Synthesis and pharmacological evaluation of mono-arylimidamides as antileishmanial agents

Graphical abstract

Leishmaniasis Arylimidamide DB766 Drug discovery a b s t r a c t Arylimidamide (AIA) compounds containing two pyridylimidamide terminal groups (bis-AIAs) possess outstanding in vitro antileishmanial activity, and the frontrunner bis-AIA DB766 (2,5-bis[2-(2-isopropoxy)-4-(2-pyridylimino)aminophenyl]furan) is active in visceral leishmaniasis models when given orally. Eighteen compounds containing a single pyridylimidamide terminal group (mono-AIAs) were synthesized and evaluated for their antileishmanial potential. Six of these compounds exhibited submicromolar potency against both intracellular Leishmania donovani and Leishmania amazonensis amastigotes, and three of these compounds also displayed selectivity indexes of 25 or greater for the parasites compared to a J774 macrophage cell line. When given orally at a dose of 100 mg/kg/day for five days, compound 1b (N-(3-isopropoxy-4-(5-phenylfuran-2-yl)phenyl)picolinimidamide methanesulfonate) reduced liver parasitemia by 46% in L. donovani-infected mice. Mono-AIAs are thus a new class of candidate molecules for antileishmanial drug development. With well over one million new cases and 20,000-40,000 deaths annually, 1 leishmaniasis is the second most prevalent infectious disease transmitted by a vector in terms of morbidity and mortality. 2 Leishmaniasis is a spectrum of disease caused by infection with approximately twenty species of Leishmania protozoa. Among the major clinical forms, visceral leishmaniasis (VL) is the most severe and is fatal if left untreated, while cutaneous leishmaniasis (CL) is typically self-healing. Nevertheless, drugs are frequently used to treat CL to limit scarring from cutaneous lesions, to speed the time to cure, and to prevent the development of the potentially disfiguring mucocutaneous leishmaniasis manifestation. The current drugs against leishmaniasis include pentavalent antimonials, amphotericin B (AmB), miltefosine and paromomycin, but they all suffer from one or more weaknesses, such as parasite resistance, toxicity, high cost, and an inconvenient route of administration. 2 New oral antileishmanial drugs are thus needed to overcome these issues and provide better treatment options against Leishmania infections.
We have demonstrated that bis-arylimidamides (bis-AIAs), a series of compounds containing two pyridylimidamide terminal groups, possess excellent in vitro and promising in vivo antileishmanial activity. In the AIAs, the imino group is bound to an anilino nitrogen atom, lowering the pK a of the amidine and increasing the lipophilicity of the molecule compared to a dicationic diamidine compound such as pentamidine. 3 DB766 and DB1960, the hydrochloride and mesylate salts of 2,5-bis[2-(2-isopropoxy)-4-(2-pyridylimino)aminophenyl]furan, respectively, display IC 50 values similar to AmB against both intracellular Leishmania donovani and intracellular Leishmania amazonensis and provide dose-dependent reduction of liver parasitemia in a mouse model of visceral leishmaniasis when administered orally. 3,4 Unfortunately, DB766 and DB1960 lack the therapeutic window required for advancement as clinical monotherapy candidates against visceral leishmaniasis. 4 As part of a broader study, we previously reported two terphenyl AIAs containing a single pyridylimidamide group (mono-AIAs) that displayed promising activity against intracellular L. amazonensis. 5 We hypothesize that lower molecular weight mono-AIAs may have improved pharmacokinetic properties and could display decreased toxicity to animals compared to bis-AIAs. The synthesis and evaluation of such a group of molecules is reported here to test this hypothesis.
Most of the mono-AIAs investigated were synthesized as outlined in Scheme 1 using methodology similar to that employed previously for making bis-AIAs. [6][7][8] The first step involved Stille coupling between tributyl(furan-2-yl)stannane and various substituted 4-bromonitrobenzenes. Bromination of the 2-(4-nitrophenyl)furan analogs with NBS at room temperature yielded the corresponding 5-bromo-2-(4-nitrophenyl)furans. Subsequent reaction of the 5-bromofurans under either Stille or Suzuki conditions provided the 2-(4-nitrophenyl)-5-arylfuran intermediates. Catalytic hydrogenation of the nitro group of the diarylfurans gave the corresponding amino analogs. Finally, reaction of the amino diarylfurans with either 2-pyridyl or 2-pyrimidyl thioimidate analogs produced, after salt formation, the mono-AIAs 1a-p. Most target compounds were prepared as the more soluble mesylate salt. 4 Hydrochloride salts were made for 1a and 1d because of the hygroscopic nature of the corresponding mesylates.
Methods employed for the evaluation of the efficacy of mono-AIA target compounds against intracellular L. amazonensis 14 and intracellular L. donovani 15 and for toxicity to murine J774 macrophages 16 have been described previously. In these assays, the reference compound amphotericin B displayed IC 50 values of 40 ± 3 nM (mean ± standard error, n = 15) and 120 ± 10 nM (mean ± standard error, n = 26) against L. donovani and L. amazonensis, respectively, while the cytotoxicity standard podophyllotoxin showed an IC 50 value of 24 ± 5 nM (mean ± standard error, n = 15) against J774 macrophages. For ease in illustrating the antileishmanial SAR of this series of compounds, the central phenyl ring, the linker, and the terminal phenyl ring are referred to as rings A, B, and C, respectively (Fig. 1).
For derivatives bearing substitutions at the 5 position of the furan 'B' ring (at the position occupied by the C ring in Fig. 1), we observed that the most active compound was 1b, where the C ring is a phenyl group. Compound 1b displayed IC 50 values of 310 and 130 nM against intracellular L. donovani and intracellular L. amazonensis, respectively (Table 1). Activity was maintained when the phenyl ring was replaced with a furan ring (1d). The antileishmanial potency decreased by 4-20 fold when this position was substituted with other groups, however, such as hydrogen (1a), 5-pyrimidyl (1c), 3-pyridyl (1e), or 4-pyridyl (1f).
The addition of a methoxy group or a fluorine atom at position R 2 or a methoxy group at position R 1 of phenyl ring C led to decreased antileishmanial potency in both the intracellular L. amazonensis and L. donovani assays ( Table 2). In the series of compounds bearing different substitutions on phenyl ring 'A', the cyclopentyloxy substituted derivative 1k was the most active compound, with IC 50 values comparable to that of 1b and amphotericin B ( Table 3). As the size of the alkoxy substituent increased from methoxy (1m) to ethoxy (1l) to isopropoxy (1b) and cyclopentyloxy (1k), antileishmanial activity increased, but switching the position of the isopropoxy substitution on the 'A' ring (1j) resulted in a loss of antiparasitic activity. Replacement of furan as the 'B' ring with other heterocycles (thiophene 1o, triazole 2, or oxazole 3) lowered the antileishmanial activity, with 3 being intermediate in potency between the highly active 1b and the moderately active 2 and 1o ( Table 4). Replacement of the 2-pyridyl terminal group with a 2-pyrimidyl group (1p) resulted in a reduction in potency against L. amazonensis but not against L. donovani (Table 4). In the host cell counterscreen, the mono-AIAs exhibited IC 50 values ranging from 5300 to >50,000 nM against murine J774 macrophages, resulting in selectivity indexes (IC 50 vs J774 macrophages/IC 50 vs Leishmania) of 2.4-100 against intracellular L. amazonensis and 3.1-76 against intracellular L. donovani in vitro.
Among these 18 mono-AIAs, 1b, 1d, and 1k displayed outstanding antileishmanial potency and good selectivity for intracellular Leishmania compared to J774 murine macrophages (selectivity indexes P25), warranting the in vivo evaluation of these derivatives. Compounds 1b, 1d, and 1k were dissolved in water and administered to healthy BALB/c mice for assessment of their in vivo toxicity. Each of these compounds was well tolerated when administered by the i.p. route at 30 mg/kg/day for 5 days and were thus evaluated at this dose in a murine model of visceral leishmaniasis. 4 Animals were infected with L. donovani LV82 promastigotes and then treated with different compounds one week post infection for five consecutive days. These mice were euthanized two weeks post infection and liver smear slides were prepared for the microscopic determination of parasitemia. When given at the dose listed above, administration of 1b, 1d, and 1k resulted in 37%, 13%, and 20% suppression of liver parasitemia, respectively, compared to untreated control groups ( Fig. 2A, B) effective of these three compounds when given i.p., the oral efficacy of 1b was also evaluated in the murine visceral leishmaniasis model. Compound 1b reduced liver parasitemia by 46% at an oral dose of 100 mg/kg/day for five days compared to the control group (Fig. 2C). The in vivo antileishmanial efficacy of 1b is thus similar to that of DB1960 and lower than that of DB766 when the compounds are administered orally at 100 mg/kg in five daily doses. 3,4 When given to infected mice at a dose of 10 mg/kg/day for five days by either the i.p. or oral route, miltefosine administration resulted in >90% inhibition of liver parasitemia, consistent with our previous observations. 4,15 To confirm the exposure and accumulation of 1b in target organs after oral administration, plasma and tissue concentrations of this compound were determined at 1, 2 and 24 h after a single oral dose at 100 lmol/kg (or 40 mg/kg) in mice. Compound 1b accumulated in the target organs such as liver and spleen at higher concentrations (C1h = 12.6 and 4.7 lM, respectively) than in the plasma (C1h = 1.9 lM) (Fig. 3). At 24 h post dose, 1b concentrations decreased substantially to 0.14 lM in the plasma and 0.6-0.7 lM in the liver and spleen.
Our efforts to explore the mechanism of antileishmanial action of the bis-AIAs led to the observation that DB766 is synergistic with posaconazole against Leishmania donovani in vitro 17 ; other investigators reported in vitro antileishmanial synergy between posaconazole and the squalene synthase inhibitor E5700. 18    combination study with 1b and posaconazole was thus performed against intracellular L. amazonensis amastigotes. While azoles have been used to treat CL, 19,20 some of these compounds exhibit low potency against certain Leishmania species in vitro. 21,22 We found this to be true for posaconazole against intracellular L. amazonensis under the conditions of our assay (IC 50 > 10 lM; posaconazole exhibits low solubility at higher concentrations), making standard isobologram analysis impossible in this case. Instead, the method described by Peters et al. 23 was employed to determine whether posaconazole enhances the potency of 1b. Figure 4 demonstrates that this azole augmented the potency of the mono AIA 1b against intracellular L. amazonensis, with 1b displaying an IC 50 value that is Although both assays involved incubating infected murine   peritoneal macrophages with posaconazole for 72 h, de Macedo-Silva exposed peritoneal macrophages to L. amazonensis for a shorter time compared to our studies prior to removing extracellular parasites (2 h vs overnight). Our studies with intracellular L. donovani have indicated that the level of macrophage infection influences the apparent in vitro antileishmanial activity of azoles (A. Joice, unpublished data). Thus, a likely explanation for the discrepancy between IC 50 values for posaconazole against intracellular L. amazonensis in the two studies is the differences in assay conditions employed. In summary, eighteen mono-AIAs were synthesized and evaluated for their antileishmanial activities. Compounds 1b, 1d and 1k displayed the highest potency against intracellular L. amazonensis and intracellular L. donovani amastigotes with their IC 50 values slightly higher than those of amphotericin B and DB766. These three mono-AIAs were well tolerated in healthy mice, indicating that mono-AIAs may be generally less toxic than bis-AIAs. 8,25 Compound 1b shows moderate in vivo antileishmanial efficacy in a mouse model of visceral leishmaniasis. Furthermore, posaconazole enhances the activity of this compound against intracellular L. amazonensis amastigotes, consistent with our observations with the bis-AIA DB766. 17 Mono-AIAs are thus promising antileishmanial leads, and further optimization of this class of compounds to improve their in vivo efficacy is needed. Given the in vitro activity of the compound 1b/posaconazole combination, combining a mono AIA with an azole drug could be a fruitful strategy for the development of antileishmanial therapies.