Identification of Inhibitors of the Leishmania cdc2-Related Protein Kinase CRK3

Article abstract of DOI:10.1002/cmdc.201100344

New drugs are urgently needed for the treatment of tropical parasitic diseases such as leishmaniasis and human African trypanosomiasis (HAT). This work involved a high-throughput screen of a focussed kinase set of ～3400 compounds to identify potent and parasite-selective inhibitors of an enzymatic Leishmania CRK3-cyclin6 complex. The aim of this study is to provide chemical validation that Leishmania CRK3-CYC6 is a drug target. Eight hit series were identified, of which four were followed up. The optimisation of these series using classical SAR studies afforded low-nanomolar CRK3 inhibitors with significant selectivity over the closely related human cyclin dependent kinase CDK2. Selective at last: [1,2,4]Triazolo[1,5-a]pyridines, aminopyrazoles, and disubstituted ureas were explored as potent and parasite-selective inhibitors of an enzymatic Leishmania CRK3-cyclin6 complex. Optimisation of these series using classical SAR studies afforded low-nanomolar CRK3 inhibitors with significant selectivity over the closely related human cyclin-dependent kinase CDK2.

Full text of DOI:10.1002/cmdc.201100344

MED
DOI: 10.1002/cmdc.201100344
Identification of Inhibitors of the Leishmania cdc2-Related
Protein Kinase CRK3
Laura A. T. Cleghorn,[a] Andrew Woodland,[a] Iain T. Collie,[a] Leah S. Torrie,[a] Neil Norcross,[a]
Torsten Luksch,[a] Chido Mpamhanga,[a] Roderick G. Walker,[b] Jeremy C. Mottram,[b]
Ruth Brenk,[a] Julie A. Frearson,[a] Ian H. Gilbert,[a] and Paul G. Wyatt*[a]
New drugs are urgently needed for the treatment of tropical
parasitic diseases such as leishmaniasis and human African try-
panosomiasis (HAT). This work involved a high-throughput
screen of a focussed kinase set of ~3400 compounds to identi-
fy potent and parasite-selective inhibitors of an enzymatic
Leishmania CRK3–cyclin 6 complex. The aim of this study is to
provide chemical validation that Leishmania CRK3–CYC6 is a
drug target. Eight hit series were identified, of which four were
followed up. The optimisation of these series using classical
SAR studies afforded low-nanomolar CRK3 inhibitors with sig-
nificant selectivity over the closely related human cyclin de-
pendent kinase CDK2.
Introduction
Neglected diseases are a major global cause of illness and
death worldwide.[1] Kinetoplastid parasites, the causative
agents of leishmaniasis, African sleeping sickness, and Chagas’
disease are a major cause of morbidity and mortality in devel-
oping countries. Approximately 350 million people are estimat-
ed to be at risk of leishmaniasis, with an estimated 12 million
people infected worldwide.[2] Despite this disease burden,
there is a lack of validated drug discovery targets and lead
compounds for these neglected diseases.[1] Current therapeutic
options are limited and unsatisfactory; therefore, there is an
urgent need to discover new drugs for the treatment of major
tropical parasitic diseases.
Many human CDK inhibitors have been developed and are
currently undergoing clinical trials despite the fact that the
CDKs inhibited by these agents can be genetically knocked out
without apparent major phenotypic changes.[9] This highlights
the need for chemical as well as genetic validation. Leishmania
CRK3–CYC6 inhibitors with micromolar potency were recently
reported by Walker et al., following a high-throughput screen
with heterocyclic and kinase libraries.[11] Grant et al. previously
described the screening of a diverse chemical library of antimi-
totic compounds for potential inhibitors of Leishmania CRK3.[12]
Although relatively successful, the broad-spectrum inhibitors
identified failed to show selectivity over the mammalian
CDK1–CYCB complex, and were in many cases equally or more
potent against CDK1.[12]
Kinases are well-known targets for a variety of diseases and
disorders,[3] but they are underexplored as targets for neglect-
ed diseases. Kinetoplastid parasites have a significant kinome
(approximately 190 protein kinases) with major differences
from the human kinome (table S1, Supporting Information).
The parasites have a large number of kinases involved in cell-
cycle control, and this may be a reflection of the complex life
cycle of these organisms. Cyclin-dependent cdc2-related kinas-
es (CRKs) are the most investigated kinase family in this
area.[4–7] The parasitic cyclin-dependent cdc2-related serine/
threonine protein kinase (CRK3) has been postulated as a po-
tential drug target for kinetoplastid species.[4] The CRK3–
cyclin 6 complex (CRK3–CYC6) is thought to have homologous
function to the cyclin-dependent kinase 1–cyclin B complex
(CDK1–CYCB) in humans,[8,9] which is essential for proliferation
and coordination of the eukaryotic cell cycle. Gene disruption
studies carried out in Leishmania mexicana to determine the
necessity of CRK3 demonstrated that both CRK3 alleles can be
replaced, although ploidy changes occurred to allow retention
of the wild-type copy of CRK3, suggesting, but not confirming,
that CRK3 is an essential enzyme for transition through the G2/
M phase checkpoint of the Leishmania cell cycle responsible
for parasite growth and survival.[8,10]
The aim of this study was to identify novel and selective
small-molecule inhibitors of Leishmania CRK3 to act as chemi-
cal probes for investigating the essentiality of Leishmania
CRK3. There are multiple Leishmania CDKs and cyclins, and
each kinase can form an active enzyme complex with more
than one cyclin. This study focussed on the L. mexicana CRK3–
[a] Dr. L. A. T. Cleghorn, Dr. A. Woodland, I. T. Collie, Dr. L. S. Torrie,
Dr. N. Norcross, Dr. T. Luksch, Dr. C. Mpamhanga, Dr. R. Brenk,
Prof. J. A. Frearson, Prof. I. H. Gilbert, Prof. P. G. Wyatt
Drug Discovery Unit, College of Life Sciences, James Black Centre
University of Dundee, Dundee, DD1 5EH (UK)
Fax: (+44)1382 386373
E-mail: p.g.wyatt@dundee.ac.uk
[b] R. G. Walker, Prof. J. C. Mottram
Wellcome Centre for Molecular Parasitology and Division of
Infection & Immunity, Faculty of Biomedical and Life Sciences
University of Glasgow, 120 University Place, Glasgow, G12 8TA (UK)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201100344.
Re-use of this article is permitted in accordance with the Terms and Con-
ditions set out at http://onlinelibrary.wiley.com/journal/
10.1002/(ISSN)1860-7187/homepage/2452_onlineopen.html.
2214
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2214 – 2224

CRK3 Inhibitors
L. major CYC6 complex (the L. mexicana CRK3 gene is 99% sim-
ilar to that of L. major), although other CRK–cyclin complexes
could be potential targets. Compounds were counter-screened
against the human cyclin-dependent kinase CDK2–cyclin A
(CYCA) complex. This was used instead of CDK1–CYCB, the
closest homologue of CRK3, as in our experience, CDK2 has
proven to be a more promiscuous enzyme than CDK1, making
it a more rigorous measure of selectivity. To check for general
toxicity, a mammalian cell line was used as a secondary coun-
ter-screen.
shown). Because the primary screen (IMAP assay—see Enzyme
assays in the Experimental Section) conditions used >10% of
substrate during the course of the reaction, the potency of hits
was reconfirmed with an orthodox “gold standard” radiometric
Earlier we reported the assembly of a focussed set of poten-
tial kinase inhibitors to identify chemical starting points for
kinase inhibitor discovery.[13] This focussed set of kinase inhibi-
tors was screened against Leishmania CRK3–CYC6 to discover
leads that can be optimised into suitable probes to chemically
validate CRK3 as a drug target.
Results and Discussion
Binding site analysis of Leishmania CRK3
Figure 1. Superposition of the ligand binding sites Leishmania CRK3 (blue
carbon atoms) homology model with a crystal structure of HsCDK2 (green
carbon atoms, PDB code 1PYE[19]) in complex with a ligand (bright-green
carbon atoms). Residues Phe82, Leu83, His84, and Gln85 in CDK2 are re-
spectively replaced with Tyr, Val, Glu, and Ala for Leishmania CRK3. The con-
struction of the homology model and methods used for visualisation are de-
scribed in the Experimental Section.
In the absence of crystal structures of the kinetoplastid CRK3s,
a homology model of L. mexicana CRK3 was built by using
human CDK2 (HsCDK2) as the template. Analysis of the ATP
binding sites revealed small differences between HsCDK2 and
Leishmania CRK3 (Figure 1). The main divergence in
amino acid side chains facing the ligand is the re-
spective replacement of Phe82 and Leu83 in HsCDK2
with tyrosine and valine. In addition, Gln85 is re-
placed with alanine in Leishmania CRK3, and His84
with glutamate. However, the latter changes are pres-
ent for amino acids where the side chains are orient-
ed away from the ligand binding site, and therefore
these probably have only a minor effect on binding.
This analysis shows that subtle changes could be ex-
plored to achieve selective inhibition of Leishmania
CRK3 over HsCDK2.
Primary screen of focussed kinase library
Our in-house kinase library containing 3383[12] com-
pounds was screened against Leishmania CRK3–CYC6
at a concentration of 30 mm. The 11 primary assay
screen plates generated a robust mean (ꢀSD) Z’
value of 0.77 (ꢀ0.04) and a mean staurosporine po-
tency (95% confidence interval) of 29 nm (27–32 nm).
The screen identified 73 compounds with inhibition
values of ꢁ40%, with 40% representing a statistically
significant threshold (>3ꢁSD of the mean of the un-
inhibited control signal across all screening plates)
for hit identification. These compounds were pro-
gressed into potency determination studies using 10-
point dilution curves. Of these, 46 compounds gave
IC50 values of ꢂ30 mm, with the most potent com-
Figure 2. Representative compounds from hit series identified by screening the DDU fo-
cussed kinase set against the Leishmania CRK3–CYC6 complex using an IMAP assay plat-
form with fluorescence polarisation detection, as described in the Experimental Section.
pound returning an IC50 value of 0.24 mm. Analysis of
the whole data set identified eight compound series
(Figure 2) and seven singletons of interest (data not The common substructure for each series is shown in bold.
ChemMedChem 2011, 6, 2214 – 2224
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2215

P. G. Wyatt et al.
MED
secondary assay platform for Leishmania CRK3–CYC6. Initial se-
lectivity was assessed using a similar radiometric HsCDK2 assay.
The secondary assays were run at an ATP concentration of <
KM for ATP, such that the IC50 value measured approximates to
Ki, thereby providing an accurate assessment of selectivity. De-
spite the 100-fold decrease in ATP substrate concentration be-
tween the IMAP-FP assay and the radiometric assay resulting
in an apparent increase in potency values, there was a good
correlation of potency values between the two platforms, and
the rank order of hit potencies was retained (figure S1, Sup-
porting Information). Mean staurosporine potency (95% confi-
dence interval) in the radiometric Leishmania CRK3 assay was
9.1 nm (8.0–10.5 nm) and 0.36 nm (0.31–0.43 nm) for HsCDK2.
The structure and purity of the hits taken into potency deter-
minations were confirmed using LC–MS, and key compounds
were re-synthesised to confirm the initial potency data. Hit
series 3, 5, 7, and 8 were followed up in hit-to-lead studies,
and the structure–activity relationship (SAR) results are report-
ed herein. Series 1 and 2 were not pursued further due to lack
of potency, series 6 was more potent against the HsCDK2
counter-screen (IC50 =1 mm) than Leishmania CRK3, and series 4
was not pursued further due to the potential Michael acceptor
moiety.
human kinases, it proved to be a nonselective inhibitor with
>50% inhibition at 10 mm for 47 of these human kinases. Due
to the lack of selectivity for parasite kinases over host kinases
and activity not being translated into cells in culture, this
series was not developed further (table S4, Supporting Infor-
mation).
Compound series 5
The set of focussed kinase inhibitors contained only seven ex-
amples of the aminopyrazole template (series 5, Figure 2), and
these showed interesting levels of activity and potential SAR
patterns (Figure 2 and Table 1, compounds 1–6). A crystal
structure of HsCDK2 in complex with a pyrazole amide scaf-
fold-based inhibitor was published previously (PDB code
1VYW).[12] Based on this structure, binding modes for pyrazole
amides bound to Leishmania CRK3 were modelled (Figure 3). It
was assumed that the pyrazole amide scaffold adopts the
same binding mode in Leishmania CRK3 as in HsCDK2 with all
nitrogen atoms forming hydrogen bonds with the amino acid
backbone of the hinge region. As a consequence, the substitu-
ents at the R1 position are oriented towards the gatekeeper
residue (Phe80), and substituents at the R2 position within a
lipophilic pocket are directed out towards the solvent (R1 and
R2 group definitions are listed in Table 1).
Compound series 3
The model appears to rationalise the binding data. Thus a
3,5-dimethoxyphenyl substituent of the pyrazole 1 (Table 1)
showed poor activity, presumably due to the inability of the
gatekeeper residue position to accommodate a group of this
size. The more linear 4-chlorophenyl group of compound 2
can be accommodated with an increased affinity relative to 3,
probably due to improved edge–face interactions of the elec-
tron-poor system with the gatekeeper. On the other side of
the active site, the lipophilic pocket can bind the phenyl
amide of 4 and the bicyclic moiety of 5, whereas larger linear
groups such as the phenoxybenzamide of 6 led to a complete
loss of binding affinity. This observation is consistent with the
The primary screen identified a range of pyrazolo[1,5-a]pyrimi-
dines (series 3, Figure 2) with modifications at both the 3- and
7-positions. The 35 compounds tested with aromatic substitu-
ents at the 3-position and diverse structural types from the
amine at the 7-position gave inhibition values ranging from 0
to 93% at 30 mm, suggesting that SAR could be derived in this
series. To confirm the data obtained from the screening sam-
ples, a further array of compounds was synthesised (table S2,
Supporting Information); these, however, displayed only mod-
erate potency and poor selectivity over HsCDK2. When an ex-
ample from the series was profiled against a panel of 76
Table 1. Leishmania CRK3–CYC6 and HsCDK2–CYCA activity of pyrazoles 1–11 (Series 5).
Compd
R1
R2
CRK3–CYC6 Inhib. [%][a]
CRK3–CYC6 IC50 [mm][b]
HsCDK2–CYCA IC50 [mm][c]
1
2
3
4
5
6
7
8
9
3,5-dimethoxyphenyl
4-chlorophenyl
phenyl
phenyl
phenyl
methyl
methyl
methyl
0
74
17
55
86
7
99
88
44
100
100
–
3.0
–
2.4
0.65
–
–
1.6
–
–
1.6
4-methoxyphenyl
2-(2,3-dihydrobenzo[b][1,4]dioxin
4-phenoxyphenyl
cyclobutyl
benzyl
phenyl
methyl
–
1.2
4-chlorophenyl
4-chlorophenyl
4-chlorophenyl
4-chlorophenyl
4-chlorophenyl
0.9
3.0
>100
2.4
0.49
>100
5.1
10
11
3-pyridyl
cyclohexyl
0.34
>100
[a] Percent inhibition of Leishmania CRK3–CYC6 activity at 30 mm. [b] Concentration required to inhibit Leishmania CRK3–CYC6 activity by 50%; data repre-
sent the mean of two or more experiments. [c] Concentration required to inhibit CDK2–CYCA activity by 50%; data represent the mean of two or more ex-
periments.
2216
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2214 – 2224

CRK3 Inhibitors
and solid-supported scavengers; this allowed the rapid synthe-
sis of 5-substituted 3-aminopyrazole-3-yl amides (Table 1, 7–
11). Amines 13 were treated with acid chlorides, and then
excess 13 was scavenged using polymer-supported (PS) isocya-
nate resin. The crude products were treated with base to hy-
drolyze any over-acylated products. Extraction and concentra-
tion yielded the desired final compounds 7–11 in 2–80% yield
without the need for further purification in most cases.
Modification of the amide substituent was examined by
using 4-chlorophenyl (R1) derivatives (Table 1, synthesised com-
pounds 7–11). Introduction of small carbocyclic rings (com-
pound 7) resulted in a threefold improvement in Leishmania
CRK3 activity relative to the initial screening hit 2, whereas ex-
tension to the corresponding benzyl derivative 8 had no
effect. Increasing ring size to phenyl 9 from cyclobutyl 7 result-
ed in a complete loss of activity against both enzymes, a modi-
fication well tolerated for HsCDK2 with the smaller cyclopropyl
substituent at the 5-position.[12,13] Interestingly, the 3-pyridyl
amide 10 did not suffer from loss of activity, possibly through
significant interaction either directly or through water-mediat-
ed hydrogen bonds. The cyclohexyl amide 11 afforded a 10-
fold increase in activity against Leishmania CRK3, but more im-
portantly, diminished HsCDK2 activity, resulting in a compound
with significant selectivity for Leishmania CRK3–CYC6, although
we cannot explain the detailed reasons for this from our
model.
Figure 3. Crystal structure of a pyrazole-based inhibitor (grey carbon atoms)
of HsCDK2 (PDB code 1VYW, green carbon atoms)[12] superimposed with the
modelled binding mode of 6 (pink carbon atoms) bound to Leishmania
CRK3 (blue carbon atoms).
binding mode of pyrazole amides in HsCDK2 and previously
established SAR data (Figure 3),[14,15] where the most potent
HsCDK2 inhibitors in this series contain an sp3 carbon atom
next to the amide group which induces an sp3 bend in the
molecule. This bend allows the accommodation of bulky sub-
stituents in this area of the binding site, whereas more linear
moieties lead to a steric clash. Although Leishmania CRK3 and
HsCDK2 share high sequence identity, inhibitor selectivity be-
tween closely related human CDK family members is well
known.[16–18] We therefore decided to pursue the preparation
of compounds selective for the parasite kinase. A small set of
compounds (7–11, Table 1) were synthesised to examine
whether selectivity could be achieved. We focussed on modify-
ing the 5-position substituent (R1) to investigate the interaction
with the gatekeeper residue (Phe80 in HsCDK2), and the
amide (R2) to probe the lipophilic pocket along the backbone
of the kinase.
As compound 11 was the most potent compound against
Leishmania CRK3–CYC6 and highly selective over HsCDK2, we
decided to retain the cyclohexyl group on the amide and then
optimise the substituent at position R1. Further modification of
R1 of 11 either as cycloalkyl, substituted phenyl, heteroaryl, or
benzyl did not afford an increase in Leishmania CRK3 activity
(table S3, Supporting Information). Comparing the SAR gener-
ated within our studies with that reported for HsCDK2 by Pe-
varello et al.[14,15] suggested some subtle differences in the
structure of HsCDK2 and Leishmania CRK3, despite their close
homology. Compound 11 proved to be highly selective against
a panel of 76 human kinases, with only three kinases—ERK8,
GSK3b and IRR—showing >50% inhibition at 10 mm (table S5,
Supporting Information).
Chemistry
The synthesis of the heterocyclic core (Scheme 1) was carried
out through condensation of acetonitrile with an appropriate
ester to give 2-cyanoketones 12 in 26–94% yield. Reaction of
12 with hydrazine hydrate in ethanol at reflux gave 5-substitut-
ed pyrazole-3-amines 13 in yields of 30–77%. A general acyla-
tion method was developed that involved microwave heating
Compound series 7
[1,2,4]Triazolo[1,5-a]pyridine 14 (Figure 3) was identified as an
initial hit from primary screening. Compound 14 had an IC50
value of 0.24 mm against Leishmania CRK3 but also inhibited
with similar potency in the HsCDK2 counter-screen (IC50 =
0.36 mm). [1,2,4]Triazolo[1,5-a]pyridine-2-ylamines are a novel
structural scaffold for Leishmania CRK3 inhibition and none
have been reported thus far for antiparasitic activity.
Putative binding mode analysis
Scheme 1. Synthesis of compounds 7–11. Reagents and conditions: a) LiN-
(SiMe3)2, THF, ꢃ788C, 1 h, then addition of R1CO2Me; b) N2H4·H2O, EtOH,
708C, 24 h; c) R2COCl (1.1 equiv), 1,4-dioxane, 1508C, 5 min (MW), then addi-
tion of PS isocyanate (2 equiv), 1608C, 5 min (MW); d) 1:1 EtOAc/NaOH (2m
solution in H2O), RT, 1 h.
We carried out modelling studies to guide the synthetic pro-
gramme. Triazolopyridine 14 was superimposed onto the
known
HsCDK2
aminoimidazo[1,2-a]pyridine
inhibitor
(Figure 4).[19] This was followed by energy minimisation in the
ChemMedChem 2011, 6, 2214 – 2224
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2217

P. G. Wyatt et al.
MED
ester 16. Cyclisation to give the
triazolopyridine scaffold 17 was
carried out by reaction with hy-
droxylamine hydrochloride in
the presence of base.[19] Diversity
can be introduced from this
point; reaction of amine 17 with
an acid or acid chloride affords
amide 18, which undergoes
facile
microwave-irradiated
Suzuki reaction with a series of
aryl boronic acids to give 20. Re-
versal of the reaction order
allows the R1 group to be kept
constant, with variation of R2 (17
to 19 to 20).
Figure 4. Overlap of the 5,6 bicyclic core of the known aminoimidazo[1,2-a]pyridine HsCDK2 inhibitor (pink) with
14 (cyan). Dashed lines represent the proposed hydrogen bond interaction with the hinge region.
binding site using the MAB force field (see the Experimental
Enzyme activity
Section). The result was a binding mode that shows a classic
kinase inhibitor hydrogen bond donor/acceptor binding pat-
tern to the hinge region.[20] Although good alignment was ob-
served for the core motifs, the substituents occupy different
areas of the binding site. The proposed binding mode includes
the following key interactions: a hydrogen bond between the
amide NꢃH of 14 and the carbonyl group of Val102 (HsCDK2)
(or Leu in Leishmania CRK3) (Figure 4), the triazolopyridine ni-
trogen hydrogen bonded to the a chain NꢃH group of Phe82
(HsCDK2), and a possible interaction between the 6-phenyl ar-
omatic hydrogen atom of 14 and the backbone carbonyl
group of Glu81 in both HsCDK2 and Leishmania CRK3 (a CꢃH
hydrogen bond). This putative binding mode was used as a
starting point for the rational design of inhibitors exploring
two main areas in the binding site: the hydrophobic gatekeep-
er region and the lipophilic pocket leading to solvent (the
region next to Gln85 in Figure 4). The substituents of the
amide and the aromatic ring at the 6-position of the triazolo-
pyridine scaffold 14 were investigated to optimise potency
against Leishmania CRK3 and Leishmania in culture while intro-
ducing selectivity over HsCDK2.
Initial work focussed on varying R1 (which we believe may in-
teract with the gatekeeper residue) whilst maintaining R2 as cy-
clohexyl (Table 2). Removal of the aromatic ring was not toler-
ated, with loss of activity in both Leishmania CRK3 and HsCDK2
with R1 =H. A phenyl group at R1 (compound 22) demonstrat-
ed that it is possible to achieve selectivity and retain potency.
By examining the substituents on the initial hit 14 with 4-hy-
droxyphenyl 23 and 3-methoxyphenyl 24 separately, 4-hydroxy
was found to be responsible for retention of HsCDK2 activity,
while monosubstituted 3-methoxy was inactive against
HsCDK2. Although not apparent in our analysis of the 3D
models, this result implies that there is the potential for
(water-mediated) hydrogen bonding to the hydroxy group,
and that this is not essential for Leishmania CRK3 activity. 3-
Methyl compound 25 was found to give the best potency and
selectivity (0.064 mm Leishmania CRK3, >100 mm HsCDK2)
among those compounds investigated in the initial plan.
The homology model suggested that bulkier aromatic and
hydrophobic R1 groups could be accommodated in the region
of the pocket proximal to the gatekeeper. We postulated that
potency could be increased by exploiting this space; therefore,
further studies focussed on this hypothesis and developing the
Chemistry
The initial synthetic strategy re-
quired the synthesis of the
common intermediate 6-bromo-
[1,2,4]triazolo[1,5-a]pyridine-2-
amine (17). From this point,
either R1 or R2 can be kept con-
stant, enabling arrays to be per-
formed by varying the other
substitution vector (Scheme 2).
Treatment of commercially avail-
Scheme 2. Reagents and conditions: a) Ethoxycarbonylisothiocyanate (1 equiv), dioxane (3 mLmmolꢃ1), under
argon, 08C!RT, 17 h; b) NH2OH·HCl (5 equiv), DIPEA (3 equiv), EtOH/MeOH 1:1 (3.76 mLmmolꢃ1), 2 h RT then 3 h
608C; c) acid chloride (1 equiv), DMAP (0.1 equiv), pyridine (4 mLmmolꢃ1), 08C!RT, 17 h;[27] d) boronic acid
(1.2 equiv), DMF/2m K2CO3 (1:1), Pd(PPh3)4 (0.01 equiv), 1408C (MW), 15 min (see Experimental Section for specific
able 2-amino-5-bromopyridine
15 with ethoxycarbonylisothio-
cyanate afforded the corre-
sponding aminothiomethylethyl reagents).
2218
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2214 – 2224

CRK3 Inhibitors
the R1 group (Table 3). The free amine 35 was inactive against
Leishmania CRK3 but retained weak HsCDK2 activity; removal
of the amide carbonyl 36 retained only moderate Leishmania
CRK3 activity and gave weak HsCDK2 activity. Replacement of
the lipophilic cyclohexyl group with a methyl substituent 37
showed a large drop in activity and similar levels of HsCDK2
activity. 2-Tetrahydrofuran, 4-tetrahydropyran, and 3-piperidine
(compounds 38–40) were incorporated to explore the possibil-
ity of water-mediated interactions to residues such as Tyr,
which replaces Phe82 in the hinge region of Leishmania CRK3
(Figure 4). Although these compounds retained selectivity, they
showed low-micromolar Leishmania CRK3 activity. This set of
compounds confirm that the lipophilic nature of the cyclohexyl
substituent is important; replacement of the heteroatom with
a carbon atom gives a minimal 10-fold increase in activity
(compare compounds 38 and 39 with 43, and 40 with 25).
These data imply the cyclohexyl substituent is in the lipophilic
pocket as suggested by the model, giving an enhancement in
potency commonly observed for HsCDK2 inhibitors.[21] Rigidifi-
cation of the cyclohexyl ring by replacement with a planar
phenyl substituent retains selectivity, but results in a 10-fold
loss in activity (compound 41). Generally, increasing the ring
size of the lipophilic probe increases the potency (Table 3, 42–
48). This affinity gain may be credited to a gain in entropy due
to the displacement of water molecules from this region of the
binding site. Insertion of a carbon linker between the amide
and the cyclic alkyl group improved activity (44–46); methyl
gave ~2-fold and ethyl ~30-fold improvements. This trend was
also observed with the phenyl substituents [phenyl 41 (IC50 =
1.1 mm), phenethyl 47 (IC50 =0.03 mm)]. Combining the optimal
groups from the SAR studies of the aromatic and amide sub-
stituents gave the expected low-nanomolar activities (Table 3,
Table 2. Leishmania CRK3–CYC6 and HsCDK2–CYCA activity (compounds
14–39) from initial SAR development at the R1 position of triazolopyridine
19.
Compd R1
CRK3–CYC6
IC50 [mm][a]
HsCDK2–CYCA
IC50 [mm][b]
14
4-hydroxy-3-methoxy-
0.23
0.36
phenyl
21
22
23
24
25
26
27
28
29
30
31
32
33
34
H
>100
0.27
0.12
0.36
0.06
0.35
0.09
0.07
0.04
1.6
>100
>100
0.86
phenyl
4-hydroxyphenyl
3-methoxyphenyl
3-methylphenyl
3-chlorophenyl
1-napthyl
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
4-indole
4-N-methylindole
3-morpholinophenyl
2,6-dimethylphenyl
2,5-dimethylphenyl
3,5-dimethylphenyl
4-methoxy-3-methyl-
phenyl
19.0
0.65
0.02
0.03
[a] Concentration required to inhibit Leishmania CRK3–CYC6 activity by
50%; data represent the mean of two or more experiments. [b] Concen-
tration required to inhibit HsCDK2–CYCA activity by 50%; data represent
the mean of two or more experiments.
SAR around the 3-methylphenyl substituent (Table 2). 1-Napth-
yl 27, 4-indole 28, and 4-N-methyl indole 29 gave results simi-
lar to that of the 3-methyl, suggesting that there is a limited
benefit. 3-Morpholine 30 was
synthesised to explore the
Table 3. Leishmania CRK3–CYC6 and HsCDK2–CYCA activity from initial SAR investigations at the R position of
triazolopyridine 19 (Scheme 3) and a benzimidazole core.
region above the gatekeeper for
favourable and potentially selec-
tive polar interactions; complete
selectivity for Leishmania CRK3
over HsCDK2 was observed, but
with only moderate enzyme ac-
tivity. Finally an additive ap-
proach with disubstituted aro-
matics was examined: 2,6- and
2,5-dimethyl groups (31 and 32)
were detrimental to activity;
however, both 3,5-dimethyl 33
and 4-methoxy-3-methyl 34
showed a two- to threefold im-
provement in potency. These re-
Compd
35
Core
R
CRK3–CYC6 IC50 [mm][a] HsCDK2–CYCA IC50 [mm][b]
A
free NH2, no amide
cyclohexylmethyl, no amide
methyl
2-tetrahydrofuran
4-tetrahydropyran
3-piperidine
phenyl
cyclobutyl
cyclopentyl
cyclopentylmethyl
cyclohexylmethyl
cyclopentylethyl
phenethyl
>100
3.0
37
64
36
37
38
39
40
41
42
43
44
45
46
47
48
A
A
A
A
A
A
A
A
A
A
A
A
B
65
12
10
12
1.1
3.3
1.1
0.04
0.01
<0.005
0.03
<0.005
26.0
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
sults imply
a bulky aromatic
group favours inhibition of
Leishmania CRK3, and not
HsCDK2.
cycloheptyl
SAR of the amide moiety was
examined by employing the
previously discovered 3-methyl-
phenyl or 3,5-dimethylphenyl as
[a] Concentration required to inhibit Leishmania CRK3–CYC6 activity by 50%; data represent the mean of two
or more experiments. [b] Concentration required to inhibit HsCDK2–CYCA activity by 50%; data represent the
mean of two or more experiments.
ChemMedChem 2011, 6, 2214 – 2224
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2219

P. G. Wyatt et al.
MED
48), showing that potency against Leishmania CRK3 can be im-
proved to the low-nanomolar range whilst retaining selectivity
over HsCDK2.
Kinase selectivity of compound series 7
Scheme 3. Synthesis of ureas 50, 52–61. Reagents and conditions: a) RꢃN=
C=O, 1,4-dioxane, 908C, 12 h.
Compounds 25, 29, 33, and 34 were screened against the
DSTT Dundee kinase panel (76 human kinases)[22] and were
found to be highly selective. HsCDK2–CYCA is in the Dundee
human kinase panel, and none of the compounds showed in-
hibition of this, confirming the selectivity observed in our
assays over the human orthologue. For 25, 29, 33, and 34 the
main kinases showing >50% inhibition at 10 mm were DYR 1A,
2, 3, and IRR. Compound 25 showed a very clean profile, only
mildly inhibiting MKK and IRR (table S6, Supporting Informa-
tion). These results confirm that the compound series is highly
selective for parasitic over human kinases, and does not afford
broad-spectrum kinase inhibition. Compound 25 was tested
against L. major wild-type promastigote cells, but limited activi-
ty was observed (IC50 =2–10 mm).
Table 4. Leishmania CRK3–CYC6 and HsCDK2–CYCA activity for the urea
series.
Compd[a]
R
CRK3–CYC6
IC50 [mm][b]
HsCDK2–CYCA
IC50 [mm][c]
50
52
53
54
55
56
57
58
59
phenyl
methyl
cyclohexylmethyl
cyclohexyl
2-methoxyphenyl
3-methoxyphenyl
4-methoxyphenyl
4-benzyloxyphenyl
4-benzoic acid methyl
ester
0.095
>100
3.9
>100
>100
>100
>100
>100
>100
>100
>100
>100
0.098
8.0
0.04
0.89
0.11
Compound series 8
Series 8 consisted of two structurally diverse disubstituted
ureas (Figure 5). Compound 49 displayed moderate Leishmania
CRK3 enzyme activity, and did not inhibit L. major wild-type
promastigote cells (EC50 >50 mm); however, none of the modifi-
cations investigated increased the activity against Leishmania
CRK3 or afforded any significant improvements in selectivity
over HsCDK2.
3.3
60
61
4-hydroxyphenyl
phenyl
3.3
>100
5.9
>100
[a] Compounds 50 and 52–60: 3-(5,5-dimethylcyclohex-2,3-ene)-1-one;
compound 61: 3-amino-cyclohex-2-eneone. [b] Concentration required to
inhibit Leishmania CRK3–CYC6 activity by 50%; data represent the mean
of two or more experiments. [c] Concentration required to inhibit
HsCDK2–CYCA activity by 50%; data represent the mean of two or more
experiments.
Enzyme activity
The requirement for a substantial lipophilic group was demon-
strated by compound 52 (R=methyl), which was inactive,
whereas the cyclohexylmethylene derivative 53 regained some
activity. Replacements of the phenyl ring of 50 with cyclohexyl
54 or substitution of the aromatic ring at the 3- or 4-positions
(56–58) were well tolerated overall, but did not show a signifi-
cant improvement in potency. However, substitution at the 2-
position (compound 55) or polar groups at the 4-position,
such as ester 59, resulted in up to a 100-fold loss of activity.
The majority of analogues demonstrated excellent selectivity
over HsCDK2, whereas compound 60, with a 4-hydroxy group
on the R2 phenyl ring, was unselective possibly due to favoura-
ble interactions with the water pocket near Asp45 in HsCDK2.
Modification of the cyclohexenyl moiety by removal of the
gem-dimethyl groups (compound 61) resulted in a complete
loss of activity.
Figure 5. Series 8 urea leads. Compound 49: Leishmania CRK3 IC50 =1.5 mm,
HsCDK2–CYCA IC50 =0.026 mm, MRC5 EC50 =12 mm; compound 50: Leishma-
nia CRK3 IC50 =0.16 mm, HsCDK2–CYCA IC50 =100 mm, MRC5 EC50 ꢁ50 mm.
Compound 50, with potent inhibition of Leishmania CRK3,
excellent selectivity over HsCDK2, and good predicted ADME
properties (see Table 5 below) represented a highly attractive
lead. To further optimise Leishmania CRK3 potency, we syn-
thesised an array of analogues by modifying the phenyl ring
(Scheme 3). The activities of compounds 50, 52–61 are sum-
marised in Table 4.
Chemistry
The ureas 50 and 52–61 were made by direct reaction of iso-
cyanates with amine 51 under elevated temperature.
2220
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2214 – 2224

CRK3 Inhibitors
Kinase selectivity
Conclusions
Compounds 58 and 59 proved to be highly selective against a
panel of 76 human kinases, with only 59 showing >50% in-
hibition at 10 mm against one kinase, ERK8.
Genetic manipulation studies have suggested that Leishmania
CRK3 kinase is essential; however, limited chemical validation
has been carried out. Screening of a focussed set of kinase in-
hibitors against Leishmania CRK3–CYC6 was used to identify
tool compounds suitable for chemical validation of this target.
The screen identified a number of series as starting points for
medicinal chemistry investigations. Optimisation of four of the
series resulted in the identification of compounds with signifi-
cant levels of activity against Leishmania CRK3–CYC6 with suit-
able selectivity against the highly homologous related human
kinase CDK2 for two series. Classical SAR studies allowed the
generation of very potent (nanomolar range) triazolopyridine
inhibitors as a novel series of potent and selective inhibitors of
Leishmania CRK3–CYC6 (IC50 <5 nm) with >10000-fold selec-
tivity over the human homologue HsCDK2. This is the first
report of low-nanomolar species-selective Leishmania CRK3 in-
hibitors. Unfortunately, no Leishmania CRK3 inhibitors from
these three series demonstrated significant activity against cell
cultures of the L. major parasite, indicating that the enzyme is
not essential for parasite proliferation. In addition, the lack of
antiparasitic activity of these potent inhibitors of Leishmania
CRK3 suggests that genetic studies alone are insufficient to
prove validity of Leishmania CRK3 as a drug target; chemical
validation is also required.
Antiparasitic activity
Key compounds from each series were assayed against
L. major wild-type promastigotes. Cell culture assays were per-
formed as described recently.[11] Only compound 25 from
series 7 translated into moderate cell activity; the other com-
pounds showed no activity up to 50 mm. Due to close se-
quence identity across the kinetoplastid (Leishmania and Trypa-
nosoma spp.) CRK3s, it was believed that these inhibitors may
be useful against other trypanosomatid parasites. Selected ex-
amples from series 3, 5, 7, and 8 were tested for antiprolifera-
tive activity against T. brucei brucei grown in culture. Unfortu-
nately, the activity against Leishmania CRK3–CYC6 did not
translate into cellular activity.
The reason for the lack of translation of Leishmania CRK3 in-
hibition into antiparasitic activity is unknown and cannot be
fully investigated, as assays to determine Leishmania CRK3 ac-
tivity in situ within parasites are not available. Unfortunately,
the lack of understanding of the signalling cascades within
these parasites means the downstream substrates of Leishma-
nia CRK3 are unknown, preventing the development of such
assays. The compounds identified across the main series stud-
ied (Table 5) have physicochemical properties regarded as suit-
able for membrane penetration, so an antiproliferative effect
against parasites would be expected to accompany inhibition
of Leishmania CRK3. Possible explanations for the lack of activi-
ty could be: 1) the compounds are excluded from the cells by
an efflux mechanism; 2) Leishmania CRK3 binds to a cyclin
other than CYC6 in the cell, forming an alternative complex for
which these compounds have much lower affinity; and
3) Leishmania CRK3 is not amenable to chemical inhibition, as
it is either not essential due to bypass mechanisms, or a suffi-
ciently high level of target inhibition cannot be achieved
chemically to observe a strong antiproliferative effect.
Experimental Section
Preparation of Leishmania CRK3:CYC6
Briefly, L. major CYC6 (LmjF32.3320; http://www.genedb.org) was
cloned into pET15b to give plasmid pGL1218; this provides an N-
terminal histidine tag. The L. major and L. mexicana CRK3 proteins
are identical, apart from a single Thr for Ala substitution at position
217. L. mexicana CRK3 was therefore used as a generic Leishmania
CRK3. L. mexicana CRK3 was cloned into pET28a to give pGL751a.
E. coli BL21 (DE3) pLys-S cells were transformed with plasmid
pGL1218 and then pGL751a to provide a CRK3–CYC6 co-expression
cell line. Cell culture (1 L volume) was induced with 1 mm IPTG at
198C overnight with agitation. After 16 h, cells were harvested by
centrifugation at 4000 g for 15 min and resuspended in ice-cold
PBS (pH 7.4) supplemented with DNAse 1 (10 mgmLꢃ1; Invitrogen)
and lysozyme (100 mgmLꢃ1; Sigma) for 60 min on ice. The cell
lysate was sonicated (four cycles of 30 s on/30 s off). The cell lysate
was harvested by centrifugation at 12000 g for 20 min, and the
soluble extract was filtered through a 0.2 mm filter syringe. The
proteins were purified by BioCAD chromatography using a metal
chelate Ni2+-charged column followed by a Hiload 16/60 Super-
dex-200 gel filtration column. The eluted CRK3–CYC6 complex was
stored in buffer (20 mm HEPES pH 7.4, 50 mm NaCl, 2 mm EGTA,
2 mm dithiothreitol (DTT), and 0.02% Brij-35) with 10% glycerol at
ꢃ808C.
Table 5. Properties of Leishmania CRK3–CYC6 inhibitors.[a]
Compd
Mr [Da] logP[b] logD[c] HBD[d] HBA[e] PSA[f] HIA[g]
11
46
50
56
306
348
258
340
4.5
4.6
3.2
3.1
4.5
4.6
3.2
3.1
2
1
2
2
4
5
4
5
58
64
58
67
+
+
+
+
CDK2
[a] Properties were calculated using the StarDrop software package
(http://www.optibrium.com/stardrop.php). [b] Partition coefficient. [c] Par-
tition coefficient at pH 7.2. [d] Number of hydrogen bond donors.
[e] Number of hydrogen bond acceptors. [f] Polar surface area. [g] Predic-
tion of human intestinal absorption sufficient for an orally dosed drug.
Recombinant full-length human CDK2 protein was expressed in an
E. coli expression system and contained an N-terminal glutathione
S-transferase (GST) tag. Human cyclin A2 (S171–L432) was also ex-
pressed in an E. coli expression system and contained an N-termi-
ChemMedChem 2011, 6, 2214 – 2224
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2221

P. G. Wyatt et al.
MED
nal GST tag. Both proteins were purified on a glutathione (GSH) Se-
pharose column, the tags were cleaved, and the cleaved proteins
were purified by gel filtration chromatography (30.4 and 34.1 kDa,
respectively). The CDK2–cyclin A2 complex was activated by using
active CAK1.
by the addition of 50 mm EDTA at 50 mL per well. FlashPlate trans-
fer, washing, and counting was as described above for CRK3. In
both primary and secondary assays, maximal signal (control) was
determined in the absence of inhibitors, whilst basal signal was de-
termined in the absence of enzyme. All assay plates included staur-
osporine as a reference inhibitor.
Enzyme assays
IMAP-FP Screening Express Kits and the carboxyfluorescein (5FAM)-
labelled peptide substrate (5FAM-GGGRSPGRRRRK-OH) were pur-
chased from Molecular Devices. [g-33P]ATP (cat.# NEG602H) and
streptavidin-coated FlashPlate PLUS 96-well plates (cat.# SMP103A)
were purchased from PerkinElmer, whilst the custom-made biotiny-
lated peptide substrate (biotin-PKTPKKAKKL-OH) used in the radio-
metric assay was supplied by Pepceuticals Ltd. All other reagents
were of analytical grade.
Data analysis
IMAP-FP output: fluorescence polarisation data were expressed in
millipolarisation (mP) units, defined as: mP=1000ꢁ[(SꢃGꢁ
P)/(S+GꢁP)], for which S (S-plane)=parallel-polarised light, P (P-
plane)=perpendicular-polarised light, and
ment- and assay-dependent factor.
G (grating)=instru-
For all compounds in the primary screen, data were normalised to
the average of the maximum uninhibited response (CTRL) as fol-
lows [Eq. (1)]:
CRK3 screen
The focussed kinase library of compounds was solubilised in DMSO
with a maximum final assay DMSO concentration of 1% (v/v). The
initial compound library screen was conducted using an immobi-
lised metal-ion affinity partitioning (IMAP) assay platform with fluo-
rescence polarisation detection (IMAP-FP) and a 5FAM-labelled
peptide substrate. Assays were carried out in black polystyrene
384-well plates at a final assay volume of 20 mL per well compris-
ing IMAP reaction buffer containing 1 mm DTT, 30 mm test com-
pound, 0.1 mm peptide substrate, 100 mm ATP, and 1 nm CRK3–
CYC6 enzyme complex. Reaction was initiated by the addition of
enzyme and incubation continued for 60 min at 208C before termi-
nation by the addition of 50 mL IMAP binding solution. Plates were
left to equilibrate at 208C for a further 30–60 min prior to reading
fluorescence polarisation (lex 480 nm/P/S, lem 535 nm) on an EnVi-
sion 2102 multilabel plate reader (PerkinElmer).
% Inhibition ¼ 1ꢃ½ðTestꢃMeanBASALÞ=ðMeanCTRLꢃMeanBASALÞꢄ ꢅ 100
ð1Þ
The quality of the primary screening assay was determined by cal-
culating the Z’ value as follows [Eq. (2)]:
Z0 ¼ 1ꢃ½ð3 SDCTRL þ 3 SDBASALÞ=ðMeanCTRLꢃMeanBASALÞꢄ
ð2Þ
For which SDCTRL and SDBASAL represent the standard deviations of
the response from 8–12 control wells (uninhibited signal) and 8–12
basal signal wells, respectively.
IC50 analysis
Radiometric assays
Data were normalised to percent inhibition and concentration–re-
sponse curves plotted using a four-parameter logistic fit equation
(model 205) in an ActivityBase (IDBS) template using XLfit.
CRK3 and CDK2: A radiometric assay was adopted for hit follow-up
and all subsequent compound potency determinations. CRK3
assays were carried out in polypropylene 96-well plates at a final
assay volume of 50 mL per well comprising 50 mm Tris-HCl buffer
(pH 7.5), 10 mm magnesium acetate, 0.2 mm EGTA, 2 mm DTT,
1 mgmLꢃ1 bovine serum albumin, 0.02% v/v Brij-35, test com-
pound, 1 mm peptide substrate, 1 mm ATP comprising 7.4 kBq per
well [g-33P]ATP, and 5 nm CRK3–CYC6 enzyme complex. Reaction
was initiated by the addition of the ATP/substrate cocktail, and
plates were incubated for 60 min at 208C on a plate shaker before
reaction termination by the addition of 50 mm EDTA at 50 mL per
well. An aliquot of 90 mL per well was transferred to 96-well strep-
tavidin-coated FlashPlates, which were then allowed to equilibrate
on a plate shaker for 20 min prior to washing with PBS containing
0.1% v/v Tween-20 (2ꢁ250 mL per well). Plates were sealed and
transferred to a TopCount NXT HTS scintillation counter (PerkinElm-
er) to quantify 33P incorporation. The counter-screen CDK2 assays
were also carried out in polypropylene 96-well plates at a final
assay volume of 50 mL per well comprising 50 mm Tris-HCl (pH 7.5)
containing 100 mm NaCl, 10 mm magnesium acetate, 0.2 mm
EGTA, 2 mm DTT, 0.02% v/v Brij-35, test compound, 1 mm peptide
substrate, 1 mm ATP comprising 7.4 kBq per well [g-33P]ATP, and
2 nm CDK2–CYCA enzyme complex. Reaction was initiated by the
addition of the ATP/substrate cocktail, and plates were incubated
for 60 min at 208C on a plate shaker before reaction termination
Molecular modelling
Homology modelling: CRK3 target sequences from T. brucei and
L. mexicana were aligned with the CDK2 sequence using T-
coffee.[23] Subsequently, Modeller 9.2[24] was used to build homolo-
gy models of CRK3, whereas the CDK2 crystal structures 1PYE[19]
and 1H1P[24] were used as templates. Modeller was run with default
settings, and only the highest-scoring structure was used to gener-
ate ligand binding modes.
Generation of putative binding modes: Relibase+[25] was used to
search for CDK2 crystal structures co-crystallised with ligands simi-
lar to the selected hit series compounds. By superimposing the
binding site of the CRK3 homology model onto the binding site of
the relevant crystal structure, an initial binding mode with CRK3
was generated. Using MOLOC, (Gerber Molecular Design), the crys-
tal structure ligand was modified to obtain the desired CRK3 inhib-
itor, and its position was optimised with the MAB force field.[26] To
allow for uncertainties in the homology model, the amino acid
side chains facing the binding sit were kept flexible during minimi-
sation.
2222
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2214 – 2224

CRK3 Inhibitors
and washed with H2O (2 mL) and Et2O (2ꢁ2 mL) to give 20 (38 mg,
0.13 mmol, 26%). H NMR ([D6]DMSO, 500 MHz): d=8.05–8.03 (2H,
Chemistry
1
1H NMR spectra were recorded on a Bruker Avance DPX 500 instru-
ment unless otherwise stated. Chemical shifts (d) are expressed in
ppm. Signal splitting patterns are described as singlet (s), broad
singlet (bs), doublet (d), triplet (t), quartet (q), multiplet (m), or
combination thereof. Low-resolution electrospray (ES) mass spectra
were recorded on a Bruker MicroTof mass spectrometer, run in pos-
itive ion mode, using either MeOH, MeOH/H2O (95:5), or H2O/
CH3CN (1:1)+0.1% formic acid as the mobile phase. High-resolu-
tion electrospray MS measurements were performed on a Bruker
MicroTof mass spectrometer. LC–MS analyses were performed with
an Agilent HPLC 1100 (Phenomenex Gemini Column 5 m C18 110A
50ꢁ3.0 mm, eluting with 20% MeOH/H2O, 0–3 min) and a diode
array detector in series with a Bruker MicroTof mass spectrometer.
All synthesised compounds were determined to be >95% pure by
LC–MS. Thin-layer chromatography (TLC) was carried out on Merck
silica gel 60 F254 plates using UV light and or KMnO4 for visualisa-
tion. TLC data are given as the Rf value with the corresponding
eluent system specified in brackets. Column chromatography was
performed using RediSep 4 or 12 g silica pre-packed columns. All
reactions were carried out under dry and inert conditions, unless
otherwise stated. Additional experimental procedures can be
found in the Supporting Information.
m, ArH), 7.80–7.79 (2H, m, ArH), 7.53–7.48 (5H, m, ArH), 6.86 ppm
(1H, s, ArH); LC–MS (ES) [M+H] (100%).
N-[5-(4-Chlorophenyl)-1H-pyrazol-3-yl]cyclohexanecarboxamide
(11): Prepared by following the same procedure as 8 to afford 11
1
(63 mg, 42%) as a white solid. H NMR ([D6]DMSO, 500 MHz): d=
12.92 (1H, s, NH), 10.54 (1H, s, NH), 7.75–7.73 (2H, m, ArH), 7.50–
7.49 (2H, m, ArH), 6.82 (1H, s, ArH), 2.39–2.33 (1H, m, CH), 1.78–
1.64 (5H, m, CH), 1.41–1.10 ppm (5H, m, CH); LC–MS (ES) [M+H],
(97%).
N-(6-Bromo[1,2,4]triazolo[1,5-a]pyridinyl)cyclohexanecarboxa-
mide (18): A mixture of triazolopyridine 16 (1.49 g, 7.0 mmol) and
DMAP (85 mg, 0.7 mmol, 10 mol%) in pyridine (anhydrous, 30 mL)
was cooled to 08C under argon, cyclohexylcarbonyl chloride
(0.94 mL, 1.03 g, 7.0 mmol) was added dropwise, and the mixture
was warmed to RT and stirred for 17 h. It was then concentrated in
vacuo, and the crude mixture was partitioned between Et2O
(20 mL) and H2O (10 mL), extracted with Et2O (2ꢁ10 mL), and the
organic layer (suspension) separated from the aqueous layer and
was passed directly through a filter funnel. The precipitate was col-
lected and dried in vacuo to afford pure cyclohexylamide 18
1
3-(4-Chlorophenyl)-3-oxopropanenitrile (12): CH3CN (1.5 mL,
29.4 mmol) was dissolved in dry THF (80 mL) under N2. The reac-
tion mixture was cooled to ꢃ788C, and LiN(SiMe3)2 (30 mL of a 1m
solution in toluene, 30 mmol) was added slowly so that the tem-
perature was kept below ꢃ608C. The reaction mixture was stirred
for 1 h at ꢃ788C, and then methyl 4-chlorobenzoate (5 g,
29.3 mmol) was added in one portion. The reaction was allowed to
warm slowly to RT and stirred for 16 h before quenching with satu-
rated aqueous NH4Cl (100 mL). The solution was further diluted
with H2O (100 mL) and EtOAc (200 mL). The resulting precipitate
was collected by filtration and washed with Et2O (100 mL) to give
(1.87 g, 83%) as a colourless solid. H NMR (500 MHz, CDCl3): d=
10.76 (1H, bs, NH), 9.27 (1H, dd, ArH, J=1.9 and 0.7 Hz), 7.77 (1H,
dd, ArH, J=9.4 and 1.9 Hz), 7.65 (1H, dd, ArH, J=9.4 and 0.7 Hz),
1.80 (2H, d, cyclohexyl-H, J=12.6 Hz), 1.74 (2H, dd, cyclohexyl-H,
J=9.6 and 3.1 Hz), 1.64 (1H, d, cyclohexyl-H, J=11.1 Hz), 1.38 (2H,
dd, cyclohexyl-H, J=12.2 and 2.7 Hz), 1.17–1.30 ppm (4H, m, cyclo-
hexyl-H); LC–MS [ES+]: m/z 395 and 397 (79Br [M+H] and 81Br
[M+H]+) tR =0.9–1.1 min.
General procedure A: Suzuki reaction on the triazolopyridine core
(18 to 19). Compound 4 (1 equiv), N,N-dimethylformamide (DMF;
4 mL(mmol 4)ꢃ1
,
saturated aqueous K2CO3 (4 mLmmolꢃ1),
1
12 as a white solid (3.39 g, 18.9 mmol, 64%). H NMR ([D6]DMSO,
Pd(PPh3)4 (1%), and boronic acid (2 equiv) were heated in a micro-
wave reactor for 10 min at 1408C, diluted with EtOAc, extracted
and washed with LiCl (5%, 1ꢁ10 mL), H2O (3ꢁ10 mL), and brine
(1ꢁ10 mL), dried over MgSO4, filtered, and the solvent was re-
moved in vacuo. Column chromatography was performed by elut-
ing with petroleum ether (40–60): ether afforded the desired aryl-
substituted triazolopyridines.
500 MHz): d=7.66 (1H, s, ArH), 7.34 (1H, s, ArH), 6.71 (2H, s, ArH),
3.35 ppm (2H, s, CH2); LC–MS (ES) [M+H], 100%.
5-(4-Chlorophenyl)-1H-pyrazol-3-amine
(13):
12
(15.9 g,
93.3 mmol) was dissolved in EtOH (250 mL), N2H4·H2O (5 mL,
102.6 mmol) was added, and the reaction was heated at reflux
with stirring for 24 h. The reaction was then cooled to RT and con-
centrated in vacuo to give a yellow oil. Trituration with CH2Cl2 gave
a white solid 13 (14.8 g, 76 mmol, 86% yield). 1H NMR (CDCl3,
500 MHz): d=9.15 (1H, bs, NH), 7.42–7.40 (2H, m, ArH), 7.34–7.32
(2H, m, ArH), 5.85 (1H, s, ArH), 3.76 ppm (2H, s, NH2); LC–MS (ES)
[M+H], 100%.
General procedure B: Amide coupling on the aryl-substituted triazo-
lopyridine core (20 to 19). Triazolopyridine 16 (1 equiv), DMAP
(10 mol%), and pyridine (anhydrous, 7 mLmmolꢃ1) was cooled to
08C under argon, and the acid chloride (1 equiv) was added drop-
wise. The mixture was warmed to RT and stirred for 17 h, concen-
trated in vacuo, and the crude mixture partitioned between Et2O
(20 mL) and H2O (10 mL), was extracted with Et2O (2ꢁ10 mL), and
the organic layer (suspension) separated from the aqueous layer;
this was passed directly through a filter funnel, and the white pre-
cipitate was collected to afford the desired product.
General method for the acylation of 3-aminopyrazoles
N-[5-(4-Chlorophenyl)-1H-pyrazol-3-yl]benzamide (8): 13 (97 mg,
0.5 mmol), DIPEA (90 mL, 0.6 mmol), and benzoyl chloride (64 mL,
0.55 mmol) were dissolved in dry 1,4-dioxane (2 mL) in a micro-
wave reaction tube, capped and heated at 1508C for 5 min in a mi-
crowave reactor. The reaction was then cooled to RT and PS isocya-
nate (0.25 g resin, 1.44 mmolgꢃ1 loading) was added. The reaction
tube was re-capped and heated at 1608C for 5 min in the micro-
wave reactor. The reaction was cooled, filtered to remove the resin,
which was washed with 1,4-dioxane (2 mL). The reaction mixture
was concentrated in vacuo, then partitioned between Et2O (3 mL)
and NaOH (3 mL of a 2m solution in H2O), and the mixture was
stirred at RT for 1 h. The resulting solid was collected by filtration
N-{6-(3-Methylphenyl)-[1,2,4]triazolo[1,5-a]pyridinyl}cyclohexa-
necarboxamide (25): Prepared according to general procedure A
on a 0.15 mmol scale to afford the product as a colourless solid
1
(36 mg, 72%). H NMR (500 MHz, CDCl3): d=8.77 (1H, s, ArH), 8.47
(1H, bs, NH), 7.78 (1H, dd, ArH, J=9.2 and 1.8 Hz), 7.65 (1H, d,
ArH, J=9.2 Hz), 7.36–7.41 (3H, m, ArH), 7.23 (1H, d, ArH, J=
7.4 Hz), 2.47 (3H, s, CH3), 2.05–2.02 (2H, m, cyclohexyl-H), 1.86–1.88
(2H, m, cyclohexyl-H), 1.70–1.75 (1H, m, cyclohexyl-H), 1.63–1.60
(2H, m, cyclohexyl-H), 1.27–1.31 ppm (4H, m, cyclohexyl-H); LC–MS
[ES+]: m/z 335 [M+H]+ tR =3.6–3.7 min.
ChemMedChem 2011, 6, 2214 – 2224
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2223

P. G. Wyatt et al.
MED
[4] M. Parsons, E. A. Worthey, P. N. Ward, J. C. Mottram, BMC Genomics
2005, 6, 127.
[5] P. Hassan, D. Fergusson, K. M. Grant, J. C. Mottram, Mol. Biochem. Parasi-
tol. 2001, 113, 189–198.
[6] C. Doerig, Biochim. Biophys. Acta Proteins Proteomics 2004, 1697, 155–
168.
[7] T. C. Hammarton, J. C Mottram, C. Doerig, Prog. Cell Cycle Res. 2003, 5,
91–101.
N-{7-(3,5-Dimethylphenyl)-[1,2,4]triazolo[1,5-a]pyridinyl}cyclo-
hexanecarboxamide (33): Prepared according to general proce-
dure A on a 0.46 mmol scale. Purification by column chromatogra-
phy eluting with Et2O afforded the product as a colourless solid
(34 mg, 21%). 1H NMR (500 MHz, DMSO): d=10.72 (1H, s, NH),
9.15 (1H, s, ArH), 7.97 (1H, dd, ArH, J=9.2 and 1.8 Hz), 7.74 (1H,
dd, ArH, J=9.2 and 0.7 Hz), 7.42 (2H, s, ArH), 7.07 (1H, s, ArH),
2.53–2.51 (1H, m, CH), 2.37 (6H, s, CH3), 1.84–1.75 (4H, m, CH),
1.68–1.65 (1H, m, CH), 1.45–1.39 (2H, m, CH), 1.32–1.19 ppm (3H,
m, CH); LC–MS [ES+]: m/z 349 [M+H]+ tR =4.8–5.0 min.
[8] X. Tu, C. C. Wang, J. Biol. Chem. 2004, 279, 20519–20528.
[9] T. C. Hammarton, J. Clark, F. Douglas, M. Boshart, J. C. Mottram, J. Biol.
Chem. 2003, 278, 22877–22886.
[10] M. Malumbres, P. Pevarello, M. Barbacid, J. R. Bischoff, Trends Pharmacol.
Sci. 2008, 29, 16–21.
[11] R. G. Walker, G. Thomson, K. Malone, M. W. Nowicki, E. Brown, D. G.
Blake, N. J. Turner, M. D. Walkinshaw, K. M. Grantm, J. C. Mottram, PloS
Neglected Trop. Dis. 2011, 5, 1–11.
[12] K. M. Grant, M. H. Dunion, V. Yardley, AL Skaltsounis, D. Marko, G. Eisen-
brand, S. L. Croft, L. Meijer, J. C. Mottram, Anti. Agents Chemotherapy
2004, 48, 3033–3042.
[13] R. Brenk, A. Schipani, D. James, A. Krasowski, I. H. Gilbert, J. Frearson,
P. G. Wyatt, ChemMedChem 2008, 3, 435–444.
[14] P. Pevarello, M. G. Brasca, R. Amici, P. Orsini, G. Traquandi, L. Corti, C.
Piutti, P. Sansonna, M. Villa, B. S. Pierce, M. Pulici, P. Giordano, K. Martina,
E. L. Fritzen, R. A. Nugent, E. Casale, A. Cameron, M. Ciomei, F. Roletto,
A. Isacchi, G. Fogliatto, E. Pesenti, W. Pastori, A. Marsiglio, K. L. Leach,
P. M. Clare, F. Fiorentini, M. Varasi, A. Vulpetti, M. A. Warpehoski, J. Med.
Chem. 2004, 47, 3367–3380.
[15] P. Pevarello, M. G. Brasca, P. Orsini, G. Traquandi, A. Longo, M. Nesi, F.
Orzi, C. Piutti, P. Sansonna, M. Varasi, A. Cameron, A. Vulpetti, F. Roletto,
R. Alzani, M. Ciomei, C. Albanese, W. Pastori, A. Marsiglio, E. Pesenti, F.
Fiorentini, R. Bischoff, C. Mercurio, J. Med. Chem. 2005, 48, 2944–2956.
[16] K. S. Joshi, M. J. Rathos, R. D. Joshi, M. Sivakumar, M. Mascarenhas, S.
Kamble, B. Lal, S Sharma, Mol. Cancer Ther. 2007, 6, 918–925.
[17] I. C. Choong, I. Serafimova, J. Fan, D. Stockett, E. Chan, S. Cheeti, Y. Lu,
B. Fahr, P. Pham, M. R. Arkin, D. H. Walker, U. Hoch, Bioorg. Med. Chem.
Lett. 2008, 18, 5763–5765.
[18] J. Sridhar, N. Akula, N. Pattabiraman, AAPS J. 2006, 8, E204–E221.
[19] C. Hamdouchi, H. Keyser, E. Collins, C. Jaramillo, J. E. De Diego, C. D.
Spencer, J. A. Dempsey, B. D. Anderson, T. Leggett, N. B. Stamm, R. M.
Schultz, S. A. Watkins, K. Cocke, S. Lemke, T. F. Burke, R. P. Beckmann,
J. T. Dixon, T. M. Gurganus, N. B. Rankl, K. A. Houck, F. M. Zhang, M.
Vieth, J. Espinosa, D. E. Timm, R. M. Campbell, B. K. R. Patel, H. B. Brooks,
Mol. Canc. Ther. 2004, 3, 1–9.
2-Cyclopentyl-N-(6-m-tolyl-[1,2,4]triazolo[1,5-a]pyridinyl)aceta-
mide (44): Prepared by following general procedure B on a
0.22 mmol scale to afford the desired product (20 mg, 27%).
1H NMR (500 MHz, CDCl3): d=8.78 (1H, bs, NH), 7.79 (1H, dd, ArH,
J=9.2 and 1.8 Hz), 7.66 (1H, dd, ArH, J=9.2 and 0.7 Hz), 7.42 (1H,
d, ArH, J=7.5 Hz), 7.37–7.40 (2H, m, ArH), 2.47 (3H, s, ArH), 2.37–
2.44 (1H, m, CH), 1.92–1.99 (2H, m, CH2), 1.63–1.72 (2H, m, cyclo-
pentyl-H), 1.59–1.62 (4H, m, cyclopentyl-H), 1.23–1.31 ppm (2H, m,
cyclopentyl-H); LC–MS [ES+]: m/z 335 [M+H]+ tR =3.7–3.8 min.
1-Cyclohexyl-3-(5,5-dimethyl-3-oxocyclohex-1-en-1-yl)urea (54):
Prepared by following general procedure A to afford 54 as a white
1
solid (47 mg, 25% yield). H NMR (DMSO, 500 MHz): d=8.32 (1H,
s, NH), 6.42 (1H, d, J=7.6 Hz, NH), 6.40 (1H, s, CH), 3.43–3.41 (1H,
m, CH), 2.23 (2H, s, CH2), 2.05 (2H, s, CH2), 1.78–1.75 (2H, m, CH),
1.65–1.63 (2H, m, CH), 1.54–1.51 (1H, m, CH), 1.33–1.26 (2H, m,
CH), 1.21–1.11 ppm (3H, m, CH) and 0.99 (6H, s, Me); LC–MS 97%
([M+H], 265).
1-(5,5-Dimethyl-3-oxocyclohex-1-en-1-yl)-3-(3-methoxypheny-
l)urea (56): Prepared by following general procedure A to afford
1
56 as a white solid (98 mg, 47% yield). H NMR (DMSO, 500 MHz):
d=8.92 (1H, s, NH), 8.65 (1H, s, NH), 7.22–7.12 (2H, m, ArH), 6.92–
6.6.91 (1H, m, ArH), 6.62–6.60 (1H, m, CH), 6.43 (1H, s, CH), 3.73
(3H, s, Me), 2.33 (2H, s, CH2), 2.11 (2H, s, CH2) and 1.02 ppm (6H, s,
Me); LC–MS 99% ([M+H], 289).
Acknowledgements
[20] J. Jie-Lou Liao, J. Med. Chem. 2007, 50, 409–424.
[21] M. Barvian, D. Boschelli, J. Cossrow, E. Dobrusin, A. Fattaey, A. Fritsch, D.
Fry, P. Harvey, P. Keller, M. Garrett, F. La, W. Leopold, D. McNamara, M.
Quin, S. Trumpp-Kallmeyer, P. Toogood, Z. Wu, E. Zhang, J. Med. Chem.
2000, 43, 4606–4616.
[22] MRC Protein Phosphorylation Unit, National Centre for Protein Kinase
Profiling, University of Dundee (UK): http://www.kinase-screen.mrc.a-
c.uk/kinase-panel.htm.
We thank the Wellcome Trust for financial support for these stud-
ies (grant ref. 077705 and strategic award WT083481) and Irene
Hallyburton (DDU, University of Dundee) for carrying out the
T. brucei and MRC5 proliferation studies. The human CDK2–CYCA
complex was obtained from the Division of Signal Transduction
Therapy in the College of Life Sciences (University of Dundee).
[23] C. Notredame, D. G. Higgins, J. Heringa, J. Mol. Biol. 2000, 302, 205–
217.
[24] A. Sali, T. L. Blundell, J. Mol. Biol. 1993, 234, 779–815.
[25] M. Hendlich, Acta Crystallogr. Sect. D Biol. Crystallogr. 1998, 54, 1178–
1182.
Keywords: CRK3 · cyclin-dependent cdc2-related kinases ·
leishmaniasis · triazolopyridines · ureas
[26] P. R. Gerber, J. Comput.-Aided Mol. Des. 1998, 12, 37–51.
[27] M. Nettekoven, B. Puellmann, S. Schmitt, Synthesis 2003, 1649–1652.
[1] K. Stuart, R. Brun, S. L. Croft, A. H. Fairlamb, R. E. Gurtler, J. McKerrow, S.
Reed, R. Tarleton, J. Clin. Invest. 2008, 118, 1301–1310.
[2] R. Reithinger, J. C. Dujardin, H. Louzir, C. Pirmez, B. Alexander, S. Brooker,
Lancet Inf. Dis. 2007, 7, 586–596.
Received: July 13, 2011
Revised: August 16, 2011
[3] S. K. Grant, Cell. Mol. Life Sci. 2009, 66, 1163–1177.
Published online on September 13, 2011
2224
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 2214 – 2224