New Class of Antitrypanosomal Agents Based on Imidazopyridines

Abstract

The present work describes the synthesis of 22 new imidazopyridine analogues arising from medicinal chemistry optimization at different sites on the molecule. Seven and 12 compounds exhibited an in vitro EC₅₀ ≤ 1 μM against Trypanosoma cruzi (T. cruzi) and Trypanosoma brucei (T. brucei) parasites, respectively. Based on promising results of in vitro activity (EC₅₀ < 100 nM), cytotoxicity, metabolic stability, protein binding, and pharmacokinetics (PK) properties, compound 20 was selected as a candidate for in vivo efficacy studies. This compound was screened in an acute mouse model against T.cruzi (Tulahuen strain). After established infection, mice were dosed twice a day for 5 days, and then monitored for 6 weeks using an in vivo imaging system (IVIS). Compound 20 demonstrated parasite inhibition comparable to the benznidazole treatment group. Compound 20 represents a potential lead for the development of drugs to treat trypanosomiasis.

Trypanosomiasis, caused by the unicellular protozoan parasites T. brucei and T. cruzi, are economically significant obstacles to human welfare. Human African Trypanosomiasis (HAT), caused by T. brucei, occurs in Sub-Saharan Africa, while Chagas disease, caused by T. cruzi, is a devastating human disease in The Americas (WHO).¹ Existing medicines for the treatment of these diseases are insufficient, antiquated, toxic, prone to resistance, and require parenteral administration.²⁻⁷

Our focus has been in the identification of compounds containing a benzothiazole core ring. A previous publication reported the antitrypanosomal activities of novel benzothiazole-containing derivatives.⁸ One such analogue, (S)-2-(3,4-difluorophenyl)-5-(3-fluoro-N-pyrrolidylamido)benzothiazole (DAP) represents a new lead against T. brucei (EC₅₀ = 35 nM). These compounds were not tested against T. cruzi in our earlier study.

Guided by these studies,^8,9 we modified the benzothiazole core ring to the imidazo[1,2-a]pyridine ring. This fused ring has been employed in drugs such as antipsychotics, anxiolytics, analgesics, and migraine therapeutics, demonstrating drug-like features associated with the core structure.

The designed library and synthesis were based on suitable modifications of R₁ and R₂ (Table 1) as well as an additional insertion of a nitrogen atom into the imidazopyridine ring at the 6- and 8-positions. With the syntheses of 22 novel compounds, the present work explores the structure–activity relationships (SAR) based on biological assays against T. cruzi, T. brucei, and mammalian cells. The biochemical targets of these compounds are unknown.

Table 1. Antitrypanosomal Activities of Imidazopyridine Derivatives.

			T. cruzi		T. brucei		CLR-8155d,e	Hep G2f,g
compd	R1	R2	EC50 (μM)a	EC50 (μM)a	EC50 (μM)b	EC50 (μM)b	EC50 (μM)	EC50 (μM)
1	piperidin-1-yl	phenyl	2.75	12.28	2.51	4.20	>50.0	>50.0
2	pyrrolidin-1-yl	phenyl	4.46	6.44	2.23	3.27	35.96	>50.0
3	(S)-3-fluoropyrrolidin-1-yl	phenyl	1.08	1.73	0.62	0.77	44.59	>50.0
4	piperidin-1-yl	3-fluorophenyl	3.10	4.53	0.97	1.48	>50.0	>50.0
5	pyrrolidin-1-yl	3-fluorophenyl	2.95	6.01	0.73	0.88	>50.0	>50.0
6	(S)-3-fluoropyrrolidin-1-yl	3-fluorophenyl	0.71	0.88	0.16	0.82	29.74	>50.0
7	piperidin-1-yl	2,3-difluorophenyl	1.00	1.58	0.72	0.87	>50.0	>50.0
8	pyrrolidin-1-yl	2,3-difluorophenyl	1.02	1.61	0.58	2.76	>50.0	>50.0
9	(S)-3-fluoropyrrolidin-1-yl	2,3-difluorophenyl	0.94	1.36	0.18	0.92	>50.0	>50.0
10	piperidin-1-yl	3,4-difluorophenyl	2.84	4.40	1.20	1.75	47.42	>50.0
11	pyrrolidin-1-yl	3,4-difluorophenyl	2.24	2.57	0.32	0.73	>50.0	>50.0
12	(S)-3-fluoropyrrolidin-1-yl	3,4-difluorophenyl	0.39	0.60	0.16	0.31	35.33	>50.0
13	(S)-3-fluoropyrrolidin-1-yl	1,3-thiazol-2-yl	>20.0	>20.0	>20.0	>20.0	>50.0	>50.0
14	(S)-3-fluoropyrrolidin-1-yl	thiophen-2-yl	2.68	3.92	1.00	2.67	>50.0	>50.0
15	2-chloropyridin-3-yl	3,4-difluorophenyl	7.34	>20.0	>20.0	>20.0	>50.0	>50.0
16	3,5-dimethyloxazol-1,2-yl	3,4-difluorophenyl	9.06	>20.0	>20.0	>20.0	>50.0	>50.0
17	pyridin-3-yl	3,4-difluorophenyl	6.27	>20.0	6.30	>20.0	>50.0	>50.0
18	1-methyl, 3-t-butylpyrazol-5-yl	3,4-difluorophenyl	0.57	1.52	0.71	>20.0	>50.0	>50.0
19	(S)-3-fluoropyrrolidin-1-yl	3,4-difluorophenyl	2.53	>20.0	>20.0	>20.0	>50.0	>50.0
20	(S)-3-fluoropyrrolidin-1-yl	3,4-difluorophenyl	0.09c	0.15c	0.02	0.05	>50.0	>50.0
21	1-methyl, 3-t-butylpyrazol-5-yl	3,4-difluorophenyl	0.17c	0.64c	6.99c	>20.0	14.50	>50.0
22	(S)-3-fluoropyrrolidin-1-yl	3,4-difluorophenyl	9.94	>20.0	>20.0	>20.0	>50.0	>50.0

^aThe values are averages of triplicate data. Benznidazole was used as a comparator compound: EC₅₀ and EC₉₀ for T. cruzi of 0.69 ± 0.09 and 1.36 ± 0.20 μM, respectively (n = 5).

^bThe values are averages of triplicate data. Pentamidine was used as a comparator compound: EC₅₀ and EC₉₀ for T. brucei of 0.00216 ± 0.0003 and 0.00613 ± 0.00056 μM, respectively (n = 6).

^cAverage of two tests.

^dHuman lymphoblasts (CRL-8155).

^eQuinacrine was used as a comparator compound: EC₅₀ for CRL-8155 of 4.23 ± 0.97 μM (n = 7).

^fHuman hepatocytes (HepG2).

^gQuinacrine was used as a comparator compound: EC₅₀ for HepG2 of 10.44 ± 1.39 μM (n = 7).

Scheme 1 shows the synthesis^10,11 of the new heterocycles. Assay results are presented in Table 1, and full experimental details are provided as Supporting Information.

Scheme 1. Synthesis of Imidazopyridine Derivatives 1–22.

Reagents and conditions: (a) appropriate diaminopyridine/pyrimidine and bromoacetophenone, NaHCO₃, MeOH, reflux, 12 h; (b) triphosgene, Et₃N, DCM, 0 °C and then appropriate 2° amine, 0 to 25 °C, 15 h, or (c) appropriate carbonyl chloride, Et₃N, DCM, 0 to 25 °C, 16 h. *The urea group was attached at the 7-position, named as X.

The compounds 1–14, 19 and 20 are fused ring ureas bearing various substituents at the 2-position of the imidazole portion of the backbone. Remaining analogues are imidazopyridine amides bearing a 3,4-difluorophenyl substituent on the same imidazole portion. In addition, the internal core of the molecule was modified by insertion of a nitrogen atom at the 6- and 8-positions (20 and 19, respectively)

We evaluated the SAR based on T. cruzi EC₅₀ assays. The effect of fluorination of the aromatic ring attached to the 2-position of imidazopyridine (R₂) was investigated at the 3-position or in the 2,3- and 3,4-difluorophenyl derivatives (1–12). The potency of piperidyl urea 7 was enhanced (less than 3-fold) by introduction of the 2,3-difluorophenyl substituent compared to 1 (phenyl substituent). However, no change in potency was observed by introduction of fluorine at the 3- and 3,4- positions (4 and 10, respectively). A similar pattern was observed for pyrrolidyl urea based analogues. The 2,3-difluorophenyl substituent enhanced the potency compared to the other structurally related analogues (8 vs 2, 5, and 11). Otherwise, the potency of 3-fluoropyrrolidyl urea based analogues 3, 6, 9, and 12 was sequentially enhanced by introduction of a fluorine atom on the phenyl group (3) at the 2,3- (9), 3- (6), and 3,4- (12) positions. 3-Fluoropyrrolidyl urea 12 bearing a 3,4-difluorophenyl group exhibited an EC₅₀ value of 0.39 μM. 3-Fluoropyrrolidyl ureas were more potent than the corresponding pyrrolidyl and piperidyl ureas in all cases.

Further modifications were made to the R₂ portion while retaining the 3-fluoropyrrolidyl urea portion of the molecule. Replacement of 3,4-difluorophenyl with thiazolyl and thiophenyl resulted in a 51-fold and 7-fold loss of activity (12 vs 13 and 12 vs 14, respectively).

The presence of the 3,4-difluorophenyl substituent on the imidazopyridine system with a 3-fluoropyrrolidyl urea led to activity improvements. Replacements of the R₁ with chloropyridinyl, dimethyloxazolyl, and pyridinyl amide groups considerably diminished the potency by 19-fold, 23-fold, and 16-fold, respectively. Only the 1-methyl, 3-t-butylpyrazolyl amide 18 was nearly as potent as 3-fluoropyrrolidyl urea 12.

A basic nitrogen atom inserted at the 8-position of the imidazopyridine ring of 19 resulted in a >6-fold loss in potency compared to compound 12. Opposite to this, the potency of 20 was enhanced more than 4-fold by a nitrogen atom inserted at the 6-position. A similar pattern was observed for 1-methyl, 3-t-butylpyrazolyl amide pairs 18 vs 21. The amide analogue 21 was similar in potency to urea analogue 20, which fits the SAR pattern reported above (amide 18 vs urea 12 analogues). In addition, the effect of a nitrogen atom inserted at the 6-position enhanced potency. The 3-fluoropyrrolidyl 20 bearing a 3,4-difluorophenyl was the most potent analogue against T. cruzi in this series of compounds (EC₅₀ = 93 nM).

For these imidazopyridine analogues, similarities and differences were seen for the SAR toward T. cruzi and T. brucei. For the T. brucei SAR, a greater enhancement in potency compared to T. cruzi was observed upon fluorination of the aromatic ring (R₂) at any position relative to 3-position or in the 2,3- and 3,4-difluorophenyl analogues (1–12). For example, with T. brucei, the pyrrolidyl urea 11 bearing a 3,4-difluorophenyl group exhibited an EC₅₀ value of 0.32 μM, seven times more potent that for T. cruzi.

The effect of fluorination of the aromatic ring was slightly different for piperidyl urea (1 vs 4, 7, and 10) when comparing T. brucei vs T. cruzi SAR. While potency against T. brucei sequentially decreased by introduction of a fluorine atom on the piperidyl ureas in the order: 2,3- > 3- > 3,4- > phenyl group; the T. cruzi potency for the same analogues sequentially decreased in the order: 2,3- > phenyl group > 3,4- > 3-positions. For pyrrolidyl ureas (2 vs 5, 8, and 11) the effect of fluorination was more effective compared to piperidyl ureas with both parasites. Incremental enhancements in potency were achieved by introduction of a fluorine atom (5) and double fluorination (8 and 11). Compound 2 (phenyl as a substituent) was the less potent pyrrolidyl urea analogues. Surprisingly, the T. brucei potencies of 3-fluoropyrrolidyl urea analogues did not change with a fluorine atom at the 3-position (6, EC₅₀ = 0.16 μM) or with the 2,3- (9, EC₅₀ = 0.18 μM) and 3,4- (12, EC₅₀ = 0.16 μM) difluorophenyl derivatives.

Replacement of the urea with an amide substantially decreased the T. brucei potency or resulted in inactivity in all cases (12 vs 15–17 and 20 vs 21), except 12 vs 18 (less than 5-fold). 3-Fluoropyrrolidyl urea analogues led to increased T. brucei potency. However, the same pattern was not true for T. cruzi. Compound 21 (R₁ = 1-methyl, 3-t-butylpyrazolyl and R₂ = 3,4-difluorophenyl) showed a T. cruzi EC₅₀ of 0.17 μM, 40 times more potent than the value for T. brucei EC₅₀ of 6.99 μM.

The urea group attached to the 7-position of the fused ring system proved to be essential for activity. Compound 22 is a regioisomer of 12 in which the urea is attached to the 6-position instead of the 7-position. The relocation of the urea moiety resulted in loss of activity (12 vs 22).

In the final T. brucei SAR analysis, inserting a nitrogen at the 8- and 6-position of the imidazopyridine urea (R₁ = 3-fluoropyrrolidyl and R₂ = 3,4-difluorophenyl) resulted in the inactive compound 19 and the most active compound 20. Compound 20 exhibited an EC₅₀ value of 18 nM (2-fold greater than existing DAP compounds).⁸

All compounds were tested for toxicity against human lymphocytes CRL-8155 and human hepatocytes HepG2. Only compounds 2, 3, 6, 10, 12, and 21 exhibited toxicity to the CRL-8155 cell line below the highest concentration tested of 50 μM (EC₅₀ range from 14.50 to 47.42 μM). The remaining compounds exhibited no detectable toxicity to either cell line.

The selectivity ratio of compounds for T. cruzi and T. brucei parasites over each of the two cell lines was calculated (see Table S2 of the Supporting Information). As an example, compound 20 exhibited a selectivity index (SI) >500 for T. cruzi and >2500 for T. brucei against either cell line; while compound 21 exhibited an SI > 290 and 80 for T. cruzi as compared to HepG2 and CRL-8155 cell lines, respectively; and poor selectivity for T. brucei against either cell line.

Selected compounds were assayed for stability to pooled mouse liver microsomes (Table 2). The half-lives of compounds 6, 9, 12, and 20 were greater than 60 min with a range of 84–100% of the test compounds remaining at the 60 min time point. Compound 21 had 47% remaining at the 60 min time point. The introduction of an additional aromatic fluorine atom (6 vs 9 or 12) offered no change in metabolic stability. All the selected compounds demonstrated excellent metabolic stability when incubated with mouse liver microsomes.

Table 2. Stability of Selected Compounds to Mouse Microsomes, Solubility, and Protein Binding^c.

	T. cruzi	T. brucei	mouse microsomesa	solubility (μM)			protein bindingb
compd	EC50 (μM)	EC50 (μM)	t1/2 (min)	pH 7.4	pH 6.5	pH 2.0	rec. (%)	FU (%)
6	0.71	0.16	>60 (84.5%)	49.7	43.8	52.0	68.5 ± 9.4	3.0 ± 0.3
9	0.94	0.18	>60 (86.3%)	6.4	6.8	53.1	68.4 ± 8.5	2.6 ± 0.2
12	0.39	0.16	>60 (100%)	32.8	49.9	45.9	64.5 ± 8.1	4.0 ± 0.7
19	2.53	>20.0		41.5	42.0	47.5	49.0 ± 1.7	8.1 ± 0.4
20	0.09	0.02	>60 (96.0%)	2.1d	2.4d	1.6d	56.1 ± 4.8	9.9 ± 0.9
21	0.17	6.99	60 (47.1%)	2.7	7.6	1.3	73.3 ± 11.2	0.7 ± 0.1

^aThe values are averages of duplicate data. Controls for mouse microsomes are included in each assay shown as average ± SEM: testosterone t_1/2, 5.68 ± 0.81 min (n = 4); dextromethorphan t_1/2, 11.11 ± 0.67 min (n = 4).

^bThe control propranolol was included in all protein binding experiments with % recovery and %FU average ± SEM as follows: % recovery 75.74 ± 3.76% (n = 12) and %FU 11.43 ± 1.88% (n = 12).

^cEC₅₀ values from the Table 1.

^dAverage of two tests.

Solubility is important for in vivo dosing formulation and bioavailability,¹² and fluorination is a well-known strategy for improving the bioavailability of drug molecules.¹³ The solubility studies were performed on selected compounds (Table 2). The experiments were carried out at three different pH values (7.4, 6.5, and 2.0), corresponding to the pH of the blood, small intestine, and stomach, respectively. Compounds 6, 9, and 12 showed higher solubility than analogues 20 and 21 at all pH conditions.

Compound 20 had a nitrogen atom embedded in the 6-position in the imidazopyridine ring resulting in a decrease in solubility at all pH conditions tested (20 vs 12). In contrast, the insertion of a nitrogen atom at the 8-position gave modest improvements for the solubility at pH 7.4 and 2.0 (19 vs 12), but caused compound 19 to suffer a greater reduction in activity when compared to compound 20.

In order to help understand ADME (absorption, distribution, metabolism, and excretion) properties and to aid in lead compound selection, plasma protein-binding experiments were carried out to determine the unbound drug fraction. Selected compounds 6, 9, and 12 showed the percent fraction unbound in 100% mouse plasma of 3.0%, 2.6%, and 4.0%, respectively (Table 2). Compound 21 exhibited the lowest unbound fraction (FU = 0.7%). The greatest unbound fraction was observed for the most active compound 20 (FU = 9.9%). The large %FU of compound 20 likely led to its increased efficacy in vivo (described below) and marks it as a promising compound for the treatment of T. cruzi and T. brucei infections.

Studies were performed with compounds 12 and 20 to see if they crossed the blood–brain barrier in order to ascertain if they could be candidates for treatment of second stage HAT infections.¹⁴ Groups of three mice were given single 5 mg/kg ip doses of the test compounds and were sacrificed 1 h postdose, at which time brains and plasma were collected for analysis. Compound 12 exhibited a mean brain concentration of 0.55 μM and a mean plasma concentration of 4.84 μM. Compound 20 had a mean brain concentration of 1.12 μM and a mean plasma concentration of 1.31 μM. This gave a brain/plasma ratio of 0.114 for compound 12 and 0.856 for compound 20, a 7.5-fold change. This large change is related to the only difference between compounds 12 and 20, i.e., the exchange of a carbon for a nitrogen atom at the 6-position of the imidazopyridine ring.

Pharmacokinetic studies were performed in mice to evaluate if compound 20 was suitable for in vivo efficacy studies. Mice (n = 3) were dosed orally with 50 mg/kg of compound with blood sampled at time points up to 24 h. Compound 20 had an (average ± SEM) C_MAX of 5.7 ± 0.9 μM, and an AUC of 3585 ± 670 min·μM.

Based on its promising in vitro activity, cytotoxicity, metabolic stability, protein binding, and PK properties, compound 20 was selected as a candidate for in vivo efficacy studies. This compound was screened in an acute model against T.cruzi (Tulahuen strain, DTU VI). Mice were dosed orally with 50 mg/kg of compound twice a day for 5 days (days 7–11 postinfection). This is a relatively short duration of treatment to establish proof of concept and is usually not expected to be sufficient to cure mice of parasitemia during acute infection (as can be seen by the benznidazole control group).¹⁵ The mice were infected subcutaneously (SC) on day 0 with 2 × 10⁴ trypomastigotes expressing a red-shifted luciferase gene.¹⁶ The parasites were imaged using an IVIS imaging system at various time points out to 6 weeks (42 days). The substrate d-luciferin potassium salt was injected SC into mice, followed by imaging 10–15 min postinjection. Images were taken for each set of mice both dorsally and ventrally. Autoexposure settings were used with a maximum exposure time set to 5 min. The signal from these images was combined and standardized by converting the readouts to radiance (photons/sec/cm²/sr). As seen in Figure 1, compound 20 led to similar levels of parasite suppression as the clinical drug benznidazole, warranting further future studies with a longer treatment duration and against a chronic model of T.cruzi infection. No toxicities to the mice were observed during the observation period.

Figure 1.

(A) Ventral images of balb/c mice infected with bioluminescent PpyRE9h-expressing T.cruzi before and after treatment with compound 20, benznidazole control (BNZ), or vehicles. Mice were injected with 120 mg/kg of d-luciferin, anesthetized, and imaged using an IVIS imaging system. Both benznidazole and compound 20 showed significant suppression of T.cruzi infection. The vehicle groups had to be removed from the experiment on day 13 due to advanced infection (one mouse from the vehicle (BNZ) group had to be removed on day 9). (B) Radiance signal of treatment groups (dorsal + ventral) throughout the course of the experiment. Dosing window highlighted in gray. The decreased signal between day 24 and day 42 is likely due to the infection advancing from an acute to a chronic infection.

Compound 20 levels in plasma were measured at the time of the seventh dose. Specifically, tail blood was collected immediately before, 1 h, and 6 h after compound administration. At 1 and 6 h postdose, the concentrations were found to be 9.0 ± 1.4 μM (n = 3) and 5.8 ± 1.9 μM (n = 3), respectively. The predose concentration of compound 20 was detectable in only one of three mice at 4.8 μM. The data demonstrate that through repeated dosing, compound 20 achieves concentrations in the blood similar to that seen in the single dose PK study. Slightly higher C_MAX values after repeated dosing in mice could be due in part to saturated clearance mechanisms.

In summary, the imidazopyridine-containing compounds have potent activity on parasite cultures, low cytotoxicity, excellent metabolic stability, and very good plasma exposure upon oral dosing in mice. The activity of compound 20 leading to substantial suppression of parasite infection in mice warrants further development as antitrypanosomal agents.

Glossary

ABBREVIATIONS

ADME

absorption, distribution, metabolism, and excretion

BPR

brain to plasma ratio

Cmpd

compound

DAP

(S)-2-(3,4-difluorophenyl)-5-(3-fluoro-N-pyrrolidylamido)benzothiazole

FU

fraction of unbound drug

HAT

Human African Trypanosomiasis

IVIS

in vivo imaging system

PK

pharmacokinetics

Rec

recovery

SAR

structure–activity relationship

SI

selectivity index

SC

subcutaneously

T. brucei

Trypanosoma brucei

T. cruzi

Trypanosoma cruzi

WHO

World Health Organization

Supporting Information Available

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsmedchemlett.7b00202.

• Full experimental details for synthesis and assays (PDF)

• Molecular formula strings and assays (XLS)

Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.