Substituted 2-Phenyl-Imidazopyridines: A New Class of Drug Leads for Human African Trypanosomiasis

Abstract

A phenotypic screen of a compound library for antiparasitic activity on Trypanosoma brucei, the causative agent of human African trypanosomiasis, led to the identification of substituted 2-(3-aminophenyl) oxazolopyridines as a starting point for hit-to-lead medicinal chemistry. A total of 110 analogues were prepared, which led to the identification of 64, a substituted 2-(3-aminophenyl) imidazopyridine. This compound showed antiparasitic activity in vitro with an EC₅₀ of 2 nM and displayed reasonable drug-like properties when tested in a number of in vitro assays. The compound was orally bioavailable and displayed good plasma and brain exposure in mice. Compound 64 cured mice infected with Trypanosoma brucei when dosed orally down to 2.5 mg/kg. Given its potent anti-parasitic properties and its ease of synthesis, compound 64 represents a new lead for the development of drugs to treat human African trypanosomiasis.

INTRODUCTION

Two protozoan agents cause Human African trypanosomiasis (HAT), Trypanosoma brucei gambiense and Trypanosoma brucei rhodesiense. Natural transmission occurs in 36 countries of sub-Saharan Africa by the bite of infected tsetse flies (WHO Fact sheet #259). After the initial cutaneous inoculation, early-stage (hemolymphatic) infection occurs. In the form occurring in West and Central Africa, (Gambian HAT), the first stage can last for months to years before progressing to second state infection in the central nervous system. In the form occurring in East and Southern Africa (Rhodesian HAT), the first stage lasts only a few weeks to months before causing late-stage disease. Once the parasites enter the central nervous system (second stage), the patients suffer neurological effects that culminate in coma and death if untreated.

Better HAT drugs are desperately needed especially for second stage HAT.¹ The recent introduction of nifurtimox-eflornithine combination therapy for Gambian HAT allows for the use of fewer doses of eflornithine, but these still need to be given by IV injections, and the treatment remains very expensive. Melarsoprol, a trivalent arsenical, is the only drug effective for second stage Rhodesian HAT. Unfortunately, toxic encephalopathy occurs in 5–18% of patients with associated mortality of 3–12% from treatment itself. Because second state HAT has 100% mortality, melarsoprol remains as first line treatment despite this toxicity. An additional concern is the increasing incidence of melarsoprol resistant strains.²

Better drugs are also desirable for treating first stage disease. Current treatment options pentamidine and suramin must be given by injection though most patients with early stage HAT are able to take oral medicines. Both drugs have considerable toxicity, and are unsafe in pregnancy.^3,4

Two promising drug candidates have recently entered early clinical development for treating HAT: a nitroheterocyle, fexinidazole, and a benzoxaborole, SCYX-7158.^6,7 These candidates follow on the heels of the diamidine drug, pafuramidine (DB-289), where efficacy was established in Phase III studies of HAT, however, nephrotoxicity observed in an expanded Phase I safety trial led to abandonment.⁸ As evidenced by pafuramidine, the attrition rate of clinical drug candidates is high (nearly 90% upon entering Phase I, ⁹ underscoring the importance of maintaining an adequate pipeline of drug candidates.

RESULTS AND DISCUSSION

A high throughput screen was designed to identify new small molecules with antiparasitic activity toward T. brucei within a library of 700,000 compounds. This library was assembled with a particular focus on drug-like properties and structural diversity. The library had been previously profiled in more than 60 other high throughput screens, which included both biochemical and cell-based assays against human and pathogen targets, and has been an excellent source of attractive starting points for medicinal chemistry in these other programs. Further, the screen history of this library also allowed rapid identification and elimination of compounds with a ‘frequent hitter’ profile.

Compounds from this library were tested for inhibitory activity on T. brucei at 3.6 IM. The screen yielded 3,889 primary hits (0.6% hit rate) that inhibited growth by > 50%. Data from more than 95% of the assay plates had a Z′ > 0.6, using DMSO as the negative control and 1 IM pentamidine as the positive control. Primary hits from the screen were further characterized using a dose-response assay format to determine the EC₅₀. In parallel, cytotoxicities of these compounds were determined against a proliferating human hepatoma cell line (Huh7). The final set of confirmed hits consisted of compounds that had EC₅₀ < 3.6 IM against T. brucei, as well as a limited or non-observable Huh7 cytotoxicity (CC₅₀ > 10 IM or SI > 10; SI = CC₅₀ / EC₅₀). The final set of confirmed T. brucei hits consisted of 1,035 inhibitors. Of this set, over 95% of the identified hits did not carry a “frequent hitter” liability (identified as a hit in > 5 screens out of a total of 60–65 screens). The 1,035 confirmed and selective hits could be grouped into about 115 distinct scaffolds, with 144 compounds having a EC₅₀ on T. brucei less than 100 nM, and a further 446 compounds having EC₅₀ less than 500 nM.

Properties of lead compound 1

Lead compound 1 was selected from the available hits based on ease of synthesis, including a lack of stereogenic carbons, as well as drug-like features including low MW and Lipinski rule-of-five compliance. Compound 1 blocks the growth of T. brucei in vitro with an EC₅₀ of 220 nM (Table 1). All compounds in this study were tested for anti-parasite activity on the cattle virulent T. brucei brucei strain BF427 in vitro, and selected examples were also tested on the human virulent T. brucei rhodesiense strain STIB900. In all cases, EC₅₀s for both parasite strains were within a factor of 2–3 of each other. Compound 1 was not cytotoxic to mammalian cells as the EC₅₀ was >50 μM for rat myoblasts (L6 cells), human lymphoblasts (CRL-8155 cells) and human hepatocytes (HepG2 cells). The compound was reasonably soluble in aqueous buffer (20 μM at pH 7.4 and 22 μM at pH 2.0) (Supplemental Table 1).

Table 1.

2-(3-Aminophenyl)-oxazolopyridines containing a N-linked 2-furanoyl group.

Compound	R1	R2	1 EC50 (μM)	2 CC50 (μM)
1	H	Cl	0.22	>50a,b,c
2	H	F	0.12	>50a
3	H	Me	0.43	>50a
4	H	CN	0.4	>50a
5	H	OMe	>10	>50a
6	H	4-Fluorophenyl	>50	>50a
7	H	3,4-Difluorophenyl	>50	>50a
8	H	4-Pyridyl	>20
9	Br	F	0.23	>50a
10	CN	F	0.38	>50a
11	Phenyl	F	0.04 (0.08)	7a
12	4-Fluorophenyl	F	0.05 (0.1)	>50b,32c
13	3-Chlorophenyl	F	0.05	>50a
14	4-MeO-Phenyl	F	0.17
15	4 Phenylphenyl	F	0.42	>50a
3 Pentamidine			1.7
4 Quinacrine				3.9b 9.8c
5 Podophyllotoxin				0.014a

¹Concentration of compound required to inhibit growth by 50% (EC₅₀) of T. brucei strain BF427 or T. brucei rhodesiense (numbers in parentheses).

²Concentration of compound required to inhibit growth by 50% (CC₅₀) of mammalian cell lines:

^arat myoblasts (L6),

^bhuman lymphoblasts (CRL-8155) or

^chuman hepatocytes (HepG2).

³Control compound for T. brucei assays: average EC₅₀ was 1.74 +/− 0.15 SEM (n=57 assays).

^4,5Control compounds for cytotoxicity assays: average CC₅₀ and SEM for CRL-8155 cells was 3.92 μM +/− 0.91 μM (n=8 assays), for HepG2 cells was 9.78 μM +/− 1.27 μM (n=10 assays), and for L6 myoblasts was 0.014.3 μM +/− 0.00065 μM (n= 15 assays).

Supplemental Table 1 summarizes in vitro ADME studies and pharmacokinetic (PK) studies in mice. When injected IP into mice, analysis of brain and plasma compound levels 15 and 60 min post-injection showed a brain-to-plasma concentration ratio of 0.54 and 0.38, respectively, indicating exposure into the central nervous system. We also measured permeability of 1 across a tight monolayer of Madin-Darby canine kidney cells that have been transfected with the human MDR1 (P-glycoprotein) drug efflux pump gene (MDCK-MDR1 cells), and permeability was measured in the presence and absence of the MDR1 inhibitor GF-120918 to gauge whether 1 is a substrate for the efflux pump ^10–12. Values of P_app in the absence and presence of GF-120918 were 417 nm/sec and 362 nm/sec, respectively. The corresponding value for propanolol, a readily permeable, non-MDR1 substrate comparator compound was 444–469 nm/sec in the presence and absence of GF-120918. The additional comparator compound Amprenavir is highly permeable by passive diffusion and is also a potent MDR1 substrate and shows P_app values of 252 and 27 nm/sec in this assay. The high permeability of 1 and its poor tendency to be pumped by MDR1 in this in vitro assay is consistent with central nervous system exposure in mice.

A liability of 1 is its poor metabolic stability in liver microsomes. The half-life for loss of 1 in mouse and human liver microsomes was 0.8 min and 6.4 min, respectively. A PK study was carried out by dosing 1 by oral gavage in mice at 50 mg/kg. The C_max value in plasma was low, 0.06 μM (T_max was 30 min), the area under the plasma concentration-time curve (AUC) was 187 min*μM, the half-life for elimination from the plasma was 53 min, and the plasma clearance was 98 mL/min/kg. The low exposure of 1 in mice is consistent with the high rate of metabolism seen in liver microsomes. In vitro metabolite identification studies in liver microsomes showed a major metabolite due to amide hydrolysis of the 4-chloroanilide group (not shown). Given this result, we tested the hydrolysis product (the aniline derived from 1) and found no inhibition of T. brucei growth in vitro when tested up to 20 μM. Thus, the anilide and not its hydrolysis product, was causing the antiparasitic activity. It could be possible that 1 kills parasites by acylating its biological target (given that the 4-chloroaniline is a relatively good leaving group), however, studies with analogues where the anilide is replaced with a urea (described below) suggest that this mechanism is unlikely. In summary, 1 was selected for hit-to-lead optimization because of good whole-cell activity on T. brucei, no cytotoxicity on mammalian cell lines, reasonable exposure into the central nervous system, and acceptable solubility. Moving forward, the goal was to design and synthesize chemical analogues to further optimize the anti-trypanosomal potency while improving metabolic stability.

Synthesis of 1 and its analogues

Scheme I shows the synthetic route to make 1 and its analogues (substituted 2-(3-aminophenyl)-oxazolopyridines). The substituted 3-nitrobenzoic acid is condensed with 2-amino-3-hydroxypyridine to generate the oxazole ring. When R¹ is chloro, it can be replaced with phenyl or substituted phenyl by Suzuki cross coupling (not shown). Reduction of the nitro group is followed by acylation.

Scheme I.

Synthesis of substituted 2-(3-aminophenyl)-oxazolopyridines. (a) 2-Amino-3-hydroxypyridine, polyphosphoric acid, 160 °C, 5 h; (b) SnCl₂.2H₂O, EtOAc, reflux; (c) R²COCl, pyridine, cat. DMAP, rt or R²CO₂H, EDCI, HOBt, CH₂Cl₂, rt.

Scheme II shows the preparation of compounds containing a phenyl substituent at the 6-position of the oxazolopyridine ring. 2-Amino-3-hydroxy-5-bromo-pyridine is made in 3 steps and then converted into the oxazolopyridine as in Scheme I. Bromine at the 6-position is replaced by phenyl or substituted phenyl by Suzuki cross coupling.

Scheme II.

Synthesis of substituted 2-(3-aminophenyl)-oxazolopyridines containing 6-substituents. (a) N, N′-Disuccinimidyl carbonate, THF, rt; (b) DMF, Br₂, rt; (c) NaOH, H₂O, reflux 5 h; (d) as in Scheme I; (e) K₂CO₃, R³-boronic acid, tetrakis(triphenylphosphine)Pd(0) (catalytic), toluene, ethanol, reflux.

Structure-activity studies of analogues of 1

Table 1 shows a number of analogues of 1 and their antiparasitic activity on T. brucei in vitro. Replacement of the aryl chlorine (R²) with fluorine (2) leads to a 2-fold improvement in potency, in contrast replacement with methyl (3) or cyano (4) reduces potency about 2-fold, and replacement by methoxy (5) abrogates activity. Replacement of chloro with various aryl groups (6–8) abrogates anti-parasite activity.

When the 6-position (R¹) of the oxazolopyridine ring in compound 2 is substituted with Br or CN there is a 2–3 fold decrease in potency (9,10) with respect to 2 but in contrast a phenyl group improves potency, with or without halogen substitution at the 3 or 4 position of the phenyl ring (11, 12, 13). When the phenyl is substituted with a 4-methoxy there is no potency gain (14) and larger substitutions lead to a decrease in potency (15).

Table 2 shows the effect of substituting or replacing the 2-furanoyl group with other moieties. While methylation of the 2-furan ring at the 3- and 5-position is tolerated (16, 17) larger substitutions such as benzofuran (18) are not tolerated. The 3-furoyl group is equally tolerated (19) but other amides including aliphatics, aromatics and heteroaromatics lead to a mild to profund decline in activity (20–29) including closely related tetrahydrofuran (22) and oxazole (24). Also inactives are sulfonamides (30, 31), while thioamide retains some activity (32). Interestingly, while methylation at the amide nitrogen is not tolerated (33), acylation is (34, 35). Complete removal of the amide (36) and both mono benzylation or di-benzylation lead to loss and decrease of activity respectively (37, 38). In an effort to identify additional R¹ moieties with increase metabolic stability, the replacement of the furanoyl amide with ureas was explored. Terminally mono-substituted and di-substituted ureas (39–44) were investigated and we discovered that some of the di-substutued ureas are tolerated with the pyrrolidine urea (43) showing slightly better potency than the similar 2-furanoyl derivatives.

Table 2.

2-(3-Aminophenyl)-oxazolopyridines containing variations of the N-linked group.

Compound	R1	R2	R3	1EC50 (μM)	2 CC50 (μM)
16	5-Methylfuran-2- carbonyl	H	F	0.20	>50a
17	3-Methylfuran-2- carbonyl	H	F	0.1	>50a
18	Benzofuran-2- carbonyl	H	F	>10	>50a
19	3-Furanoyl	H	F	0.15	>50a
20	Acetyl	H	F	>20	>50a
21	Trifluoroacetyl	H	Cl	(>50)
22	Tetrahydrofuran-2- carbonyl-	H	F	>10
23	Benzoyl	H	Cl	7.1	>50a
24	Oxazole-5-carbonyl-	H	F	1.9	>50a
25	2-Thiophenoyl	H	Cl	1.5	>50a
26	3-Pyridine-carbonyl-	H	F	7.0	>50a
27	1-Methyl-1H- imidazole-5-carbonyl	H	F	>20	>50a
28	Pyrazine-2-carbonyl-	H	F	0.9	>50a
29	N-Methyl-pyrrole-2- carbonyl-	H	F	1.1	>50a
30	Methylsulfonyl	H	Cl	6.1	>50a
31	Phenylsulfonyl	H	Cl	>20	>50a
32	2-Furancarbothioyl-	H	F	0.41	>50a
33	2-Furanoyl	Methyl	F	>10	>50a
34	2-Furanoyl	2-Acetyl	F	0.5
35	2-Furanoy	2-Furanoyl	F	0.12	>50a,b,c
36	H	H	Cl	>20
37	Benzyl	H	F	>20
38	Benzyl	Benzyl	F	1.1	>50a
39	Methylcarbamoyl	H	F	3.8	>50a
40	Isopropylcarbamoyl-	H	Cl	1.0	>50a
41	Phenylcarbamoyl-	H	Cl	12.0	>50a
42	Dimethylcarbamoyl-	H	Cl	0.4	>50a
43	1-Pyrrolidinoyl-	H	Cl	0.09	>50b,c
44	1-Piperidinoyl-	H	Cl	1.9

¹Concentration of compound required to inhibit growth by 50% (EC₅₀) of T. brucei strain BF427 or T. brucei rhodesiense (numbers in parentheses).

²Concentration of compound required to inhibit growth by 50% (CC₅₀) of mammalian cell lines:

^arat myoblasts (L6),

^bhuman lymphoblasts (CRL-8155) or

^chuman hepatocytes (HepG2).

Controls shown at bottom of Table 1.

Additional structure-activity relationships came from study of the analogues shown in Scheme III where changes at the core were explored. The results show that the oxazolopyridine bicycle cannot be easily replaced with a monocycle (45–49). Moreover, the N at the 4-position of the oxazolopyridines is essential for activity as it cannot be replaced with C or moved to the 5-position (50–57).

Scheme III.

An alternative to the oxazolopyridine core came by replacing the oxygen to a nitrogen in the bicycle leading to the imidazopyridine core (Table 3). These analogues were made as in Scheme I starting from the appropriately substituted 2,3-diaminopyridines (not shown). Compound 58 showed comparable potency to 2 and chloro and fluoro substitution at C6 (R²) of the core was also beneficial (59, 60). As expected from the previous SAR, chloro substitution at C5 (61) led to an inactive compound and it was shown that chloro substitution was also tolerated at C7 (62). In general the SAR for the imidazopyridines was found to be similar to that for the oxazolopyridines (only selected compounds are shown). The pyrrolidine urea was a good replacement for the furanoyl amide as was the case for the oxazolopyridines (63). An expected improvement in potency was further achieved by adding a phenyl group at the 6 position of the core which led to compound 64 with an EC₅₀ against T. brucei of 2 nM. A wide variety of substituents can be added to this 6-phenyl group, and the phenyl can also be replaced with several heterocyclic group with maintenance of anti-parasite potency (65–86).

Table 3.

2-(3-Aminophenyl)-imidazopyridines containing a N-linked 2-furanoyl group.

Compound	R1	R2	1 EC50 (μM)	2 CC50 (μM)
58	2-Furanyl	H	0.2
59	2-Furanyl	Cl	0.07	>50 a, c 1.7 b
60	2-Furanyl	F	0.2
61	2-Furanyl	5-Cl	10
62	2-Furanyl	7-Cl	0.12
63	N-Pyrrodinyl	Cl	0.05	15 a, >50b,c
64	N-Pyrrodinyl	Phenyl	0.002	24 b, 30 c
65	N-Pyrrodinyl	3-Methoxy-phenyl	0.01	3.6 b, 35 c
66	N-Pyrrodinyl	2-Methoxy-phenyl	0.005	13 b, >50c
67	N-Pyrrodinyl	3-Cl-phenyl	0.002	2.8 b,>50c
68	N-Pyrrodinyl	2-chlorophenyl	0.005
69	N-Pyrrodinyl	3-acetylphenyl	0.003
70	N-Pyrrodinyl	3-Me-phenyl	0.002	5 b, 10 c
71	N-Pyrrodinyl	3-Trifluoromethoxy-phenyl	0.03	11 b, >50c
72	N-Pyrrodinyl	3-Methyl-4-fluoro-phenyl	0.02	>50 b
73	N-Pyrrodinyl	3-NH2-Phenyl	0.01	5 b, 40 c
74	N-Pyrrodinyl	3-Furanyl	0.004	25 b, >50c
75	N-Pyrrodinyl	3-Thiophenyl	0.002	6 b, >50c
76	N-Pyrrodinyl	2-Thiophenyl	0.004	9 b
77	N-Pyrrodinyl	3-Pyridyl	0.005	14 b,>50c
78	N-Pyrrodinyl	5-(2-chloropyridyl)	0.01	3.8b, >50c
79	N-Pyrrodinyl	4-(2-chloropyridyl)	0.004	>50b,c
80	N-Pyrrodinyl	5-(3-methylpyridyl)	0.01	>50b,c
81	N-Pyrrodinyl	5-(2-methoxypyridyl)	0.005	>50b,c
82	N-Pyrrodinyl	5-(3-pyrrolidino)	0.02	8b, >50c
83	N-Pyrrodinyl	5-(3-chloropyrimidinyl)	0.005	38b, >50c
84	N-Pyrrodinyl	5-pyrimidinyl	0.03	12b, >50c
85	N-Pyrrodinyl	5-(2-methoxypyrimidinyl)	0.06	3.9 b, 9.8 c
86	N-Pyrrodinyl	5-(2-chloropyrimidinyl)	0.04	14.3a

¹Concentration of compound required to inhibit growth by 50% (EC₅₀) of T. brucei strain BF427 or T. brucei rhodesiense (numbers in parentheses).

²Concentration of compound required to inhibit growth by 50% (CC₅₀) of mammalian cell lines:

^arat myoblasts (L6),

^bhuman lymphoblasts (CRL-8155) or

^chuman hepatocytes (HepG2).

Controls shown at bottom of Table 1.

Compound 64 was studied in detail. It produced a very steep dose response curve for antiparasitic activity on T. brucei in vitro with an EC₉₀ of 2 nM, and complete growth inhibition was observed. Cytotoxicity toward mammalian cells was minimal (24–30 μM, Table 3). Along with other analogues, it was characterized in in vitro ADME assays (Supplemental Table 1) and showed aqueous solubility at pH 7.4 of 8.1 μM which increased to 32 μM at pH 2.0, presumably because of imidazole ring protonation. Compound 64 was metabolized upon incubation with mouse and human liver microsomes in vitro with half-lives of 10 and 31 min, respectively. This represented a significant improvement in metabolic stability over the initial lead 1, presumably because the latter is more sensitive to hydrolysis. Compound 64 was a modest inhibitor of hepatic cytochrome P450 3A4-isoform in vitro with an IC₅₀ of 1.5 μM. Oral gavage dosing of 64 in mice at 50 mg/kg gave a maximal concentration in the plasma (C_max of 4.3 μM) with a time to C_max (T_max) of 30 min. The plasma concentration versus time curve is provided as Fig. 1 of Supporting Information. The area under the plasma concentration versus time curve (AUC) after oral dosing was 1125 min*μM, the half-life for elimination from the plasma was 204 min, and the plasma clearance (Cl_p) was mL/min/kg. The in vivo clearance was substantially lower than what expected from the in vitro metabolism in liver microsmes likely due to the high plasma protein binding (95% in mouse plasma). Compound 64 showed excellent permeability across MDCK-MDR1 cells and was not a significant substrate for the MDR1 efflux pump (P_app = 401 and 503 (+ MDR1 inhibitor) nm/sec). Consistent with this is the significant distribution of 64 into mouse brain, with a brain-to-plasma concentration ratio 60 min after IP dosing of 0.5 and 1.1 (2 independent mouse experiments). PK studies of 64 in rats are summarized in Supporting Materials Fig. 2 and Table 2. Briefly, rats were dosed by IV injection and by oral gavage and the bioavailability of 64 was 35% upon delivery of a 21.3 mg/kg oral dose. The apparent volume of distribution during steady-state (V_SS) was 2.4 L/kg. PK data in mice for additional compounds is provided in Supporting Material Table 2.

Analysis of efficacy of 64 in parasite-infected mice

Given the good physiochemical and PK properties of 64, we decided to test it for efficacy in a mouse model of acute HAT infection. Mice were injected with T. brucei rhodesiense on day 0, and 64 was given by oral gavage 48 h after parasite injection. Mice were dosed for a total of 5 days using the dosing regimens shown in Table 4. Parasitemia in the blood and survival were followed out to 60 days. Cures in all mice were obtained with once and twice dosing at 20 and 50 mg/kg (Table 4). With 5 mg/kg dosing, all mice were cured with twice daily dosing, and 2 of 5 mice were cured with once daily dosing. Twice daily dosing at 2.5 mg/kg cured 2 of 5 mice, and twice daily dosing with 1 mg/kg cured 0 of 5 mice. These results show that 64 is highly efficacious in a mouse model of acute trypanosomiasis.

Table 4.

Efficacy of 64 in T. brucei infected mice. Mice were infected on Day 0 and treated for five consecutive days beginning 48 hours post-infection when patent parasitemia was detectable.

Exp	Dose (mg/kg)	Doses per day	Cured	Day of euthanasia*
1	50	2	5/5	60
1	20	2	5/5	60
1	5	2	5/5	60
1	50	1	5/5	60
1	Veh	2	0/5	4
2	20	1	5/5	60
2	5	2	5/5	60
2	5	1	2/5	13, 18, 27, 60, 60
2	2.5	2	2/5	16, 16, 20, 60, 60
2	1	2	0/5	4, 4, 9, 11, 13
2	Veh	2	0/5	4

^*Mice that cleared parasites and remained parasite free through day 60 postinfection were declared cured.

Mice dosed with 64 orally at 50 mg/kg twice per day for 14 days had no difference in weight change compared to vehicle treated mice and no observed clinical toxicity, indicating good tolerability of the compound above fully efficacious doses.

CONCLUDING REMARKS

A high throughput phenotypic screen successfully identified the oxazolopyridines as antiparasitic agents against T. brucei. Medicinal chemistry optimization to improve potency and PK properties led to the discovery of the imidazopyridines as a novel class of potent antiparasitic T. brucei compounds. The substituted 2-phenyl-imidazopyridine 64 is a highly potent antiparasitic agent against T. brucei, and cures parasite-infected mice when given orally when dosed down to 2.5 mg/kg. Thus, 64 represents a significant lead compound for the development of a new drug class to treat HAT. Studies are in progress to further optimize the drug-like properties of 64 and to explore its potential in the treatment of second stage (central nervous system involvement) HAT. Additional toxicology studies are also in progress as part of a pre-clinical drug development program.

EXPERIMENTAL SECTION

The synthesis of all compounds is given in Supporting Information. All target compounds were purified by preparative HPLC and are judged to be >95% pure.

Phenotypic screen for T. brucei growth arrest compounds

To screen compound libraries for antiparasitic activity on T. brucei, we employed the bloodstream form of the well-characterized T. brucei isolate, Lister 427. During the screen, the parasite was grown in 1,536-well plates in 5.5 μL of HMI-9 medium in the presence of library compounds. All wells including negative controls contained a final of 0.4% DMSO. After incubating the plates at 37 °C for 48 h, the parasite density was determined using the CellTiter-Glo reagent (Promega), a firefly luciferase assay system that measures the amount of cellular ATP present in plate wells.

Solubility measurement

A glass tube containing 8 mL of buffer (20 mM ammonium acetate, pH 7.4) was placed in a 37 °C water bath. The test compound was added from a DMSO stock solution to give the desired concentration of compound with 1% DMSO final concentration. The tube was incubated at 37 °C for 5–10 min. Sufficient compound was used to give a visibly cloudy solution. After incubation overnight at 37 °C, the tube was centrifuged at 1,000×g for 20 min at 37 °C. Seven mL of supernatant was transferred to a new tube, and buffer was removed in a Speed-Vac concentrator. One mL of distilled water was added to the residue, and the sample was concentrated to dryness as above. To the residue was added 0.3 mL of water and 0.3 mL of methanol. The amount of compound was determined by HPLC using a C18-reverse-phase column. HPLC of a standard amount of test compound was used to convert the UV peak area to moles of compound. In some cases, solubility was carried out in 0.01 M HCl.

Distribution of compounds between mouse plasma and brain

Mice (in groups of 3) were injected with test compounds (5 mg/kg IP) and sacrificed at the indicated times for collection of plasma and brain. Compound was dissolved in 0.4 mL of dosing solution (7% Tween 80, 3% ethanol, 5% DMSO, 0.9% saline) for IP injections. The brains were weighed and immediately frozen, then later homogenized in acetonitrile using Dounce homogenizer. Prior to animal studies, recovery of test compound was carried out by adding a known amount to a mouse brain in the test extraction solvent and performing the homogenization. Compound recovery was determined by liquid chromatography/tandem mass spectrometry analysis relative to a standard compound amount. Blood was taken from the same mice in heparinized capillary tubes (Fisherbrand) for determination of compound concentration in plasma. The concentration of compound in the brain was obtained by dividing the moles of compound in the brain by the brain volume (obtained from the brain weight assuming 1 g is 1 mL) and correcting for the brain vasculature volume of 2.5% by weight.

In vitro parasite growth arrest assay

Compounds were tested for antiparasitic activity on T. brucei brucei (strain BF427) as described.¹³ Cells were tested in triplicate in the presence of serial dilutions of compound, and growth is quantified with Alamar Blue.¹⁴ Pentamidine isothionate (Aventis, Dagenham, UK) was included in each assay as a positive control (EC₅₀ 1.2 +/− 0.3 nM).

Mammalian cell cytotoxicity assay

Selected compounds were tested for cytotoxicity at University of Washington against CRL-8155 cells (human lymphoblasts) and HepG2 cells (human hepatocellular carcinoma) as previously described (PMID: 22720744). Briefly, cells were exposed to serial dilutions of compounds for 48 hours and cytotoxicity was assayed using the AlamarBlue® (Life Technologies). Each dilution was assayed in quadruplicate with standard error of the mean values averaging < 15%. Concentrations causing 50% growth inhibition (CC₅₀) were calculated by non-linear regression using GraphPad Prism software (San Diego, CA). L6 rat myoblasts were tested by similar methods at the Swiss TPH as previously described (PMID: 12512078).

In vitro liver microsome studies

Liver microsome metabolism studies were carried out as described.^15,16 Compounds were incubated with liver microsomes (mouse or human), and samples were extracted at time points for liquid chromatography/tandem mass spectrometry analysis. Reference drugs dextromethorphan and testosterone were used to standardize each assay.

Inhibition of liver cytochrome P450s

Inhibition of human cytochrome P450 (3YP3A4 isoform) by test compounds were measured with a commercial kit (BD Biosciences) according to the manufacturer’s instructions.

Permeability across monolayers of MDCKII-MDR1 cells

This assay uses Madin-Darby canine kidney cells that were transfected with the human MDR1 (P-gp) gene.^10–12 Permeability across these monolayers were measured in triplicate as described. ¹⁷ The assay was performed with and without the addition of GF-120918, an inhibitor of the MDR1 efflux pump to determine if the compound is a pump substrate. Propranolol was used as a permeable, non-MDR1 substrate control, and Amprenavir as a permeable, MDR1 substrate control.

PK studies in mice

Test compound was administered to mice by oral gavage followed by blood sampling at intervals of 30, 60, 120, 180, 240, and 360 min. Compound was dosed orally at 50 mg/kg in 0.2 mL of dosing solution (7% Tween 80, 3% ethanol, 5% DMSO, 0.9% saline).

Experiments were performed with groups of 3 mice per compound as published.^17,18 Plasma was separated and extracted with acetonitrile for measurements of compound concentrations by liquid chromatography/tandem mass spectrometry.

PK studies in rats

Test compound was administered to Sprague-Dawley jugular canulated rats (Charles River) by either oral gavage or IV injection followed by blood sampling from the jugular vein at 5, 15, 30, 60, 90, 120, 240, 360, 480, 720 and 1440 min. The oral dose was administered to each rat at 20 mg/kg at time = 0 in a 1 mL volume of dosing solution (7% Tween 80, 3% EtOH, 5% DMSO, 09.% saline.) IV injections were administered at 5 mg/kg from time = 0 to 3 minutes in a 1 mL volume of dosing solution, and blood was sampled at the same time points via the jugular vein. Experiments were performed with groups of 2 rats each for the oral and IV dosing. Plasma was separated and extracted with acetonitrile for measurements of compound concentrations by liquid chromatography/tandem mass spectrometry. PK calculation were performed using Phoenix WinNonlin software (Pharsight).

Plasma protein binding

Dialysis membrane sheet (MW cutoff 3.5 kDa, HTDialysis Cat. 1135) was soaked for 1 h in distilled water and then in 20% ethanol for 30 min. The membrane was clamped between two Teflon plates containing a row of opposing wells. Test compound in DMSO was added to 0.12 mL of serum to give a concentration of 9 μM, a small aliquot was taken as a 100% recovery standard, and the solution was placed on one side of the membrane. The well on the other side of the membrane was charged with an equal volume of dialysis buffer (14.2 g/L Na₂HPO₄, 8,77 g/L NaCl, pH 7.4), and the block was placed on an orbital shaker for 18 h at 37 °C. An aliquot was taken from each side of the membrane. One-fourth volume of acetonitrile was added to each aliquot, and the samples were centrifuged to precipitate protein. Test compound in the supernatants was quantified by liquid chromatography/tandem mass spectrometry to determine the concentration on each of the membrane and the total recovery of test compound from the device. A control dialysis was carried out with dialysis buffer on both sides of the membrane to ensure that equilibration of test compound across the membrane was achieved. The fraction of compound bound to protein was calculated as bound/(unbound + bound), and the value was taken only if compound recovery was >50% (recovery was 84% for 64).

Anti-parasite efficacy studies in mice

Studies were carried out using the standard operating procedure used by WHO screening centers.¹⁹ The studies were done in compliance with University of Washington IACUC approved protocol. Female Swiss-Webster mice age 6–8 weeks in groups of 5 were infected with 2 × 10⁴ T. brucei rhodesiense STIB 900 strain on day 0, then administered compound (in 3 dose levels) or vehicle for 5 days. Treatments were administered orally at a dose and schedule anticipated to maintain plasma concentrations well above the EC₅₀ (see Table 5). The first dose was 48 h after parasite injection, and dosing was 12 or 24 h apart. Dosing was carried out by oral gavage as for PK studies (see above) Parasitemia was monitored for 60 days by microscopic analysis of blood collected from tail bleeds. Cures were defined by sustained clearance of microscopic parasitemia through the end of the 60 day observation period. Mice were euthanized when high-levels of parasitemia were evident on peripheral blood slides.

ABREVIATONS

AUC

area under the plasma concentration-time curve

Cl_p

plasma clearance

HAT

human African trypanosomiasis

MDCK

Madin-Darby canine kidney cells

MDCK-MDR1

MDCK cells transfected with human P-glycoprotein

PK

pharmacokinetics