 is considered a marginal outlier, then 9 of 10 good inhibitors of the parasite-induced channels are active antiplasmodial agents by the present definition. These compounds had negligible effects on the degradation of methemoglobin by crude Plasmodium extracts and did not inhibit malarial cysteine protease falcipain 2 or interfere with heme binding (17). However, although antiplasmodial activity is better correlated to the inhibition of parasite-induced channels than to the other routes of inhibition investigated, we cannot exclude the possibility that the chalcones have alternative targets within the infected cell.

The structural distributions of chalcones that were good inhibitors of sorbitol-induced hemolysis revealed that a large number of them (7 of 10) are alkoxylated chalcones, namely, dimethoxy- and methoxychalcones. No good inhibitor was found among the trimethoxy- or ethoxychalcones, even though the trimethoxychalcones are well represented among the active antiplasmodial agents. It would appear that most good inhibitors of sorbitol-induced hemolysis are also active antiplasmodial agents, but not all active antiplasmodial chalcones inhibit sorbitol-induced hemolysis. Of the 21 active members identified in Table 1, more than half did not impede sorbitol-induced hemolysis to a significant extent (≤40% of that for the control) at 10 μM. An earlier structure-activity relationship study (19) showed that the alkoxylated and hydroxylated chalcones in Table 1 have distinct physicochemical requirements (lipophilicity, charge distribution, volume, and area) for antiplasmodial activity. Different physicochemical requirements were observed even among the alkoxylated chalcones (the trimethoxy-, dimethoxy-, and methoxychalcones). Thus, it is plausible that structurally different classes of chalcones have different modes of antiplasmodial activity. Dimethoxy and methoxychalcones (which were strongly represented among the good inhibitors of sorbitol-induced hemolysis) may have the right features to promote an interaction with the parasite-induced channels. As to what these features might be, the small number of compounds available for comparison allows only a tentative structure-activity relationship to be proposed. (i) Inhibition is sensitive to the number of groups on ring B, as seen from the lack of good inhibitors among the 2,3,4-trimethoxychalcones. (ii) Alkoxy substituents are preferred to hydroxy substituents on ring B, since only 3 of the 10 good inhibitors were hydroxylated chalcones. (iii) Ring A can be substituted with electron-donating groups (hydroxy, methoxy, methyl, or ethyl groups) or electron-withdrawing groups (fluoro groups). The fluoro substituent is encountered most frequently among the active inhibitors. (iv) Lipophilicity does not appear to be an important factor for inhibition. Most of the active compounds have lipophilicities (determined from high-pressure liquid chromatography-derived capacity factors) in the range of 3.3 to 3.6, but as many compounds with lipophilicities in the same range do not have inhibitory activities (Table 1). Previous investigations have generally emphasized the importance of lipophilicity for good inhibitory activity (6, 12). (v) Ring A can be pyridine or quinoline. In the case of quinoline, there appears to be a preference for 3-quinolinyl (compound 27) instead of 4-quinolinyl (compound 28).

Compared to other known inhibitors of the parasite-induced channels, the present series of chalcones are not exceptionally potent. There is a possibility that the metabolites of chalcones act synergistically with the parent substance. Zhang et al. (25) reported on a dihydrochalcone as the major metabolite of an experimental chalcone in an in vitro investigation with human liver microsomes. If a similar biotransformation occurs with the present series of chalcones, the dihydrochalcone metabolites are likely inhibitors of the parasite-induced channels, since a similar activity has been reported for phloretin, a dihydrochalcone (15).

To date, the most potent inhibitors of the parasite-induced channels are found among the substituted amino benzoic acids, some of which have been shown to inhibit parasite-induced choline transport with IC₅₀ in the nanomolar range (12, 20). These compounds inhibited chloride channels in mammalian cells to various degrees, although some members block the parasite-induced pathway without significantly affecting anion transport mechanisms in host tissues (12). A more serious limitation of these compounds is their anionic character at physiological pH, which promotes binding to plasma proteins (23). This factor may account for the poor correlation between antiplasmodial activities in vitro (normally determined in growth medium with 10% serum proteins) and inhibition of parasite-induced transport (determined in physiological saline). By contrast, the present series of chalcones are not acidic and, because they lack an anionic character, would not be expected to bind to plasma proteins. Moreover, the selectivity ratios of some chalcones indicated that they discriminated between Plasmodium and mammalian cells (Table 1). To the best of our knowledge, there is no reference in the literature to chalcones that inhibit chloride channels or other anion transport pathways.

In summary, we have shown that chalcones inhibited the new permeation pathways induced by the parasite in the host erythrocyte membrane. A significant number of chalcones that were good inhibitors of sorbitol-induced hemolysis were concurrently identified as active antiplasmodial agents. Most of these are dimethoxy- and methoxychalcones. However, more than half of the chalcones with good antiplasmodial activities did not inhibit the parasite-induced channels to a significant degree. This observation reinforces the point that structurally different chalcones may exert their antiplasmodial activities by different routes or target additional pathways in the parasitized cell. Compared to other known inhibitors of the parasite-induced channels, the chalcones are notable in two respects: selective toxicity against Plasmodium and the absence of an anionic character, which would render these compounds less likely to bind to plasma proteins. Therefore, they are promising lead compounds with activities against the parasite-induced channels.