Evaluating Aziridinyl Nitrobenzamide Compounds as Leishmanicidal Prodrugs

Abstract

Many of the nitroaromatic agents used in medicine function as prodrugs and must undergo activation before exerting their toxic effects. In most cases, this is catalyzed by flavin mononucleotide (FMN)-dependent type I nitroreductases (NTRs), a class of enzyme absent from higher eukaryotes but expressed by bacteria and several eukaryotic microbes, including trypanosomes and Leishmania. Here, we utilize this difference to evaluate whether members of a library of aziridinyl nitrobenzamides have activity against Leishmania major. Biochemical screens using purified L. major NTR (LmNTR) revealed that compounds containing an aziridinyl-2,4-dinitrobenzyl core were effective substrates for the enzyme and showed that the 4-nitro group was important for this activity. To facilitate drug screening against intracellular amastigote parasites, we generated leishmanial cells that expressed the luciferase reporter gene and optimized a mammalian infection model in a 96-well plate format. A subset of aziridinyl-2,4-dinitrobenzyl compounds possessing a 5-amide substituent displayed significant growth-inhibitory properties against the parasite, with the most potent agents generating 50% inhibitory concentrations of <100 nM for the intracellular form. This antimicrobial activity was shown to be LmNTR specific since L. major NTR^+/− heterozygote parasites were slightly resistance to most aziridinyl dinitrobenzyl agents tested. When the most potent leishmanicidal agents were screened against the mammalian cells in which the amastigote parasites were propagated, no growth-inhibitory effect was observed at concentrations of up to 100 μM. We conclude that the aziridinyl nitrobenzamides represent a new lead structure that may have the potential to treat leishmanial infections.

INTRODUCTION

Leishmaniasis represents a series of insect-transmitted, blood-borne diseases caused by more than 20 different protozoan parasite species belonging to the genus Leishmania. These infections are endemic throughout many tropical and subtropical countries where 350 million people are at risk of infection (1). Estimates indicate that up to 12 million individuals are currently infected by these protozoan parasites, with up to 2 million new cases and 50,000 deaths occurring each year (1). Recently, due to military activity, population migration, modern medical practices, intravenous drug usage, and global warming, the number of new cases in areas of nonendemicity has increased, stimulating interest from pharmaceutical companies in these previously neglected infections (2–4). Drugs currently represent the only treatments available to combat leishmaniasis. For more than 60 years, front-line therapies have been based on pentavalent antimonial compounds, but their use is problematic, as they are toxic and administration requires medical supervision, with clinical resistance now commonplace (5, 6). In light of this worrying situation, a range of alternative treatments employing preparations such as amphotericin B, paromomycin, and miltefosine are now available, but these too are far from ideal, as they can be expensive and require medical administration, with some having teratogenic and other unwanted toxicity problems (7). Therefore, there is an urgent requirement for new, safer, cost-effective antileishmanial treatments.

Nitroaromatic compounds are used predominantly as broad-spectrum antibiotics to treat various urinary and gastrointestinal tract infections. They are characterized by possessing at least one nitro group attached to an aromatic ring that usually has a heterocyclic structure (e.g., imidazole, furan, or thiazole) (8). However, following concerns over their safety, the use of many nitroaromatics has been discontinued in Europe and the United States, although they are commonly prescribed elsewhere (9–11; reviewed in reference 12). It is now apparent that several nitro-based compounds are not as toxic as initially thought (13–15). Such observations have stimulated renewed interest in this group of agents, with calls for the reinstatement of nitrofurantoin as a treatment for uncomplicated urinary tract infections, while several others have emerged as lead structures to treat various microbial infections and different forms of cancer, with fexinidazole and PA-824 undergoing evaluation to treat visceral leishmaniasis (16–24).

Most nitroaromatic compounds used in medicine function as prodrugs and must undergo activation before mediating their therapeutic effects, reactions catalyzed by nitroreductases (NTRs). Based on oxygen sensitivity, the flavin cofactor, and the reduction products, NTRs can be broadly divided into two classes, type I and type II NTRs (25). In many bacterial and eukaryotic microorganisms, flavin mononucleotide (FMN)-containing type I NTRs use NAD(P)H to drive the sequential two-electron reduction of the conserved nitro (-NO₂) group to a hydroxylamine derivative (-NHOH) via an unstable nitroso intermediate (-NO) (see reactions 1 and 2). The hydroxylamine can be processed further to produce DNA-damaging/cross-linking adducts and other toxic molecules such as open-chain nitriles and reactive dialdehydes (18, 25–32). Crucially, this reaction can occur in aerobic and anerobic environments, indicating that the reactions described above do not involve O₂; as such, FMN-type I catalyzed activity is said to be “O₂ insensitive.” For bacteria and trypanosomes, this mechanism of activation appears to be key for the selective toxicity of many nitroaromatic prodrugs, as cells selected for resistance frequently show mutations in, or decreased expression of, their FMN-type I NTR gene complement (33–35).

$$R\text{-}{NO}_2+2e^{-}\to R\text{-}NO$$ (1)

$$R\text{-NO}+2e^{-}\to R\text{-}NHOH$$ (2)

In contrast, the ubiquitous NAD(P)H-dependent type II NTRs use flavin adenine dinucleotide (FAD) or FMN as a cofactor to catalyze the one-electron reduction of the conserved nitro group to form an unstable nitro anion radical (-NO₂^·−). In the presence of O₂, this radical can undergo futile cycling to produce superoxide anions (O₂⁻) and regenerate the parental compound (36, 37) (see reactions 3 and 4 below). Mammalian enzymes, such as NAD(P)H quinone oxidoreductase 1, can catalyze a two-electron reduction reaction with many nitroaromatic prodrugs under aerobic conditions; however, mammalian nitroreduction does not utilize FMN-containing type I enzymes (38).

$$R\text{-}{NO}_2+2e^{-}\to R\text{-}{{NO}_2}^{\cdot -}$$ (3)

$$R\text{-}{{NO}_2}^{\cdot -}+O_2\to R\text{-}{NO}_2+{O_2}^{\cdot -}$$ (4)

Here, we exploit the activity of a FMN-type I NTR expressed by the protozoan parasite Leishmania major to conduct biochemical and phenotypic screens on a small library of azirindyl nitrobenzamide (ANB) compounds, agents we have previously shown to display significant antitrypanosomal activities (39). Two of these compounds were highly active against L. major in its intracellular form and displayed high selectivity toward the parasite. We postulate that compounds based on a 5-(aziridin-1-yl)-2,4-dinitrobenzamide core represent a promising new class of leishmanicidal agents.

MATERIALS AND METHODS

Chemicals.

Structures of aziridinyl nitrobenzamides (ANBs) are shown in Table 1. The synthesis of NH1 to NH8 is described elsewhere (40). CB1954 was purchased from Sigma-Aldrich, and NH9 to NH12 were supplied by the Department of Therapeutics, U.S. NCI.

TABLE 1.

Structures of aziridinyl nitrobenzamide compounds

Cell culturing.

L. major (MHOM/IL/80/Friedlin) promastigote parasites were grown at 27°C in modified M199 medium (Life Technologies) (41). Transformed parasites were grown in this medium supplemented with G418 (20 μg ml⁻¹ on agar plates, 40 μg ml⁻¹ in broth) or blasticidin (10 μg ml⁻¹). L. major metacyclic parasites were harvested from promastigote cultures as described previously (42). These were used to infect differentiated human acute monocytic leukemia (THP-1) cells at a ratio of 20 parasites per mammalian cell. The L. major-infected monolayers were incubated overnight at 37°C under a 5% (vol/vol) CO₂ atmosphere in mammalian growth medium and then washed with RPMI 1640 to remove residual parasites. L. major amastigote parasites were maintained in differentiated THP-1 cells at 37°C under a 5% (vol/vol) CO₂ atmosphere in RPMI 1640 medium.

The human acute monocytic leukemia cell line (THP-1) was grown at 37°C under a 5% (vol/vol) CO₂ atmosphere in RPMI 1640 medium (PAA Laboratories Ltd.) supplemented with 2 mM pyruvate, 2 mM sodium glutamate, 2.5 U ml⁻¹ penicillin and 2.5 μg ml⁻¹ streptomycin, 20 mM HEPES (pH 7.4), and 10% (vol/vol) fetal calf serum. Differentiation of THP-1 to produce macrophage-like cells was carried out using phorbol 12-myristate 13-acetate (PMA) (Sigma-Aldrich) (20 ng ml⁻¹) (43, 44).

In vivo studies.

All animal experiments were conducted under license in accordance with United Kingdom Home Office regulations. L. major parasites were passaged through female BALB/c mice by subcutaneous injection of 2 × 10⁷ purified metacyclic parasites in 100 μl RPMI medium without serum into the shaved rump. Amastigotes were harvested from skin lesions and allowed to transform back to promastigotes in 5 ml M199 medium.

Antiproliferative assays.

All assays were performed in a 96-well plate format. L. major promastigote parasites (5 × 10⁵ ml⁻¹) or differentiated THP-1 cells (2.5 × 10⁴ ml⁻¹) were seeded in 200 μl growth medium containing different concentrations of nitroaromatic agent. After incubation at 27°C for 5 days (L. major) or at 37°C for 3 days (THP-1), 2.5 μg resazurin (20 μl of 0.125 μg ml⁻¹ stock–phosphate-buffered saline) was added to each well and the plates were incubated for a further 8 to 16 h. Cell densities were determined by monitoring the fluorescence of each culture using a Gemini fluorescent plate reader [Molecular Devices (UK) Ltd., Wokingham, United Kingdom] at an excitation wavelength of 530 nm, an emission wavelength of 585 nm, and a filter cutoff at 550 nm, and the drug concentration that inhibits cell growth by 50% (IC₅₀) was established.

Growth inhibition of luciferase-expressing L. major amastigotes was monitored as follows. THP-1 cells seeded at 2.5 × 10⁴ ml⁻¹ in 200 μl of growth medium containing PMA (20 ng ml⁻¹) were incubated at 37°C in a 5% (vol/vol) CO₂ atmosphere for 3 days. Macrophage monolayers were washed with mammalian growth medium and then infected with purified luciferase-expressing L. major metacyclic cells (5 × 10⁵ cells ml⁻¹) resuspended in 200 μl of mammalian growth medium. Following incubation overnight at 37°C in a 5% (vol/vol) CO₂ atmosphere, the cultures were washed twice in growth medium to remove noninternalized parasites and the supernatant was replaced with fresh growth medium containing the compound under investigation. Compound-treated infections were incubated for a further 3 days at 37°C under a 5% (vol/vol) CO₂ atmosphere. The growth medium was then removed, and the cells were lysed in 50 μl cell culture lysis reagent (Promega). Activity was then measured using the luciferase assay system (Promega) and light emission measured on a ß-plate counter (PerkinElmer). The luminescence was proportional to the number of live cells. The IC₅₀ for each compound was then established.

Plasmids and parasite genetic manipulation.

DNA fragments corresponding to the L. major 5′ rRNA spacer/promoter and 3′ spacer rRNA regions were amplified from L. major Friedlin genomic DNA and sequentially cloned into Trypanosoma cruzi vector pTRIX-Luc (39), replacing the equivalent T. cruzi 5′ rRNA spacer/promoter and 3′ spacer rRNA sequences. A polypyrimidine tract/spliced leader additional site sequence located upstream of the T. cruzi MPX gene was then isolated and inserted between the L. major 5′ rRNA spacer/promoter region and luciferase reporter to form the integrative vector pLmRIX-Luc. Following linearization, this DNA was introduced into L. major promastigotes in the logarithmic phase of growth using a human T-cell Nucleofector kit and an Amaxa Nucleofector (Lonza AG) set to program U-033.

A DNA fragment encoding the catalytic domain of L. major NTR (LmNTR) was amplified from genomic DNA with the primers ggatccCTCGACGCCGTCGAGGCCGTCG and gaattcCTAGAACTTGTTCCACCGCAC; lower-case italics correspond to restriction sites incorporated into the primers to facilitate cloning. The fragment was digested with BamHI and HindIII and then cloned into the corresponding sites of vector pTrcHis-C (Invitrogen) to form the plasmid pTrcHisC-LmNTR.

Enzyme assay.

Recombinant His-tagged LmNTR was purified as described previously (45). Type I NTR activity was measured spectrophotometrically by monitoring the formation of the 2- and 4-hydroxylamine derivatives from the parent ANB (λ = 420 nm, ε = 1,220 M⁻¹ cm⁻¹) (46). A standard reaction mixture (1 ml) containing 50 mM Tris-Cl (pH 7.5), 100 μM NADH, and 100 μM electron acceptor was incubated at room temperature for 5 min. The background reaction rate was determined and the assay initiated by addition of the LmNTR (35 μg). For nifurtimox, activity was measured spectrophotometrically by monitoring reduction of this compound (λ = 435 nm, ε = 19,000 M⁻¹ cm⁻¹) (47). Enzyme activities were expressed in nmol of NADH oxidized per minute per mg LmNTR with the assumption that four NADH molecules are oxidized per molecule of ANB or nifurtimox reduced (46, 47).

RESULTS

Construction and evaluation of luciferase-expressing Leishmania major.

Several drug-screening systems are now in place for use with Leishmania, each having its own advantages and disadvantages. To facilitate screening of nitroaromatic compounds, which are often colored, we developed a L. major line that constitutively expresses luciferase. The integrative vector pLmRIX-Luc was generated by sequentially cloning DNA fragments containing the 5′ and 3′ L. major Friedlin rRNA promoter/spacer sequences on each side of an expression cassette derived from pTEX containing the luciferase and neomycin phosphotransferase genes (39). To assist luciferase mRNA processing, an untranslated sequence corresponding to the T. cruzi mitochondrial peroxiredoxin gene polypyrimidine tract/spliced leader addition site was then inserted between the 5′ L. major rRNA region and reporter gene. The DNA fragment containing L. major rRNA/luciferase/neomycin phosphotransferase sequences was then purified following restriction digestion and electroporated into promastigote-form parasites. After selection with G418, the luciferase activity from several clones was determined and shown to be up to 1,000-fold higher than that of the parental line (Fig. 1A). The effect of luciferase expression on various L. major life cycle stages was then evaluated. This showed that the reporter did not influence (i) growth of promastigote parasites, (ii) the ability of promastigote cells to differentiate into infective metacyclic forms, (iii) invasion of tissue culture-derived macrophages by metacyclics, or (iv) growth of intracellular amastigote parasites (Fig. 1B and E). Additionally, the luciferase-expressing parasites were passaged through BALB/c mice and amastigotes recovered from infected animals could readily differentiate to the promastigote form and be grown in culture. Therefore, it is implicit that luciferase expression has no effect on L. major growth, differentiation, and infectivity.

FIG 1.

Luciferase expression in insect- and mammal-stage L. major. (A) The luciferase activity, expressed in relative light units, of six recombinant L. major promastigote clones (Luc1 to Luc6) was determined and compared to that of the parental line (wt). A total of 1 × 10⁶ cells were used in each analysis. (B) Growth of L. major wild-type (squares; solid line) and LmNTR^+/− BLA heterozygote (diamonds; dotted line) promastigote parasites was monitored until cultures were in the stationary phase of growth. From day 5 onward, the number of metacyclic parasites in wild-type (triangles; solid line) and LmNTR^+/− BLA heterozygote (triangles; dotted line) promastigote cultures was determined following purification by agglutination. The data are expressed as percent load of metacyclics in the total L. major population. All curves shown are derived from a single data set and are representative of the results of experiments performed in triplicate. (C) Purified L. major metacyclic parasites engineered to express luciferase were used to infect differentiated THP-1 cells. Over a 4-day postinfection period, extracts were generated from each cell line and the luciferase activity was determined. Following background correction, the luciferase activity was plotted against time. All curves shown were derived from a single data set and are representative of the results of experiments performed in triplicate. (D) Correlation between promastigote load (between 625 and 160,000 cells) and luciferase activity. Three independent readings were taken for each parasite load, and the values are means ± standard deviations. (E) Correlation between luciferase activity and L. major amastigote load using various parasite/macrophage cell ratios (0:1 to 80:1).

A prerequisite for any drug-screening assay requires that reporter levels provide an accurate representation of cell number. Using extracts derived from serially diluted promastigotes, a linear relationship between luciferase activity and parasite load was observed over the range of 625 to 160,000 cells (Fig. 1C). In the case of amastigotes, such a correlation was shown by evaluating reporter levels in lysates from a fixed number of macrophages infected for 16 to 20 h with various L. major loads (Fig. 1D); luciferase activity was linear when using a parasite:mammalian cell ratio of 2.5 to 80:1. With this established, the luciferase-expressing cells were then appraised as to whether they could be used for in vitro drug screens in a 96-well plate format using nifurtimox as the reference compound. For promastigote assays, cells were grown in the presence of different concentrations of the nitrofuran for 6 days and the extracts tested for luminescence activity (Fig. 2A). From the resultant dose-response curves, nifurtimox exhibited an IC₅₀ of 7.50 ± 0.40 μM, in line with the IC₅₀ (6.30 ± 0.10 μM) obtained when using the fluorescent vital dye resazurin. For the intracellular amastigote assays, differentiated THP-1 cells were infected with recombinant L. major at a ratio of 20 parasites per mammalian cell. Following 72 h growth postinfection in the presence of drug, the luciferase activity for each culture was determined and dose-response curves were generated (Fig. 2B). From these plots, the IC₅₀ for nifurtimox was calculated to be 0.58 ± 0.06 μM. When the susceptibility of differentiated THP-1 cells to nifurtimox was determined using resazurin, the IC₅₀ was calculated to be >100 μM (Table 2), demonstrating that this nitrofuran has a selective toxicity of >170 for the intracellular parasite. Therefore, based on our data, nifurtimox shows selective killing of the L. major amastigotes and the observed leishmanicidal activity is not due to death of the mammalian cell.

FIG 2.

Susceptibility of L. major to nifurtimox. Dose-response curves for luciferase-expressing L. major promastigotes (A) and amastigotes (B) exposed to nifurtimox treatment were determined, and IC₅₀s were calculated from the results. The data represent the means of the results from three independent experiments ± standard deviations.

TABLE 2.

Susceptibility of L. major and THP-1 cells to aziridinyl nitrobenzamides^a

Compound	Group	L. major IC50 (μM)		Differentiated THP-1 IC50 (μM)c	Selective toxicityd
		Promastigotesb	Amastigotesc
Nifurtimox		7.50 ± 0.40	0.58 ± 0.06	>100	>172
CB1954	Ia	0.42 ± 0.01	0.05 ± 0.02	>100	>2000
NH1	Ia	7.32 ± 0.63	11.15 ± 2.01	>100	>9
NH2	Ia	1.48 ± 0.18	2.95 ± 1.14	>100	>34
NH9	Ia	>10	0.67 ± 0.05	>100	>149
NH10	Ia	0.49 ± 0.04	0.56 ± 0.05	>100	>178
NH11	Ia	0.91 ± 0.07	0.06 ± 0.01	>100	>1667
NH12	Ia	>10	1.32 ± 0.02	>100	>76
NH3-5	Ib	>10	>10	ND	ND
NH6	II	6.85 ± 0.18	2.10 ± 0.22	>100	>48
NH7-8	II	>10	>10	ND	ND

^aND, not determined.

^bData are means from 4 experiments ± standard deviations.

^cData are means from 3 experiments ± standard deviations.

^dThe therapeutic index of a compound was calculated as a ratio of the IC₅₀ against differentiated THP-1 cells to the IC₅₀ against amastigote parasites. Nifurtimox was used as a reference compound.

Metabolism of nitrobenzamides by the leishmanial NTR.

Many nitroaromatic prodrugs undergo activation in reactions catalyzed by FMN-type I NTRs. Here, we evaluated whether CB1954 and 12 related ANBs could function as substrates for purified His-tagged L. major NTR using NADH as an electron donor (Fig. 3). Normally, assays involving this parasite enzyme can be readily followed by monitoring the change in absorbance at 340 nm, corresponding to the oxidation of NADH. However, many ANBs undergo a considerable change in absorbance at this particular wavelength. Instead, enzyme activity was assayed by monitoring the change in absorbance at 420 nm, which corresponds to the appearance of the hydroxylamine metabolite (46).

FIG 3.

Activity of LmNTR toward different aziridinyl nitrobenzamides. (A) SDS-PAGE gel (10%) stained with Coomassie blue. Lane 1, size standards; lane 2, purified His-tagged LmNTR. (B) The activity of purified recombinant LmNTR was assessed by using various ANBs (100 μM) as substrates at a fixed concentration of NADH (100 μM). Enzyme activity, expressed in nmol of NADH oxidized per minute per mg LmNTR (nmol·mg⁻¹ min⁻¹), was then calculated using an ε value of 1,220 M⁻¹ cm⁻¹ with the assumption that four molecules of NADH are oxidized per molecule of ANB reduced (46). NFX (nifurtimox) was used as the control. For nifurtimox, the change in absorbance at 435 nm was followed and enzyme activity determined using an ε value of 19,000 mM⁻¹ cm⁻¹, again assuming that four molecules of NADH were oxidized per molecule of nitrofuran reduced (47). The enzyme activity values are the means of data from three assays ± standard deviations. The ANB group designations as listed in Table 1 are noted.

Of the 13 compounds screened, most (10) were metabolized by LmNTR at a rate higher than that noted for nifurtimox. These “active” compounds were related in that they all contained two nitro groups at positions 2- and 4- relative to the 1-aziridinyl ring. In contrast, the three ANBs that contain a single 2-nitro group were deemed “poor” substrates for the parasite enzyme, having negligible activity values.

Leishmanicidal activity and mammalian cytotoxicity of aziridinyl nitrobenzamides.

To determine whether there was a correlation between LmNTR activity and antileishmanial activity, all compounds were screened against L. major promastigotes and luciferase-expressing amastigotes. Initial tests using a fixed concentration of compound (10 μM) were set up to evaluate the growth-inhibitory properties of the ANB series. Of the 13 compounds tested, 5 had no effect on either parasite stage at this concentration (Table 2) whereas 2 others had no effect on promastigote cells but did have activity against L. major in intracellular form. Three of the five compounds that had no effect against promastigote and amastigote parasites correspond to the ANBs previously identified as “poor” LmNTR substrates. To evaluate the growth-inhibitory activities of those structures identified by the initial screen, a series of secondary assays were performed using various concentrations of the ANB. For each agent, dose-response curves were drawn from which the compound's IC₅₀ was determined (Table 2) (Fig. 4). Of the six compounds tested against promastigotes, CB1954, NH10, and NH11 had IC₅₀s < 1 μM. These, plus NH9, also displayed submicromolar IC₅₀s against amastigote parasites, with two (CB1964 and NH11) yielding values < 100 nM. All compounds that displayed leishmanicidal activity against either of the L. major forms were assayed for cytotoxicity against differentiated THP-1 cells (Table 2) (Fig. 4). In all cases, the ANBs screened had no growth-inhibitory effect against this mammalian line at concentrations of up to 100 μM. Comparison of the mammalian cell IC₅₀ with the equivalent value obtained against amastigote parasites revealed that CB1964 and NH11 displayed >2,000- and >1,667-fold selective toxicity toward the intracellular pathogen, respectively. This clearly demonstrates that the antiparasitic activity displayed by these two compounds was specifically due to growth inhibition of L. major and not due to induction of host cell toxicity.

FIG 4.

Dose-response curves of L. major and mammalian cells for aziridinyl nitrobenzamides. Various concentrations of leishmanicidal ANBs were tested against L. major and THP-1 cells. The growth-inhibitory effect of each treatment was evaluated, and dose-response curves were constructed for promastigotes (•), amastigotes (■), and differentiated THP-1 cells (▲). In all cases, drug treatments were performed in triplicate and the plots shown represent the average growth inhibition obtained at each concentration ± standard deviation. The curves for compounds CB1954, NH6, and NH10 are shown.

To demonstrate that LmNTR played a role in prodrug activiation in the parasite itself, the susceptibility of LmNTR^+/− heterozygous promastigotes to CB1954, NH10, and NH11 was evaluated; attempts (16 independent transformations) to generate L. major LmNTR^−/− null mutants failed to produce recombinant cells, leading us to speculate that this activity is essential in insect-stage parasites (48). This screening demonstrated that cells expressing lower levels of the oxidoreductase were between 2- and 4-fold more resistant to the ANB than controls; wild-type L. major had IC₅₀s of 0.42 ± 0.01 μM, 0.49 ± 0.04 μM, and 0.91 ± 0.07 μM against CB1954, NH10, and NH11, respectively, while LmNTR^+/− heterozygotes displayed IC₅₀s of 1.62 ± 0.02 μM, 1.00 ± 0.07 μM, and 1.77 ± 0.01 μM against these agents.

DISCUSSION

Several high-throughput phenotypic screening approaches are now available to facilitate the search for novel drugs targeting the medically relevant, intracellular stage of Leishmania. Of these, the luciferase-based systems have proven to be sensitive, rapid, reproducible, and versatile, with tagged parasites now being used to follow the fate of the pathogen during the course of an animal model infection (49–52). Here, we constructed a luciferase-expressing L. major line making use of an existing T. cruzi integration vector, modifying the expression construct such that it would integrate into the leishmanial ribosomal array (39). Characterization of the recombinant line established that expression of the reporter had no effect on L. major promastigote growth and metacyclogenesis or proliferation of the intracellular amastigote form. Using luciferase-expressing L. major, growth inhibition assays were established with nifurtimox employed as a selective compound for promastigote or amastigote cells in a standardized 96-well plate format.

Nitrobenzamide-based compounds that contain an aziridinyl ring or mustard substituent have been evaluated as potential therapies targeting hypoxic cancers and the trypanosomal infections African sleeping sickness and Chagas' disease (53–55). These structures invariably function as prodrugs, with their toxicity dependent on reduction of the nitro group(s) to its hydroxylamine derivatives, reactions catalyzed by FMN-type I NTRs (18, 21). This bioreductive trigger promotes an electronic reconfiguration on the compound's aromatic ring, leading to presentation of cytotoxic moieties to the cell (55). For the treatment of hypoxic cancers, the FMN-type I NTR activity must be introduced into mammalian cells using gene- or antibody-based approaches before addition of the nitrobenzamide whereas in trypanosomes an essential endogenous enzyme can be exploited to catalyze the nitro reduction described above (18, 21, 55). As Leishmania parasites also express a FMN-type I NTR, it is postulated that nitroaromatic compounds may have potential for the treatment of various forms of leishmaniasis (24).

Following the finding that CB1954, the archetypal ANB, displayed potent antitumor activity against the Walker 256 carcinoma, a series of derivatives differing in the numbers and locations of nitro groups and other substituents attached to a central benzyl ring core have been synthesized (40, 56). Using a small library (14 compounds) of these structures in biochemical screens, we demonstrated that compounds having two nitro groups located at the 2- and 4- positions on an aromatic ring backbone relative to the aziridinyl ring were readily metabolized by LmNTR. The location of an amide (or amine)-containing substituent at the 5- (group Ia) or 6- (group II) position did not affect this in vitro activity. In contrast, replacement of the 4-nitro group with an H or SO₂Me substituent (group Ib) generated compounds that were not metabolized by LmNTR. These biochemical studies revealed that LmNTR exhibits a substrate preference that is slightly different from that of Trypanosoma brucei NTR (TbNTR). In contrast to the LmNTR results, the trypanosomal enzyme was able to metabolize only compounds having the group Ia configuration. This highlights the importance of the 4-nitro group during prodrug activation, with both LmNTR and TbNTR unable to metabolize compounds with a group Ib arrangement (39).

When tested against Leishmania, all group Ia structures showed potent growth-inhibitory properties for promastigote and/or amastigote parasites, thus mirroring the biochemical observations. Several of these ANBs yielded IC₅₀s below 1 μM against intracellular L. major, including CB1954 and NH11, which had sub-100 nM values. In contrast, none of the group Ib ANBs had an effect on parasite growth, consistent with observations made using the trypanosomal NTRs and mammalian cells expressing Escherichia coli FMN-type I NTRs, again highlighting the importance of the 4-nitro group and its reduction products in mediating cytotoxicity (39, 40, 57). Of the group II compounds, surprisingly, only NH6 displayed any leishmanicidal activity. NH6 had lower potency than its structural isomer CB1954 (CB1954 was 42- and 16-fold more effective at inhibiting amastigote and promastigote parasite growth than NH6, respectively). The decreased potency displayed by group II compounds may reflect the ability of these structures (or their reduction products) to access regions of the cell where activation (or downstream leishmanicidal processes) takes place. This could be in part due to the spatial arrangement of these substituents (located at the 6- position on the benzyl ring), which may hinder the presentation of the adjacent azirindinyl cytotoxic moiety to biomolecular targets within the cellular environment.

To conclusively demonstrate the link between LmNTR and leishmanicidal activities, the susceptibility of the L. major LmNTR^+/− heterozygous cell line to the most effective compounds (CB1954, NH10, and NH11) was investigated (49). In this context, parasites with lower levels of NTR displayed relative resistance to all compounds tested, mirroring observations made for the trypanocidal, NTR-activated nitroaromatic prodrugs (35, 39, 45, 47). This, in conjunction with the observation that none of the group Ia compounds displayed cytotoxicity to the macrophage-like cells within which the intracellular L. major parasites were cultured, suggests that the group Ia ANBs act by specifically targeting the parasite itself rather than by promoting death to the mammalian cell.

We have now identified two leishmanidical ANB-based compounds (CB1954 and NH11) that display significant potency toward the L. major forms that replicate inside the mammalian host. These structures were shown to mediate their antiparasitic activities following specific activation in a reaction catalyzed by a FMN-type I NTR, an activity present in the pathogen but absent from the mammalian macrophage cells. Indeed, these compounds display little or no cytotoxicity to the mammalian macrophage-like cell, having >1,660-fold selectivity when targeting intracellular parasites, although when in vivo rodent models are used, CB1954 does exhibit hepatotoxicity and neurotoxicity and causes gastrointestinal tract disturbances (58, 59); the toxicity/pharmacokinetics of other ANBs screened here have not been determined. Despite this, group Ia compounds, particularly, CB1954 and NH11, warrant further attention in the development of novel leishmanidical therapies and ideally represent one component of a new combinatorial treatment.