Optimization of purine-nitrile TbcatB inhibitors for use in vivo and evaluation of efficacy in murine models

Article abstract of DOI:10.1016/j.bmc.2010.09.073

There are currently only four clinical drugs available for treating human African trypanosomiasis (HAT), three of which were developed over 60 years ago. Despite years of effort, there has been relatively little progress towards identifying orally availab

Full text of DOI:10.1016/j.bmc.2010.09.073

Bioorganic & Medicinal Chemistry 18 (2010) 8302–8309
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Optimization of purine-nitrile TbcatB inhibitors for use in vivo
and evaluation of efficacy in murine models
Jeremy P. Mallari a,b, Fangyi Zhu b, Andrew Lemoff b, Marcel Kaiser c, Min Lu b, Reto Brun c, R. Kiplin Guy a,b,
⇑
a Graduate Program in Chemistry and Chemical Biology, University of California, San Francisco CA 94143-2280, United States
b Department of Chemical Biology and Therapeutics, St. Jude Children’s Research Hospital, Memphis TN, 38105, United States
c Swiss Tropical and Public Health Institute, Socinstrasse 57, 4002 Basel, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
There are currently only four clinical drugs available for treating human African trypanosomiasis (HAT),
three of which were developed over 60 years ago. Despite years of effort, there has been relatively little
progress towards identifying orally available chemotypes active against the parasite in vivo. Here, we
report the lead optimization of a purine-nitrile scaffold that inhibits the essential TbcatB protease and
its evaluation in murine models. A lead inhibitor that had potent activity against the trypanosomal pro-
tease TbcatB in vitro and cultured parasites ex vivo was optimized by rationally driven medicinal chem-
istry to an inhibitor that is orally available, penetrates the CNS, has a promising pharmacokinetic profile,
and is non-toxic at 200 mg/kg in a repeat dosage study. Efficacy models using oral administration of this
lead inhibitor showed a significantly increased survival time in Trypanosoma brucei brucei infected mice
but little effect on Trypanosoma brucei rhodesiense infected mice.
Received 3 August 2010
Revised 24 September 2010
Accepted 30 September 2010
Available online 8 October 2010
Keywords:
Parasite
TbCatB
Protease inhibitor
Human African trypanosomiasis
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
synthetase; ascofuranone,9,10 which targets ubiquinol oxidase;
benzyl diamino pyrimidines,11 which target dihydrofolate reduc-
The four drugs (melarsoprol, eflornithine, pentamidine, and
suramin) currently approved for treating human African trypano-
somiasis (HAT) have toxic side effects and require intravenous
administration, which is difficult in the rural areas where HAT is
endemic.1 In addition, only melarsoprol is effective against the
CNS stage of both forms of HAT.1,2 These limitations drive the need
to discover and develop new chemotypes to treat sleeping
sickness, especially those with lower toxicity, improved oral
absorption, and CNS penetration. Much of the current effort for
HAT inhibitor development has focused on reoptimization of
established chemical scaffolds, especially the dicationic aryl bis-
amidines. In fact, other than eflornithine the only new small mol-
ecules to move into clinical trials for HAT since the 1970’s are
fexinidazole and the pentamidine prodrug DB289 (pafurami-
dine).3–7 There have been few novel inhibitor scaffolds that show
in vivo efficacy against a murine Trypanosoma brucei infection,
and even fewer scaffolds with established oral efficacy. Some
examples of chemotypes active in animal models include
buthionine sulfoximine,8 which targets gamma glutamyl cysteine
tase; 6-diazo-5-oxo-L-norleucine and a-amino-3-chloro-4,5-dihy-
dro-5-isoxazoleacetic acid,12 which target cytidine triphosphate
synthetase; diazomethane peptide analogues,13,14 which target
multiple cysteine proteases;
sis; and S-adenosyl-
-methionine analogues,16 which target poly-
amine metabolism. Among these leads, only DB289, -carnitine,
D
-carnitine,15 which inhibits glycoly-
L
D
and ascofuranone, have demonstrated efficacy by an oral route.
Only DB289 and ascofuranone are curative, and both require multi-
ple daily treatments to clear an infection.5–7,10 Despite the early
success of these compounds, only DB289 is particularly amenable
to medicinal chemistry and further development. In addition,
DB289 has poor CNS permeability and is therefore not effective
against late stage infections. There is clearly a pressing need for
the identification of novel orally available and CNS permeable
trypanocides.
We have previously reported a purine-nitrile scaffold with
potent activity against cultured trypanosomes.17,18 These com-
pounds were targeted to the trypanosomal protease TbcatB, one
of two papain-like cysteine proteases in the parasite. TbcatB is
essential for parasite growth and facilitates iron acquisition via
lysozomal degradation of host transferrin.19,20 Prior work in our
group optimized potency and selectivity against TbcatB and
demonstrated a strong correlation between TbcatB inhibition and
trypanocidal activity. This series of inhibitors also kills cultured
parasites with a high degree of selectivity relative to multiple
mammalian cell lines.17,18
⇑
Corresponding author. Tel.: +1 901 495 5714; fax: +1 901 495 5715.
E-mail address: kip.guy@stjude.org (R.K. Guy).
Abbreviations: HAT, human African trypanosomiasis; ADME, absorption,
distribution, metabolism, and excretion; NAETL, no adverse event threshold levels;
DMF, dimethylformamide; DMSO, dimethyl sulfoxide; ACN, acetonitrile; PAMPA,
parallel artificial membrane permeability assay; FOB, functional observational
battery; CBC, complete blood count.
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.09.073

J. P. Mallari et al. / Bioorg. Med. Chem. 18 (2010) 8302–8309
8303
Herein, we report the optimization of the absorption, distribu-
tion, metabolism, and excretion (ADME) of this inhibitor series
using both in vitro methods and an in vivo murine model. A total
of six inhibitors were evaluated for acute toxicity in Balb/C mice
and found to have no adverse event threshold levels (NAETL) of
greater than 200 mg/kg. ADME of a test set of three compounds
was examined and these data indicated that oxidation at the N9
side chain was a major issue for this inhibitor series. Modifications
to the N9 side chain were tested for in vivo and in vitro stability,
leading to the development of a best candidate 3f. Inhibitor 3f
was optimized for oral absorption and metabolic stability while
chromatography with a resulting yield of 60%. Purity of all com-
pounds was confirmed by two orthogonal analytical LC/MS proce-
dures, using both C4 and C18 columns.
2.2. In vitro inhibitor testing
Compounds were selected for testing in a mouse model of tox-
icity based on protease inhibition profile, trypanocidal potency,
cellular therapeutic index, and stability profile, as previously de-
scribed.17,18 Prior to testing in mice, each inhibitor was evaluated
using an in vitro ADME assay panel to identify potential in vivo lia-
bilities. Aqueous solubility and passive permeability (PAMPA) were
determined using a physiologically relevant pH range to predict
oral absorption and cell membrane permeability. Compound sta-
bility was tested in DMSO, PBS, and mouse plasma to assess non-
metabolic degradation. In addition, stability in liver microsomes,
hemolytic activity, and cytotoxicity was determined according to
previously reported methods to predict expected hepatic CL and
potential toxicity.18,21–25
maintaining potent trypanocidal activity (EC50 = 0.53
oral doses of 3f at 1000 mg/kg are non-toxic. Concentrations of
exceeding 10 M could be achieved in both the plasma and brain
lM). Single
l
tissue, and these levels persisted for over 24 h following a single
200 mg/kg oral dose of the lead inhibitor. This compound was also
characterized in a repeat dosage study, and no negative symptoms
were observed after daily treatments of 200 mg/kg for 10 consecu-
tive days. Finally, 3f was evaluated for efficacy in Trypanosoma
brucei brucei infected mice and, although the treatment was not
curative, post-infection survival time was significantly increased.
2.3. Formulation
However, the compound was not active in vivo against
a
Trypanosoma brucei rhodesiense infection.
Inhibitor formulations were developed using a standardized
high throughput UPLC based assay with either UV or MS detec-
tion.22 Briefly, several combinations of three different organic
co-solvents (ethanol, propylene glycol, and PEG 400) were diluted
in various buffered aqueous solutions (pH 4, 7.4, and 9) and
dispensed into a 96-well plate (Supplementary data). These co-
solvents effectively solubilize hydrophobic compounds and are
compatible with IV, IP, and oral routes of administration. A DMSO
inhibitor stock solution of each compound was then added to each
well to a final concentration of 1% DMSO and 10 mM inhibitor. The
plate was sealed and agitated on an orbital shaker overnight then
centrifuged to pellet undissolved compound. A sample was then
taken from the center depth of each well for UPLC analysis to
determine the concentration of inhibitor in solution.
2. Materials and methods
2.1. Chemistry
Purine-nitriles were synthesized using the general routes previ-
ously described (Scheme 1).17,18 Briefly, 1 was reacted with the
appropriate amine in 2-butanol to install the 6-amino substituent.
The crude reaction mixture was concentrated in vacuo and resus-
pended in DMF with K2CO3. Heating this mixture with the desired
alkyl bromide afforded alkylation at N9. Purification was accom-
plished by silica medium pressure chromatography giving overall
yields for the two reactions ranging from 25% to 60%. Subsequent
reaction with sodium cyanide in DMSO with microwave accelera-
tion afforded nitriles 3a–f, which were purified by preparative
C18 chromatography with yield of 30–70%. Suzuki coupling of 3a
and 2,6-dimethylphenylboronic acid was performed with Na2CO3
and Pd(PPh3)4 in 1,4-dioxane with microwave acceleration to give
inhibitor 4a. The target inhibitor 4a was purified by preparative C18
2.4. Preliminary in vivo ADME characterization
Initial in vivo studies were designed to quickly provide a preli-
minary assessment of acute toxicity, inhibitor metabolites, and
semi-quantitative oral absorption and plasma exposure. Studies
Br
Br
NH
NH
NH
N
N
N
N
N
N
N
i a
ii a
iii
N
N
H
N
Cl
N
N
N
N
Cl
N
HO
HO
N
N
2a
3a
4a
N
H
Cl
1
i b
ii b
Cl
Cl
Cl
Cl
NH
N
NH
N
N
N
N
N
N
N
H
Cl
N
R
2b
3b-f
Scheme 1. Synthesis of purine-nitriles. Reagents and conditions: (i) (a/b) 3-bromobenzylamine or 3,4-dichlorobenzylamine, 2-butanol, 60 °C, 3–12 h; (ii) (a/b) alkyl bromide
or 3-hydroxypropylbromide, K2CO3, DMF, 60 °C, 10–16 h; then NaCN, DMSO, W, 180 °C, 160 W, 5–20 min; (iii) 2,6-dimethylphenylboronic acid, Na2CO3, Pd(PPh3)4,
1,4-dioxane, W, 150 °C, 300 W, 5–40 min.
l
l

8304
J. P. Mallari et al. / Bioorg. Med. Chem. 18 (2010) 8302–8309
were performed using 12 mice per compound, which were treated
in a four-stage dose escalation study. One advantage of the exper-
imental design is that acute toxicity data are collected simulta-
neously with preliminary pharmacokinetic data, using the same
animals.
and treated on days 3, 4, 5, and 6 with 3f (50, 100, or 200 mg/kg)
or vehicle only. Four mice were used for each dosage group, and sur-
vival time was determined as described above (Fig. 2).
3. Experimental
The maximum tolerated dosage (MTD) for oral administration
of each inhibitor was determined by testing increasing dosages
starting with vehicle only, followed by 50, 100, and 200 mg/kg. Fol-
lowing treatment, mice were observed using a modified Irwin
Functional Observational Battery (FOB) for 15 min for acute ad-
verse symptoms before continuing the dose escalation.26 All mice
were monitored for 48 h after completion of the dose escalation.
For each dosage group, a single retro orbital bleed of approximately
0.1 ml was collected for UPLC–MS analysis from one mouse at 1, 2,
and 4 h after drugging. At 48 h animals were sacrificed, a terminal
blood draw was collected by cardiac puncture, and a gross nec-
ropsy was performed. The blood sample was used for complete
blood count (CBC) and serum chemistry analysis, as well as to pro-
vide a final PK time point.
3.1. 6-((3,4-Dichlorobenzyl)amino)-9-(4,4,4-trifluorobutyl)-9H-
purine-2-carbonitrile (3f)
3-(Trifluoromethyl)propyl bromide was purchased from Matrix
Scientific. All other chemicals were purchased from Sigma–Aldrich.
All reaction solvents were anhydrous. LC/MS data were collected
on a Waters system using XTerra C18 columns (MeOH/H2O, 1%
HCOOH). Synthesis was performed as previously described.17,18
Briefly, 2,6 dichloropurine 1 (0.25 g, 1.3 mmol) was suspended in
of 2-butanol (3.9 ml, 0.34 M) and treated with the appropriate
benzylamine by drop-wise addition (3.9 mmol, 3.0 equiv). The
reactions were stirred for 3–12 h at 60 °C with monitoring by thin
layer chromatography (silica gel 60). The crude product mixture
was concentrated under vacuum and resuspended in anhydrous
DMF (5 ml, 0.26 M). K2CO3 (solid, 0.54 g, 3.9 mmol, 3.0 equiv)
was added, followed by drop-wise addition of the desired alkyl
bromide (5.2 mmol, 4.0 equiv). The reaction was stirred 12–18 h
at 60 °C and monitored by LC/MS. The crude reaction was concen-
trated under vacuum and purified by flash chromatography (30:1
DCM/MeOH v/v) using 40 M columns (Biotage) on a Biotage SP1
flash chromatography instrument to give a waxy solid (60% yield).
The purified intermediate (2-chloro-N-(3,4-dichlorobenzyl)-9-
(4,4,4-trifluorobutyl)-9H-purin-6-amine) 2-chloro-N-(3,4-dichlo-
robenzyl)-9-(4,4,4-trifluorobutyl)-9H-purin-6-amine (0.5 mmol)
was dissolved in DMSO (5.0 ml, 0.1 M) in a 5 ml microwave reac-
tion vial (Biotage). Sodium cyanide (170 mg, 3.5 mmol, 7.0 equiv)
was added. The vial was sealed and heated to 180 °C for increments
2.5. Efficacy studies
In vivo efficacy against T.b. brucei was modeled using an estab-
lished system.13,27 On day 0, 16 female Balb/C mice, aged
10–12 weeks, were infected withT.b. brucei (strain 221) by intraperi-
toneal injection of approximately 800 parasites. Infected mice were
randomly selected for treatment with either 3f (200 mg/kg) or vehi-
cle on days 0–4. Blood was collected by saphenous vein puncture
every third day, starting on day 4, to measure parasitemia by hemo-
cytometer. Animals displaying symptoms indicative of unaccept-
ably high parasitemia were euthanized. In vivo efficacy against T.b.
rhodesiense (strain STIB900) was modeled in female NMRI mice.
Mice were infected by ip injection of 2 Â 104 parasites on day 0
Figure 1. In vivo inhibitor metabolism.

J. P. Mallari et al. / Bioorg. Med. Chem. 18 (2010) 8302–8309
8305
reaction plate was incubated at 37 °C and shaken at a speed of
60 rpm. 0.5 h, 1 h, 2 h time points were taken. At each time point,
450 ll pre-cooled acetonitrilewas added to three rows inthereaction
plate, and the liquid was then transferred to the quenched plate. The
quenched plate was then centrifuged (model 5810R, Eppendorf,
Westbury, NY) at 4000 rpm for 20 min. 200 ll supernatantwas trans-
ferred to a 96-well plate and analyzed by UPLC–MS (Waters Inc., Mil-
ford, MA). The compounds and internal standard were detected by
SIR. The log peak area ratio (compound peak area/internal standard
peak area) was plotted against time and the slope was determined
*
to calculate the elimination rate constant [k = (À2.303) slope]. The
half life (h) was calculated as t1/2 = 0.693/k.
Figure 2. Brain and plasma concentration of inhibitor 3f dosed at 200 mg/kg.
3.7. Determination of hemolytic potential
Inhibitors were tested as previously reported.23 Briefly, 1 ml of
whole human blood was centrifuged at 1000g for 5 min. The super-
natant was decanted and the pellet was washed three times with
of 10 min in a Biotage Initiator microwave until the reaction was
complete as shown by LC/MS, a total of 30–90 min. The crude prod-
uct mixture was decanted from insoluble NaCN and purified by re-
verse-phase HPLC (Waters XTerra preparative C18 column, ACN/
H2O, 1% HCOOH, linear gradient from 40–70% ACN) to afford nitrile
3f with a yield of 70%.
1 ml PBS, pH 7.4. 400
in 9.6 ml PBS. In a 96-well plate, 2
to 98 l PBS and mixed. In the first 2 rows of the 96-well plate,
100 l PBS was added as a negative control. 100 l of resuspended
human erythrocytes were added to each well. Final assay hemato-
crit was 2% and analyte concentration was 10 M. 0.4% saponin
ll human erythrocytes were resuspended
l
l of 1 mM analyte was added
l
l
l
3.2. 1H NMR peak listing for inhibitor 3f
l
was used as a positive control for full hemolysis and 1%, 10% and
20% ethanol utilized as partial hemolysis controls. The plate was
incubated at 37 °C with shaking at 60 rpm for 5 h and then centri-
1H NMR ((CD3)2SO) (400 MHz): d 9.02 (1H, m), 8.46 (1H, s), 7.60
(2H, m), 7.34 (1H, d, J = 8.0 Hz), 4.67 (2H, d, J = 6.0 Hz), 4.27 (2H, t,
J = 6.8 Hz), 2.34 (2H, m), 2.06 (2H, quintet, J = 7.6 Hz) LC/MS: M+H+:
429.15.
fuged at 1000g for 5 min. 75
ll supernatant from each well was
transferred to a 96-well UV plate (Corning). 75
ll supernatant from
saponin well was added to wells A1 and B1, and serial diluted
along the row to wells A11 and B11. Absorbance at 414 nm was
determined for each well, and percent hemolysis was determined
using a standard calibration curve. All data was collected in
triplicate.
3.3. Cloning and expression of TbcatB
TbcatB was cloned, expressed, and purified as previously
described.18
3.4. Enzyme inhibition assays
3.8. Metabolite identification
Inhibitors were screened against TbcatB, rhodesain, human
cathepsin B, and human cathepsin L as previously described.18,20
All metabolite identification was performed using a Waters
Acquity UPLC system (Waters, Milford, MA) coupled to either a
Waters Acquity single quadrupole mass spectrometer or a Waters
Acquity triple quadrupole mass spectrometer, both fitted with an
electrospray ionization (ESI) source. Chromatographic separations
3.5. Cell proliferation assays
Inhibitors were screened against T.b. brucei, HEK 293, Hep G2,
were achieved on a 21 Â 50 mm, 1.7
lm Waters Acquity BEH col-
BJ, and Raji as previously described.18,20
umn. The column eluent was split after the column, with half going
to the mass spectrometer and half going to a PDA detector. All
analyses were performed in positive ion mode. Instrument control,
data acquisition, and data processing were performed using
Masslynx v. 4.1.
3.6. Liver microsomal stability
Microsomal stability was determined as previously described21,23
with some modification. 0.63 ml of liver microsome protein (20 mg/
ml, female CD-1 mice, BD Biosciences) was mixed with 0.05 ml of
0.5 M EDTA solution and 19.32 ml potassium phosphate buffer
(0.1 M, pH 7.4, 37 °C) to make 20.00 ml of liver microsome solution.
One part of 10 mM DMSO compound stock was mixed with 4 part
of acetonitrile to make 2 mM diluted compound stock in DMSO and
Mobile phase for the single quadrupole MS was 0.1% formic acid
in water and methanol. The proportion of methanol was increased
from 40% to 100% over 4 min. Mass spectra were acquired over a
range of m/z = 110–600. The triple quadrupole MS mobile phase
consisted of 0.1% formic acid in water and 0.1% formic acid in
acetonitrile. The proportion of acidified acetonitrile was increased
from 10% to 90% over 1.8 min, and brought back to 10% over
acetonitrile. 30.4
microsomal solution and vortexed to make microsomal solution with
compound. 180 l of the microsomal solutions with different com-
pounds were dispensed into respective rows of a 96-well storage
plate (pION Inc., MA). For 0 h time point, 450 l pre-cooled acetoni-
M caffeine) was added
lldilutedcompoundstockwasaddedto2.4 mlliver
0.2 min. MS/MS spectra were acquired over
a
range of
l
m/z = 50–450. Collision energies were varied from 15–30 eV to
achieve varying amounts of fragmentation to aid in structural
elucidation.
l
trile (À20 °C) with internal standard (100
l
to the first three columns before the reaction starts. 1.25 ml of micro-
someassay solution A (BD Biosciences) was combined with 0.25 mlof
solution B (BD Biosciences) in 3.5 ml of potassium phosphate buffer
3.9. Mouse MTD and chronic toxicity studies
Female balb/c mice aged 10–12 weeks were purchased from
Charles River Laboratories and housed individually. Food and
water were provided ad libitum. Inhibitors 3b and 3d were
suspended in 5% ethanol, 50% propylene glycol, and 45% PBS (pH
(0.1 M, pH 7.4). 45 ll of this solution was added to each well of the
96-well storage plate (reaction plate). Liquid in the first three
columns was moved to another storage plate (quenched plate). The

8306
J. P. Mallari et al. / Bioorg. Med. Chem. 18 (2010) 8302–8309
7.4). Compounds 3c and 3e were suspended in 10% ethanol, 10%
propylene glycol, 40% PEG 400, and 40% PBS (pH 7.4). Compound
4a was suspended in 10% ethanol, 10% propylene glycol, 40% PEG
400, and 40% sodium citrate (1 mM, pH 6.0). Compound 3f was
suspended in 20% ethanol, 40% propylene glycol, 10% PEG 400,
and 30% PBS (pH 7.4). Oral dosing at 50, 100, 200, 400, and
1000 mg/kg was carried out with inhibitor suspensions of 10, 20,
40, 80, and 200 mg/ml, respectively. For each mouse, 0.1 ml of
inhibitor suspension was given by oral gavage for every 20 g of ani-
mal weight. To determine the IV MTD for inhibitor 3f, solubilized
inhibitor was given by lateral tail vein injection (0.1 ml solution
per 20 g animal). Dosing of 3f at 2, 4 and 8 mg/kg was carried
out with inhibitor solutions of 0.4, 0.8, and 1.6 mg/ml, respectively.
0.1 ml blood was collected retro-orbitally from a different
mouse within each treatment group at 1, 2, and 4 h after treat-
ment. Only one collection was made per mouse during the entire
study period. In some cases the sample size was expanded to six
animals and additional samples were collected at 8, 12, and 24 h.
All samples were treated with EDTA. Mice were observed at 1, 2,
4, 8, 24, 32, and 48 h according to a modified FOB. Gait, alertness,
respiration, pilorection, vasodilation, and vocalization were moni-
tored and recorded at each time point. At 48 h the animals were
sacrificed and a terminal cardiac bleed was collected from each
animal. Plasma samples were prepared as previously described.25
Briefly, each blood sample was centrifuged for 2 min at
10,000 rpm in a desktop centrifuge, and the plasma was collected.
Plasma was diluted in three parts of ice cold acetonitrile, vortexed
vigorously, and centrifuged at 13,000 rpm for 10 min in a desktop
centrifuge. The supernatant was collected and inhibitor concentra-
tion was determined in a UV/HPLC based assay.
Compounds were tested for hemolytic activity as well as stability
in DMSO, PBS, mouse plasma, and liver microsomes. Solubility
and PAMPA permeability were determined, and dosing formula-
tions were developed for in vivo studies. In vivo studies assessed
each inhibitor’s metabolic stability, oral availability, plasma expo-
sure, and MTD. Subsequent optimization was driven by these
in vivo parameters. A total of six inhibitors (3b–3f, and 4a) were
tested in this study.
4.1. In vivo ADME optimization
ADME testing began with inhibitors 3b–c and 4a. Each of these
compounds was non-toxic in cell proliferation and hemolysis
assays, and stable in DMSO and PBS (pH 3, 5, and 7.4) over the
course of the study (14 days). Stability of 3b, 3c, and 4a in both
microsomes and mouse plasma was modest (microsome t1/2
=
0.53–1.08 h, plasma t1/2 = 5.1–38.3 h) (Table 1). PAMPA flux values
for 3b and 3c were reasonable, though it was not possible to deter-
mine a value for 4a due to its low aqueous solubility (Table 2).
In vivo, the initial inhibitor set displayed no detectable plasma
exposure due to either poor oral availability or poor metabolic sta-
bility. Blood samples of mice treated with 4a did not show detect-
able amounts of either the inhibitor or potential inhibitor
metabolites, suggesting little or no oral absorption. Inhibitors 3b
and 3c were also not detected in the blood, although high levels
(>50 lM) of a single, predominant, oxidatively modified metabolite
(5a and 5b, respectively) was observed for each compound (Sup-
plementary data). MS/MS analysis indicated these metabolites ar-
ose from oxidation of the inhibitor’s terminal hydroxyl groups on
the N9 side chain to carboxylic acids 5a and 5b (Fig. 1). After oral
dosing of 3c, the parent inhibitor was completely converted to
5b by the 1-h time point and no other metabolites were detectable.
Following the oral dosing of 3b, 5a concentration peaked at 1 h, fol-
lowed by a drop in 5a levels and a corresponding increase in 5b
concentration at 2 and 4 h. 3b also gives rise to a minor metabolite
(m/z = 435), which appears to be the result of the hydroxylation of
the N9 sidechain of 5a. Low levels of an unidentified metabolite (m/
z = 386) were also detected for 3c. These data suggest that 3b is
first oxidized to 5a then subsequently converted to 5b, presumably
by b-oxidation. The 5b metabolite does not appear to give rise to
any other species. These metabolic modifications are undesirable,
as previous work has shown that such carboxylic acid derivatives
The chronic toxicity of 3f was determined in a group of five
mice treated by oral gavage with 200 mg/kg daily for 10 days. Sur-
viving mice were monitored out until 30 days. One mouse from the
cohort was euthanized and submitted for blood analysis and nec-
ropsy on day 10 and day 30, respectively.
3.10. Efficacy studies
Parasites were grown to a density of 107 parasites/ml in HMI-9
media supplemented with 10% fetal bovine serum (Omega), 10%
Serum Plus (JRH Biosciences), and penicillin/streptomycin. Para-
sites were pelleted by centrifugation then resuspended and serially
diluted in unsupplemented HMI-9 media to a density of 8 Â 103
parasites/ml. Mice were inoculated by IP injection of 0.1 ml of this
culture at 0 h on day 0. Mice weights were recorded and each ani-
mal was randomly assigned to either a drug treatment or vehicle
only control group Mice were drugged by oral gavage with a sus-
pension of inhibitor 3f (40 mg/kg, final dose of 200 mg/kg) or with
vehicle only (20% ethanol, 40% propylene glycol, 10% PEG 400, and
30% PBS (pH 7.4)). Gavage volume was 0.1 ml per 20 g animal
weight. Oral dosing was continued once a day at 24, 48, 72, and
96 h post-infection. Blood was collected every third day starting
are active against TbcatB (5b IC50 = 2.6 0.6
lM) but inactive
against the parasite (5b EC50 >25
lM), presumably due to poor cell
permeability.
Taken together, these results indicate that those compounds
that provided the best balance of in vitro characteristics have sig-
nificant ADME liabilities that hinder their further development.
The large and hydrophobic 6-amino biphenyl moiety of 4a, the
most TbcatB selective inhibitor in the series, is apparently respon-
sible for the inhibitor’s poor oral absorption. Likewise, the N9
hydroxyalkyl sidechains of inhibitors 3b–c, which provide reason-
able solubility and potency, are rapidly oxidized. Despite the insta-
bility of the 3b–c, these compounds appear to have good oral
absorption. Modification of the side chain alcohol was considered
a more tractable strategy for mitigating these liabilities than
attempting to address the poor oral absorption of the more hydro-
phobic subseries. Further efforts focused on optimizing the stabil-
ity of the N9 side chain while leaving fixed the 3,4-dichlorobenzyl
group at the 6-amino position.
on day 4 to monitor parasitemia. Approximately 10 ll of blood
was collected by saphenous vein puncture in a heparinized collec-
tion tube. Blood samples were diluted 100 fold in PBS and parasite
density was counted by hemocytometer. Animals displaying symp-
toms indicative of heavy parasitemia (extensive pilorection, lack of
responsiveness) were euthanized by CO2 asphyxiation followed
cervical dislocation.
Preventing side chain oxidation by replacing the hydroxyl of 3c
with the corresponding methyl ether 3d was initially assessed. In
vitro, this modification did not significantly improve microsomal
stability (t1/2 = 0.53 h) or PAMPA permeability, but did significantly
improve plasma stability (t1/2 = 40.9 h) relative to 3c (Tables 1
and 2). In vivo, the parent compound 3d was not detected after oral
4. Results
An initial set of three inhibitors (3b, 3c, and 4a) was chosen for
ADME characterization based on previously determined trypanoci-
dal activity, in vitro therapeutic index, and TbcatB selectivity.17,18

J. P. Mallari et al. / Bioorg. Med. Chem. 18 (2010) 8302–8309
8307
Table 1
Inhibitor activity and stability in vitro
Inhibitor
R
EC50 versus T. brucei (
lM)
Therapeutic index (fold)
% Hemolysis
Plasma t1/2 (h)
Microsome t1/2 (h)
3b
3c
3d
3e
3f
5-Hydroxypentyl
3-Hydroxypropyl
3-Methoxypropyl
n-Butyl
0.38 0.04
1.4 0.1
2.4 0.4
0.5 0.3
0.53 0.06
0.27 0.3
>20
>20
>10
>20
>20
>20
0.70 0.03
0.70 0.05
0.70 0.08
0.90 0.04
0.4 0.1
38.3
5.1
40.9
246.6
177.0
25.7
1.08 0.09
0.60 0.02
0.53 0.04
0.74 0.07
Stable
3-(Trifluoromethyl)propyl
4a
1.1 0.1
0.43 0.04
Therapeutic index is defined as the inhibitor EC50 value against the most sensitive mammalian cell line divided by EC50 versus T.b. brucei. Hemolytic potential was determined
at inhibitor concentrations of 10 lM. Plasma and liver microsome assays were carried out at 37 °C.
Table 2
Inhibitor solubility and permeability
Inhibitor
pH 3
PAMPA flux
pH 5
Aqueous soln (
l
M)
Formulation soln (
l
M)
Formulation
pH 7.4
3b
3c
3d
3e
3f
960 60
1500 100
1800 800
1512
960 60
1900 300
1300 300
1770 60
16.1 0.6
10.4 0.3
4.2 0.6
0.7 0.1
0.3 0.1
1.0 0.2
10,340 80
5/50/0/45, pH 4.0
10/10/40/45, pH 7.4
5/50/0/45, pH 7.4
10/10/40/45, pH 7.4
20/40/10/30, pH 7.4
10/10/40/40, pH 6.0
4060 30
890 10
4a
Aqueous solubility for each compound was determined in deionized water. Formulations are shown as the ratio of ethanol/propylene glycol/PEG 400.
dosing, although micromolar levels of the O-demethylation product
3c were transiently observed for 1–2 h following drug treatment.
This is significant in that metabolite 3c reaches trypanocidal concen-
methyl moiety would reduce p450-mediated oxidation at the O-1
position. This strategy was highly successful in vitro, and 3f
displayed excellent stability in both plasma (t1/2 = 177.0 h) and
microsomes (no observable transformation). In vivo performance
was similarly improved, and 3f achieved a peak plasma concentra-
tion of approximately 10 lM with no detectable metabolites at any
collection time. At oral dosages tested (50, 100, 200, 400, and
1000 mg/kg), plasma concentrations of the inhibitor peaked at
trations (ꢀ2
lM) in vivo. However, 3c was completely oxidized to 5b
within 4 h after dosing, thus indicating that further optimization
was required. The identities of metabolites 5a and 5b were deter-
mined by comparison to the MS/MS spectrum of a chemically syn-
thesized sample of 5b (Supplementary data).
It was expected that replacement of the N9 moiety with a sat-
urated alkyl chain might improve oxidative stability, and 3e was
tested to explore this hypothesis. In vitro, this approach resulted
in a significant improvement in plasma stability (t1/2 = 246.6 h),
although microsomal stability was comparable to previous com-
pounds (t1/2 = 0.74 h). It was not possible to determine PAMPA per-
meability for this compound do to its low solubility (Table 1 and 2).
Stability was significantly improved in vivo, and trypanocidal con-
between 6 and 10 lM within 1–4 h of dosing (data not shown).
Although peak concentration was similar for all treatment groups,
time required to clear the drug increased with dosage. Following
treatment with 50 or 100 mg/kg, the inhibitor was cleared within
24 h, while at dosages of 400 mg/kg or higher, inhibitor remained
at micromolar concentrations for 48 h or longer. This profile sug-
gests that the inhibitor is orally available at all tested dosages, that
the peak plasma concentration of the drug is limited by its solubil-
ity, and that a high volume of distribution is likely providing a res-
ervoir for retention of the drug in vivo. This is further supported by
the peak plasma concentrations measured following IV administra-
tion of 3f (Supplementary data) as well as the inhibitor concentra-
tions measured in brain tissue (see below). These data suggest that
the hydrophobic inhibitor partitions into other tissues after enter-
ing the circulation, and equilibrates back into the bloodstream as
the drug is cleared.
centrations (1–2 lM) of the parent compound were observed in
the plasma for 1–2 h after dosing. However, by 4 h post dosing
the parent inhibitor was undetectable. Three unidentified metabo-
lites (5c–e) with m/z values of 511.1, 335.0, and 391.1 were ob-
served for 3e, and obvious oxidation products of the O and O-1
positions were not observed. These metabolites appear to be signif-
icantly less hydrophobic, as indicated by their shortened retention
times by reverse-phase HPLC.
Although the site of transformation of 3e remained unclear, oxi-
dation and subsequent modification of either the O or O-1 carbons
were thought to be the most likely scenarios. To block possible oxi-
dation at these positions, the terminal methyl group of the 3e n-
butyl substituent was replaced with a trifluoromethyl group (com-
pound 3f). We hypothesized that eliminating available hydrogen
atoms at the terminal methyl group would prevent oxidation of
the O position, while introduction of the electronegative trifluoro-
To further assess the potential of 3f as a lead compound, long-
term solid state stability was tested at simulated tropical condi-
tions (37 °C, 100% humidity) and was found to be completely stable
for at least 30 days. This stability profile is promising and espe-
cially important in HAT endemic areas that often lack electricity re-
quired for climate controlled drug storage. The stability, oral
availability, and plasma exposure of 3f indicates it is a promising
lead compound for HAT.

8308
J. P. Mallari et al. / Bioorg. Med. Chem. 18 (2010) 8302–8309
4.2. Efficacy studies
All animals with measurable parasitemia at day 4 died by day 6,
while the mice without detectable parasitemia lived for 7–19 days.
In the T.b. rhodesiense infection model mice were treated with
50, 100, or 200 mg/kg of 3f daily on days 3–6 pi. There was no sig-
nificant difference between the survival time of mice treated with
vehicle only (11.5 0.6 days) and those treated with 3f (data not
shown). As this result was at odds with the T.b. brucei findings, sen-
sitivity of the T.b. rhodesiense strain used in this experiment was
tested and it was found that the EC50 for this strain was roughly
10-fold weaker than seen with T.b. brucei. Thus, the pharmacody-
namic result makes sense in the context of the in vitro and
in vivo data.
Due to the encephalopathic nature of HAT, a major concern for
new drug development is CNS availability. Of the four current
clinical drugs, only melarsoprol is effective against CNS stage T.b.
rhodesiense and T.b. gambiense infections, and the paucity of CNS
permeable drugs drives the need for drugs effective against late
stage HAT. To begin assessing the potential of 3f against parasites
in the CNS, 6 mice were treated at 200 mg/kg, and plasma and
brain tissue were collected to determine plasma and CNS inhibitor
concentrations. Brain concentrations of 3f paralleled those in the
plasma, with concentrations peaking within 4 h, and remaining
relatively constant for at least 24 h before declining by 48 h. Peak
concentrations in the brain (ꢀ20
lM) were approximately 2 fold
5. Discussion
higher than those in the plasma over the first 24 h. These data indi-
cate that the inhibitor freely crosses the BBB, and is therefore a
promising lead for developing drugs against late stage HAT.
A single oral dose of 200 mg/kg established a steady peak plas-
ma concentration for approximately 24 h, and this dosage was
determined to be most suitable for a daily treatment regimen.
Chronic toxicity of the inhibitor was assessed in a repeat dose
study (200 mg/kg daily for 10 days) to determine an appropriate
course of treatment. No toxicity was observed at any point by
any of the measures used (CBC analysis, serum chemistry, func-
tional observation battery). The toxicity of 3f was also tested at
IV dosages of 2, 4, and 8 mg/kg. Some negative symptoms were ob-
served immediately after dosing (decreased movement and tran-
sient darkening of the eyes), but all mice recovered within 30 s
and no subsequent adverse effects were observed. Following oral
dosing, plasma concentrations of 3f were approximately 20-fold
5.1. ADME studies
We observed a good correlation between inhibitor stability in
mouse liver microsome preparations and actual CL in mice. The
poor in vivo stability of inhibitors 3b–e was predicted by their
short half lives in this assay (between 0.4 and 1.1 h). Additionally,
the microsome studies successfully predicted the 5a metabolite for
compound 3b, but not the downstream metabolite 5b. This dis-
crepancy may be explained by the lack of a functional b-oxidation
pathway in microsomal fractions, which is a likely route for the
conversion of 5a to 5b. The two metabolites observed in vivo for
3c (5b and the unidentified metabolite with m/z = 386) were iden-
tical to those observed in the microsomal assay. O-Demethylation
was observed for 3d both in vivo and in vitro, and the three metab-
olites of 3e observed in this assay were identical to the three ob-
served in mice. Only 3f was completely stable, consistent with
in vivo data. Although mouse metabolism data was not obtained
for 4a due to poor oral absorption, in liver microsomes the N9 side
chain of this compound was primarily converted to the corre-
sponding carboxylic acid (m/z = 429).
higher than its EC50 (0.53 0.06
lM) and 10-fold higher than its
EC90 (0.8 M) against cultured parasites. No toxicity was observed
l
in the oral MTD study at up to 1000 mg/kg, and 3f was determined
to be a highly promising candidate for efficacy testing.
4.3. Efficacy results
Due to low aqueous solubility, PAMPA data was only obtained
for inhibitors 3b–d (Table 2). Although the rapidly metabolized
parent species were not observed in vivo, the metabolites of these
compounds were detected at relatively high concentrations
Compound 3f was evaluated for efficacy in standard murine
models for acute T.b. brucei and T.b. rhodesiense infection.13,27 In
the T.b. brucei infection model, all mice from the control group
showed significant disease progression by day 6 pi and were
euthanized according to the protocol. In mice treated with 3f
(200 mg/kg) for a 5-day period beginning on day 0, 4/8 animals
showed detectable parasitemia on day 4, and by day 6 only half
of the mice from this group required euthanasia (Fig. 3). Two of
the 3f treated mice lived longer than 14 days, and there was a sig-
nificant increase in average survival time (p = 0.068) for inhibitor
treated animals (9.6 days) relative to the vehicle only controls
(5.6 days). Parasitemia at day 4 was predictive of survival time.
(>50 lM), suggesting good oral absorption of these inhibitors.
These results are consistent with the PAMPA assay data. Therefore,
for this compound series, it is fruitful to utilize the combination of
in vitro microsome assays together with solubility and passive per-
meability models to predict the bioavailability of individual
compounds.
5.2. Efficacy studies
As described above, a pharmacokinetics driven approach was
used to select a candidate TbcatB inhibitor that best matched the
product profile of oral availability, metabolic stability, potency,
and CNS accumulation. The best candidate 3f was then examined
in murine disease models and showed activity against the cattle
disease but not the human disease.
Inhibitor 3f is somewhat less effective at delaying disease pro-
gression in a murine T.b. brucei infection compared to previously
reported irreversible diazomethyl ketone inhibitors.13 However, it
should be noted that diazomethyl ketones require repeated IP
administration at higher dosages (250 mg/kg), and that acute tox-
icity was a limiting factor at these dosages. The reversible purine-
nitrile scaffold displays more desirable inhibition kinetics, and 3f
represents a significant improvement in terms of oral absorption
and toxicity compared to the peptidyl diazomethyl ketones.
Although 3f treatment was not curative, the clear efficacy against
T.b. brucei in delaying disease progression coupled with the
Figure 3. Mouse survival time post-infection.

J. P. Mallari et al. / Bioorg. Med. Chem. 18 (2010) 8302–8309
8309
pharmacokinetic behavior of this inhibitor demonstrates that this is
a promising lead. The compound is quite stable in vivo with no
anti-trypanocidal chemotype that is CNS permeable, orally avail-
able, and efficacious in vivo against T.b. brucei.
detectable metabolites. Peak plasma concentration (6.5–9.7 lM)
was reached within 1–4 h after oral administration for all tested dos-
ages, and peak plasma concentration was maintained for over 24 h
following a single oral dose of 200 mg/kg or more.
Acknowledgments
This work was supported by the American Lebanese Syrian
Associated Charities (ALSAC) and St. Jude Children’s Research
Hospital (SJCRH). We acknowledge the contributions of the High
Throughput Analytical Chemistry Core at SJCRH. We would like
to thank Zachary Mackey and Jim McKerrow for their advice during
the efficacy experiments. We would also like to thank Jill Riggs for
all of her help and technical assistance.
Infections were not cleared by treatment with 3f in the T.b. brucei
model, though plasma levels of the drug were maintained for nearly
one week at concentrations 10-fold higher than the inhibitor’s
in vitro EC90. It is possible that this discrepancy is due to nutrients
or growth factors only present in vivo that change parasite sensitiv-
ity to inhibitor. Or, it may be the case that there is an in vivo selection
for drug resistant parasites. Alternatively, it may be that the avail-
able concentration of drug in vivo is lower than expected, possibly
due to differences in non-specific protein/inhibitor binding. Given
the established efficacy of this inhibitor, it is likely that improving
peak plasma concentration will significantly improve in vivo activ-
ity. Based on solubility and preliminary pharmacokinetic data, it is
likely that the peak plasma concentration of 3f is limited by aqueous
solubility and not by in vivo stability or oral absorption. It is notable
that the structurally similar carboxylic acid metabolites 5a and 5b,
derived from inhibitors 3b–c, achieved plasma concentrations of
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.09.073.
References and notes
1. WHO Human African Trypanosomiasis Drugs. http://www.who.int/trypano
somiasis_african/drugs/en/.
2. Kennedy, P. G. J. Clin. Invest. 2004, 113, 496.
50–100 lM. This suggests that it should be possible to improve 3f
plasma exposure by optimizing solubility.
3. Janssens, P. G.; De Muynck, A. Ann. Soc. Belg. Med. Trop. 1977, 57, 475.
4. Ogada, T.; Fink, E.; Mbwabi, D. Trans. R. Soc. Trop. Med. Hyg. 1973, 67, 280.
5. Mdachi, R. E.; Thuita, J. K.; Kagira, J. M.; Ngotho, J. M.; Murilla, G. A.;
Ndung’u, J. M.; Tidwell, R. R.; Hall, J. E.; Brun, R. Antimicrob. Agents Chemother.
2009, 53, 953.
6. Thuita, J. K.; Karanja, S. M.; Wenzler, T.; Mdachi, R. E.; Ngotho, J. M.;
Kagira, J. M.; Tidwell, R.; Brun, R. Acta Trop. 2008, 108, 6.
7. Jennings, F. W.; Urquhart, G. M. Z. Parasitenkd 1983, 69, 577.
8. Arrick, B. A.; Griffith, O. W.; Cerami, A. J. Exp. Med. 1981, 153, 720.
9. Minagawa, N.; Yabu, Y.; Kita, K.; Nagai, K.; Ohta, N.; Meguro, K.; Sakajo, S.;
Yoshimoto, A. Mol. Biochem. Parasitol. 1997, 84, 271.
10. Yabu, Y.; Yoshida, A.; Suzuki, T.; Nihei, C.; Kawai, K.; Minagawa, N.; Hosokawa,
T.; Nagai, K.; Kita, K.; Ohta, N. Parasitol. Int. 2003, 52, 155.
11. Chowdhury, S. F.; Villamor, V. B.; Guerrero, R. H.; Leal, I.; Brun, R.; Croft, S. L.;
Goodman, J. M.; Maes, L.; Ruiz-Perez, L. M.; Pacanowska, D. G.; Gilbert, I. H.
J. Med. Chem. 1999, 42, 4300.
3f is inactive in vivo against T.b. rhodesiense, presumably due to
decreased potency against T.b. rhodesiense (EC50 = 4.9
lM) relative
to T.b. brucei (EC50 = 0.53 M). It is also possible that pharmacoki-
l
netic differences between the two mouse models used may have
also contributed to differences in in vivo efficacy. The lack of effi-
cacy against T.b. rhodesiense does not necessarily diminish the po-
tential of the purine-nitrile scaffold against HAT, but underscores
the importance of testing inhibitor potency against the human
pathogen early in a development program in addition to the more
commonly utilized T.b. brucei model.
6. Conclusion
12. Hofer, A.; Steverding, D.; Chabes, A.; Brun, R.; Thelander, L. Proc. Natl. Acad. Sci.
U.S.A. 2001, 98, 6412.
13. Scory, S.; Caffrey, C. R.; Stierhof, Y. D.; Ruppel, A.; Steverding, D. Exp. Parasitol.
1999, 91, 327.
We describe herein the optimization of a series of purine-nitrile
inhibitors in order to enhance in vivo ADME parameters and to
study the efficacy of these compounds in two murine models of
African trypanosomiasis. Several inhibitors were tested for acute
toxicity in mice, and all were determined to be non-toxic at doses
up to 200 mg/kg. 3f was found to be non-toxic at oral doses up to
1000 mg/kg. Based upon PK modeling of the initial inhibitor series,
structural modifications were made to optimize for in vivo stabil-
ity, and the orally available and CNS permeable inhibitor 3f was
discovered. This compound displayed peak concentrations in plas-
14. Troeberg, L.; Morty, R. E.; Pike, R. N.; Lonsdale-Eccles, J. D.; Palmer, J. T.;
McKerrow, J. H.; Coetzer, T. H. Exp. Parasitol. 1999, 91, 349.
15. Manganaro, M.; Mascellino, M. T.; Gradoni, L. Parasite 2003, 10, 147.
16. Bitonti, A. J.; Byers, T. L.; Bush, T. L.; Casara, P. J.; Bacchi, C. J.; Clarkson, A. B., Jr.;
McCann, P. P.; Sjoerdsma, A. Antimicrob. Agents Chemother. 1990, 34, 1485.
17. Mallari, J. P. S. A.; Caffrey, C.; Kosinski, A.; Connely, M.; McKerrow, J. H.; Guy, R.
K. J. Med. Chem. 2009, 52, 6489.
18. Mallari, J. P.; Shelat, A. A.; Obrien, T.; Caffrey, C. R.; Kosinski, A.; Connelly, M.;
Harbut, M.; Greenbaum, D.; McKerrow, J. H.; Guy, R. K. J. Med. Chem. 2008, 51,
545.
19. Mackey, Z. B.; O’Brien, T. C.; Greenbaum, D. C.; Blank, R. B.; McKerrow, J. H.
J. Biol. Chem. 2004, 279, 48426.
ma and brain tissue of approximately 10
lM and of approximately
20 M in brain tissue. These inhibitor levels were maintained for
l
20. O’Brien, T. C.; Mackey, Z. B.; Fetter, R. D.; Choe, Y.; O’Donoghue, A. J.; Zhou, M.;
Craik, C. S.; Caffrey, C. R.; McKerrow, J. H. J. Biol. Chem. 2008, 283, 28934.
21. Di, L.; Kerns, E. H.; Li, S. Q.; Petusky, S. L. Int. J. Pharm. 2006, 317, 54.
22. Ionsource Principles of MS Quantitation. http://www.ionsource.com/tutorial/
msquan/requantoc.htm.
23. Institute of Laboratory Animals, G. S. o. M., Kyoto-University, Japan Kyoto Rat
Phenotype Database, The National Bioresource Project for the Rat in Japan.
http://www.anim.med.kyoto-u.ac.jp/NBR/strainsx/FOB_menue.aspx.
24. Shin, S. Y.; Lee, S. H.; Yang, S. T.; Park, E. J.; Lee, D. G.; Lee, M. K.; Eom, S. H.;
Song, W. K.; Kim, Y.; Hahm, K. S.; Kim, J. I. J. Pept. Res. 2001, 58, 504.
25. Blanchard, J. J. Chromatogr. 1981, 226, 455.
over 24 h following a single oral dose of 200 mg/kg. In addition,
no negative symptoms were observed for this inhibitor in a chronic
toxicity study in which mice were given 200 mg/kg of inhibitor
daily by oral gavage for 10 consecutive days. Finally, this inhibitor
was efficacious in a murine disease model, significantly extending
the survival time of T.b. brucei infected mice relative to a vehicle
control. However, the same compound was not efficacious in T.b.
rhodesiense infected mice, most likely due to lower potency against
the human parasite relative to T.b. brucei. This inhibitor is signifi-
cant in that it represents one of the only examples of a novel
26. Clarke, S. E.; Jeffrey, P. Xenobiotica 2001, 31, 591.
27. Abdulla, M. H.; O’Brien, T.; Mackey, Z. B.; Sajid, M.; Grab, D. J.; McKerrow, J. H.
PLoS Negl. Trop. Dis. 2008, 2, e298.