Novel Compounds Active against Leishmania major

Abstract

Leishmania major is an important trypanosomatid pathogen that causes leishmaniasis, which is a serious disease in much of the Old World. Current treatments include a small number of antimony compounds that, while somewhat effective, are limited by serious side effects. We have screened a small portion of a unique chemical library and have found at least three novel compounds that are effective against L. tarentolae and L. major in vitro and in a murine macrophage model of L. major infection. These compounds were effective in both assays at doses significantly lower than those of sodium stibogluconate (Pentostam) and represent possible candidates for drug development.

Leishmania is a trypanosomatid protozoan that is transmitted by phlebotomine sand flies and that causes leishmaniasis (14). Like many other tropical diseases, the leishmaniases are related to economic development and human-made environmental changes that result in increased exposure to the sand fly (7). The disease assumed importance in the United States when many Operation Desert Storm veterans were exposed to sand flies and were afflicted with leishmaniasis (10). The overlapping geographical distribution of human immunodeficiency virus (HIV) infection-AIDS and leishmaniasis is increasing due to the spread of the disease from rural into urban areas (1). Leishmania-HIV coinfection is regarded as an “emerging” infectious disease, for in certain countries up to 70% of adult cases of leishmaniasis are related to HIV infection-AIDS.

The primary treatment of leishmaniasis has been pentavalent antimony complexed to carbohydrate in the form of sodium stibogluconate (Pentostam) (2, 9) or meglumine antimoniate (Glucantime) (3). These compounds have been in use since 1940, but their mode(s) of action is still not known. Severe reactions, including death, occur in 10% of those who are treated (20).

Amphotericin B and pentamidine are second-line drugs used for the treatment of leishmaniasis (16); however, as is well known, amphotericin B and pentamidine are both toxic. Encouragingly, a new drug, miltefosine (Impavido) (18), which is a signaling pathway inhibitor, has recently been approved for the treatment of leishmaniasis in India. However, experience with antibiotics, including antibacterials, antifungals, and antivirals, indicates that resistance to the currently used drugs is the rule rather than the exception; this necessitates the continued search for new drugs.

The Developmental Therapeutics Program (DTP) of the National Cancer Institute (NCI) maintains a repository of synthetic and natural products. These compounds have been screened in the DTP anticancer assays, and thus, the cytotoxicity of each compound can be determined by searching the NCI database.

We have screened ∼15,000 of these compounds for their activities against, first, Leishmania tarentolae, taking advantage of its ease of growth in vitro. Compounds that were active against L. tarentolae promastigotes but that were cytotoxic or that contained chemical structures not suitable for drug development were eliminated. The remaining compounds were then tested for their activities against Leishmania major promastigotes in vitro and, finally, for their activities against L. major amastigotes in a murine macrophage model. We have found three novel compounds that were active against L. major in vitro and in macrophages.

MATERIALS AND METHODS

Chemicals.

Chemicals and antibiotics were purchased from Sigma-Aldrich Inc. (St. Louis, MO). Brain heart infusion medium was purchased from Difco (Detroit, MI). RPMI 1640 tissue culture medium (with glutamine, without bicarbonate, and with pH indicator) and Schneider's Drosophila insect medium (with glutamine) were purchased from Invitrogen (Carlsbad, CA).

Culture conditions.

L. tarentolae was obtained from Robert E. Nelson of the University of California, Los Angeles. Working stocks of L. tarentolae cells were prepared by inoculating 1.5 × 10⁶ cells/ml from glycerol stocks into 10 ml of brain heart infusion medium plus hemin (0.005% [wt/vol]) (BHIH) in tissue culture flasks (25 cm²; treated nonpyrogenic polystyrene; Corning, Inc.). Cultures were incubated without agitation at 24°C for 3 to 4 days. After every 4 to 5 days of incubation, 9 ml was removed and replaced with 9 ml of fresh BHIH.

Working stocks of L. major strain LV39 were prepared by inoculating 1.5 × 10⁶ cells/ml from glycerol stocks into 10 ml of Schneider's insect medium with fetal bovine serum (1% [vol/vol]) and hemin (0.005% [wt/vol]) in tissue culture flasks. The cultures were incubated at 24°C for 4 days. Every 5 to 6 days, 9 ml was removed and replaced with 9 ml of fresh medium.

Compound source.

Approximately 15,000 compounds were obtained frozen in 96-well microtiter plates from DTP (Bethesda, MD). Each plate contained 80 compounds dissolved in 100% dimethyl sulfoxide (DMSO) at either 1 mM or 10 mM.

Compound screening assays.

L. tarentolae promastigotes were inoculated into fresh BHIH at 1 × 10⁶ cells/ml in sterile 96-well microtiter plates (200 μl per well). Approximately 0.5 μl of each compound was added to each well by using a 96-prong metal replicator. Controls included DMSO only, no additions, 0.01% (wt/vol) sodium azide (final concentration), and 16 μg/ml amphotericin B (final concentration). The plates were incubated at 24°C for 72 h, and the A₆₀₀ of each well was determined with a SpectraMax 340 plate reader (Molecular Devices). The A₆₀₀ of each well was compared to the A₆₀₀ of the no-drug control wells, and the percent growth inhibition was calculated.

Antifungal assays.

An overnight culture of Candida albicans (ATTC 90028) was grown in 50 ml culture of Sabouraud dextrose broth medium with shaking at 140 rpm at 35°C. Cell numbers were determined by using a hemocytometer, and RPMI 1640 medium was inoculated at a final cell concentration of 1 × 10⁴ cells/ml in sterile 96-well microtiter plates (200 μl per well). A glycerol stock of Aspergillus fumigatus (ATTC 16424) conidia was thawed; the conidia were counted using a hemocytometer and were used to inoculate RPMI 1640 medium at a final cell concentration of 1 × 10⁵ conidia/ml in sterile 96-well microtiter plates (200 μl per well). Two microliters of each compound (0.010 μM; final concentration) was added to individual wells. Amphotericin B (0.05 μM; final concentration), 2 μl sterile DMSO, and no additions were used as controls. The plates were incubated for 24 h at 35°C, and the amount of growth in each well was determined visually by comparison to the growth in the control wells.

Antibacterial assays.

An overnight culture of Bacillus subtilis (ATTC 11774) was grown overnight in nutrient broth with shaking at 37°C. The optical density at 600 nm (OD₆₀₀) of the resulting culture was 0.86. The culture was diluted 1:1,000 with fresh medium and was used directly in sterile 96-well microtiter plates (200 μl per well). An overnight culture of Escherichia coli DH5α cells was grown in Luria-Bertani medium (10% [wt/vol] tryptone, 5% [wt/vol] yeast extract, 10% [wt/vol] NaCl) with shaking at 140 rpm at 35°C. The OD₆₀₀ was determined, and the culture was diluted to result in an OD₆₀₀ of 0.1_. This was used directly in sterile 96-well microtiter plates (200 μl per well). Two microliters of each compound (0.010 μM; final concentration) was added to individual wells. Ampicillin B (0.05 μM; final concentration), 2 μl sterile DMSO, and wells with no drug were used as controls. The plates were incubated at 37°C for 18 h. Inhibition of growth was determined visually by comparison to the growth in the control wells.

MIC.

Stock cultures of L. tarentolae were used to inoculate fresh BHIH at a final concentration of 1 × 10⁶ cells/ml in 96-well microtiter plates (200 μl per well). Serial dilutions of each active compound were added, beginning with a concentration of 0.125 μM and ending with a 1:2,048 dilution of this concentration. Medium only and cells with medium with or without DMSO were used as controls. The plates were incubated at 25°C for 3 days. Inhibition of growth was determined visually by comparison to the growth in the control wells.

This assay was also performed by using Leishmania major promastigotes, as described above, with the following modifications: stock cultures were used to inoculate Schneider's insect medium plus hemin (0.005% [wt/vol]) and fetal calf serum (1% [wt/vol]) at a final cell concentration of 1 × 10⁶ cells/ml in 96-well plates (200 μl per well).

Amastigote macrophage assay.

The technique of Titus et al. (19) was used with 2 × 10⁵ peritoneal exudate cell macrophages and 1.25 × 10⁶ L. major cells per well. Triplicate cultures were treated with 1 μl of various concentrations of each compound in DMSO (see Table 1 for the concentrations tested). Controls included medium with or without DMSO as well as Pentostam at final concentrations of 75 μg/ml or 50 μg/ml. The coverslips were removed from the 24-well dishes 3 days later, rinsed in 1× phosphate-buffered saline, fixed (J322A-1), stained (J322A-2), and counterstained (J322A-3) (Jorgensen Labs, Inc.) The coverslips were allowed to dry completely before they were mounted on frosted microscope slides with Vectamount (Vector Labs).

TABLE 1.

Summary of MICs against L. tarentolae and L. major^a

NSC compound no.	MIC (μg/ml)		Concn (μg/ml) tested with macrophages	Efficacy in macrophage model
	L. tarentolae	L. major
377	0.015	0.25	0.03	−
4972	0.2	1	0.4	−
687330	20	5	2	−
108944	11	5.5	1.1	−
163443	15.5	8	1.6	−
635975	8	8	0.8	−
165883	0.5	9	0.9	−
46613	9.5	9.5	1	−
63984	10	10	1	−
634650	10	10	1	−
112789	11	10.5	1.1	−
150169	10.5	10.5	1.1	−
97703	11	11	1.1	−
687667	11	11	1.1	−
168466	2.5	11	1.1	−
150554	1.5	11.5	1.2	−
50155	3	11.5	1.2	−
58319	3	12	1.2	−
128305	12.5	12.5	1.3	−
648422	109	13	1.3	−
89110	3	13	6.5	−
369294	109	13.5	11	−
647613	14	14	1.4	−
57584	14	14	1.4	−
281245	15.5	15.5	1.6	−
62791	16	16	1.6	−
176555	132	16.5	13	−
99504	67	17	13	−
273829	17.5	17.5	1.8	−
63984	10	20	1	−
13726	6.5	27	13.6	−
118770	7	27	11	−
7962	1.5	28.5	3.5	−
83633	29	29	12	+
62791	16	31.5	1.6	−
32673	0.13	33	0.3	−
56939	0.3	34.5	0.5	−
69298	5	38	9.5	−
281668	37	74	7.4	−
1199	11	90	9	−
351520	6.5	106	10.6	+
13512	29.5	117	12	+
35761	70.5	140	7	−
128164	38	152	15	−
73109	156	156	15.6	−
Pentostam	Not tested	250	75	+

^aThe MIC values against L. tarentolae and L. major were determined as described in Materials and Methods. In addition, the efficacy of each compound tested at the indicated concentrations in the macrophage infection model was determined as described in the Materials and Methods (+, activity; −, no activity).

Microscopy.

The numbers of amastigotes per 100 macrophages and the numbers of macrophages infected with at least one amastigote in compound-treated and untreated cells were determined by visual examination under ×400 magnification by using bright-field microscopy. Digital photographs of treated and untreated, fixed, mounted cells were obtained by using a Zeiss AuxioVert 200 M microscope equipped with a Sensicam camera, and the images were processed by using Intelligent Imagining Innovation software.

RESULTS

Screening of compounds for activity against Leishmania.

Approximately 15,000 compounds were screened for in vitro inhibition of growth of L. tarentolae, as described in the Materials and Methods. Since we wished to discover compounds that were active against several Leishmania species, we first screened compounds for their activities against L. tarentolae, taking advantage of its ease of manipulation and rapid growth. Seven hundred eighty-six compounds inhibited the growth of L. tarentolae promastigotes by >90% (results not shown). We searched the NCI database (http://dtp.nci.nih.gov/docs/dtp_search.html) to eliminate compounds that were cytotoxic (50% lethal concentration, <30 μM) or that had structures or active moieties that would preclude their use as drugs. In total, 663 compounds were eliminated.

Testing for antifungal and antibacterial activities.

The remaining 123 compounds were screened for their antifungal and antibacterial activities as described in the Materials and Methods. Seven compounds had activities against either C. albicans or A. fumigatus and were eliminated. Compounds were then tested for their effects on the growth of B. subtilis and E. coli. None of the compounds was active against E. coli, but 22 were eliminated due to activity against B. subtilis.

Determination of potency.

Thus, there remained 94 compounds that were active against L. tarentolae but that were not cytotoxic and that did not have antifungal or antibacterial activity. We determined the potency of each compound against L. tarentolae as described in the Materials and Methods. Of these, we chose 43 compounds that had MICs against L. tarentolae less than ∼155 μg/ml for testing against L. major. These results are presented in Table 1. Note that in 16 cases, the MICs against both species were similar, while in 7 cases the MIC against L. major was lower than that against L. tarentolae, and in the remaining 20 cases, the MIC was higher. The reason(s) for these differences and similarities is unknown but may reflect differences between these species.

Testing of each compound for efficacy by using a murine macrophage model.

Murine macrophages were infected with L. major as described in the Materials and Methods and treated with each of the 43 compounds that had activity against L. major in vitro. Based on previous experience (4, 5, 6, 8, 11, 12, 13, 17) with Pentostam, we chose to use concentrations for testing in the macrophage amastigote model that were significantly lower than those used for testing against promastigotes in vitro. Although most compounds were marginally active in this assay, three compounds (NSC compounds 83633, 351520, and 13512) showed significant efficacies (Table 1).

After the initial screen in the macrophage model, we tested each of the three compounds (NSC compounds 83633, 351520, and 13512) at three concentrations and monitored the number of infected macrophages and the number of amastigotes per 100 macrophages as two parameters for efficacy. Representative photomicrographs of infected, nontreated cells and of infected, treated cells are shown in Fig. 1A and B, respectively. Note the complete absence of amastigotes in the treated cells.

FIG. 1.

Macrophages infected with L. major. Murine macrophages were obtained and infected with L. major as described in the Materials and Methods. After treatment with DMSO only (A) or with NCI compound 83633 (B), the cells were fixed and mounted. Digital microphotographs were obtained as described in the text. The arrows in panel A point to intracellular amastigotes. Magnifications, ×200.

For the untreated control cells, ∼20% of macrophages were infected with an average of ∼60 amastigotes per 100 cells (Fig. 2A and B). In contrast, treatment with Pentostam (75 μg/ml and 50 μg/ml; [based on μg of sodium stibogluconate/ml]), as expected, significantly reduced the number of infected macrophages (Fig. 2A) and the number of amastigotes per 100 cells (Fig. 2B) to 1.2% and 2%, respectively, at a concentration of 75 μg/ml and 8.2% and 14%, respectively, at a concentration of 50 μg/ml. Critically, compounds 351520, 13512, and 83633 each significantly reduced (P < 0.05 in comparison to the results for the control) the number of infected cells and the number of amastigotes per 100 cells (Fig. 2A and B).

FIG. 2.

Treatment of L. major-infected macrophages with Pentostam and NCI compounds. Macrophages were obtained, infected with L. major promastigotes, and treated with NCI compound 83633, 351520, or 13512 or Pentostam, as described in the text. The numbers of intracellular amastigotes per 100 macrophages (A) and the numbers of macrophages infected with at least one amastigote (B) were determined by visual examination of fixed, mounted specimens. The concentrations of each compound are indicated. The values represent the averages of three separate determinations (except for Pentostam at 50 μg/ml, which had one determination [and, therefore, a P value could not be determined]), with the error bars indicating the range of values. *, values significantly different (P < 0.05) from those for the controls with DMSO only, as determined by Student's t test.

Structures of active compounds.

The structures of each of the active compounds are presented in Fig. 3 (the structure of Pentostam is included for comparison).

FIG. 3.

Structures of compounds 83633, 531520, and 13512 and Pentostam. The compound structures were obtained from the National Cancer Institute website, and that of Pentostam was reported previously (2). The structure of each compound was redrawn by using Chemsketch software from ACD/labs (http://www.acdlabs.com/download/).

DISCUSSION

The World Health Organization estimates that ∼500,000 new cases of leishmaniasis occur each year (21, 22). U.S. military personnel deployed to Afghanistan and Iraq in support of Operations Enduring Freedom and Iraqi Freedom are particularly at risk, with infection rates as high as 2.3% (15). Treatment often requires transport to Brook Army Medical Center or Water Reed Army Medical Center for a 10- to 20-day course of Pentostam, with effects ranging from mild rash and myalgia to more severe chemical pancreatitis (23).

In the search for new potent and safe anti-Leishmania compounds, we have screened a small portion of the unique NCI compound library, using a tiered assay. Of the initial 15,000 compounds, ∼5% inhibited the growth of L. tarentolae. However, only ∼120 of these were suitable for further development due to either cytotoxicity or structure considerations. Of these, 23 had antibacterial activities and 7 had antifungal activities. Whether the seven compounds with antifungal activities should have been eliminated is a matter of debate; however, we specifically did not want to discover another azole or polyene but, rather, wished to discover novel compounds that had new modes of action. We chose the most active 50 of the 90 remaining compounds for testing against L. major. Interestingly, seven compounds that were active against L. tarentolae did not have activity against L. major, even at the highest concentration tested.

Each of the 43 compounds that were active against both L. tarentolae and L. major promastigotes in vitro was tested in the murine macrophage model of L. major infection. We found that some compounds that were very potent in the in vitro promastigote growth inhibition assays had little or no activity in the amastigote macrophage assay (e.g., NSC compound 377). This may be a reflection of the exclusion of the compound by permeability considerations or by rapid intracellular degradation. The lack of a precise correlation between the MIC obtained in vitro and the efficacy (or, in this case, intracellular efficacy) obtained in vivo is, unfortunately, a well-known observation in the screening of antibacterials and antifungals.

However, we found three compounds that were potent in clearing infected macrophages. Treatment with concentrations as low as ∼2% (e.g., 1.1 μg/ml for compound 351520) of that of Pentostam resulted in a significant reduction in the number of infected cells and the number of intracellular parasites (Fig. 2A and B, respectively). Interestingly, none of these compounds and Pentostam share obvious core pharmacophores (Fig. 3). This suggests that each compound may act by a different mechanism. The determination of the mechanism of action of each compound is currently unknown but is the subject of intense investigation in our laboratories.

Given the increasing incidence of leishmaniasis and the need for additional and less toxic treatment options, we have discovered three compounds that are 7- to 25-fold more active than Pentostam using a modest and simple screen. These results encourage additional screening and the further testing of the current compounds.