Design of potent and selective hybrid inhibitors of the mitotic kinase nek2: Structure-activity relationship, structural biology, and cellular activity

Article abstract of DOI:10.1021/jm201683b

We report herein a series of Nek2 inhibitors based on an aminopyridine scaffold. These compounds have been designed by combining key elements of two previously discovered chemical series. Structure based design led to aminopyridine (R)-21, a potent and selective inhibitor able to modulate Nek2 activity in cells.

Full text of DOI:10.1021/jm201683b

Article
pubs.acs.org/jmc
Design of Potent and Selective Hybrid Inhibitors of the Mitotic
Kinase Nek2: Structure−Activity Relationship, Structural Biology, and
Cellular Activity†
Paolo Innocenti,‡ Kwai-Ming J. Cheung,‡ Savade Solanki,‡ Corine Mas-Droux,§ Fiona Rowan,‡,§
Sharon Yeoh,∥ Kathy Boxall,‡ Maura Westlake,‡ Lisa Pickard,‡ Tara Hardy,∥ Joanne E. Baxter,∥
G. Wynne Aherne,‡ Richard Bayliss,§,∥ Andrew M. Fry,∥ and Swen Hoelder*,‡
‡Cancer Research UK Cancer Therapeutics Unit, Division of Cancer Therapeutics, The Institute of Cancer Research, 15 Cotswold
Road, Sutton, Surrey SM2 5NG, United Kingdom
§Chester Beatty Laboratories, Division of Structural Biology, The Institute of Cancer Research, 237 Fulham Road, London SW3 6JB,
United Kingdom
∥Department of Biochemistry, Henry Wellcome Building, University of Leicester, Lancaster Road, Leicester, LE1 9HN, United
Kingdom
S
* Supporting Information
ABSTRACT: We report herein a series of Nek2 inhibitors based on an aminopyridine scaffold. These compounds have been
designed by combining key elements of two previously discovered chemical series. Structure based design led to aminopyridine
(R)-21, a potent and selective inhibitor able to modulate Nek2 activity in cells.
propynamide 1 has been shown to be a nanomolar range
inhibitor acting through an irreversible mechanism (Figure 1).3a
We previously described a pyrazine series of Nek2 inhibitors
(exemplified by 2, Figure 1) that attained an IC50 of 0.23 μM,
displayed high selectivity, but failed to achieve cellular activity
because of poor permeability.5 In addition, we published results
of a series of benzimidazole compounds that showed
comparable activity toward Nek2 and good permeability (see
structure (R)-3, Figure 1).6
The structure−activity relationship (SAR) of this series was
found to be nonlinear: high Nek2 inhibition was achieved when
the basic piperidine and the fully substituted benzyl ether were
present in the same molecule. Removal of either group resulted
in a significant drop in potency. Despite these interesting
features, these compounds also did not show any cellular
activity, most likely due to lack of sufficient potency. Herein we
report the exploration of hybrids of the two series that
ultimately led to potent compounds with the ability to
modulate the phosphorylation status of a Nek2 substrate in
cells.
INTRODUCTION
■
(Never in mitosis gene a)-related kinase 2 (Nek2) is a serine/
threonine kinase involved in key mitotic processes. It localizes
to the centrosomes and takes part in the regulation of spindle
pole organization and separation by phosphorylation of
centrosomal Nek2-associated protein 1 (C-Nap1), rootletin,
and ninein-like protein (Nlp).1 Nek2 has also been reported to
act as a regulator of the spindle assembly checkpoint (SAC)
through interaction or phosphorylation of highly expressed
cancer 1 (Hec1), mitotic arrest deficient-like 1 (Mad1), and
mitotic arrest deficient-like 2 (Mad2).2 In recent publications,
the role of Nek2 was probed by RNAi knockdown experiments
and through inhibition by a covalent inhibitor. These studies
suggest that Nek2 is not essential for centrosome separation
but plays a supportive part.3 Moreover, definitive evidence for a
role in the SAC is still lacking. Despite recent and somewhat
conflicting literature reports on its requirement for mitotic
progression, Nek2 is an interesting cancer target. High levels of
Nek2 expression have been found in many tumors, and RNAi
knockdown experiments have shown antiproliferative effects in
HeLa, MDA-MB-231, HuCCT1, and MCF7 cells in vitro and
in vivo.4
The paucity of potent, selective, and cell-active inhibitors has
hampered in-depth investigation of the role of Nek2. Recently,
Received: December 15, 2011
Published: March 12, 2012
© 2012 American Chemical Society
3228
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
Figure 1. Nek2 inhibitors described in the literature.
a
Scheme 1. Synthesis of Aminopyrazines
a
Reagents and conditions: (a) H2SO4, MeOH, 80 °C; (b) ( )-1-(2-(trifluoromethyl)phenyl)ethanol, di-tert-butyl azodicarboxylate, PPh3, DCM, 0
°C to rt; (c) bis(pinacolato)diboron, NaOAc, Pd(dppf)Cl2·DCM, DMF, microwave, 100 °C; (d) 3-bromo-5-chloropyrazin-2-amine, NaHCO3,
Pd(dppf)Cl2·DCM, DMF/water, microwave, 100 °C; (e) RB(OR)2 or RB(OH)2, Na2CO3, Pd(PPh3)4, DMF/water, microwave, 125−135 °C; (f)
NH3, MeOH, 75−90 °C.
The aminopyridine series was made by a synthetic sequence
similar to that used for the aminopyrazine series. Ethers 7 and
12a were prepared starting from phenol 6 under Mitsunobu
conditions with ( )-, (S)-1-(2-(trifluoromethyl)phenyl)-
ethanol, or 4-methoxybenzyl alcohol (Scheme 2). Ether 12b
was made from compound 6 by treatment with methyl iodide
and potassium carbonate. Boronic esters 8 and 13 were formed
using standard palladium chemistry8 and subsequently coupled
with 5-bromo-3-iodopyridin-2-amine under Suzuki conditions
to afford bromopyridines 14.10
Aminopyridines 16a−i were formed by Suzuki coupling of
the appropriate boronic esters or acids with bromo derivatives
14 (Scheme 2).10,12 Derivatives 16j−l were prepared starting
from 1-(5-bromothiophen-3-yl)-N,N-dimethylmethanamine13
and suitably prepared boronic acids 15.10 Ammonolysis of
esters 16 furnished the target primary amides 17. Compound
rac-21 was prepared via a slightly different route: amino-
CHEMISTRY
■
The pyrazine compounds were prepared starting from
commercially available methyl 4-bromo-2-hydroxybenzoic acid
5, which was transformed into methyl ester 6 by treatment with
sulfuric acid in methanol (Scheme 1). Introduction of the
benzyl ether functionality to afford compound rac-7 was carried
out under Mitsunobu conditions7 in the presence of ( )-1-(2-
(trifluoromethyl)phenyl)ethanol, whereas transformation into
the corresponding boronic acid ester rac-8 was accomplished
by means of a standard palladium catalyzed reaction.8 Coupling
with known 3-bromo-5-chloropyrazin-2-amine9 afforded chloro
derivative 9, which was further elaborated through Suzuki
reactions to produce esters 10 (Scheme 1).10 The boronic acids
or esters required for the couplings described above were either
commercially available or prepared according to standard
procedures.11 Finally, esters 10 were heated with ammonia in
methanol to give primary amides 11 (Scheme 1).
3229
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of Aminopyridines
a
Reagents and conditions: (a) ( )-, (S)-1-(2-(trifluoromethyl)phenyl)ethanol or 4-methoxybenzyl alcohol, di-tert-butyl azodicarboxylate, PPh3,
DCM, 0 °C to rt (rac-7, (R)-7, and 12a); (b) MeI, K2CO3, DMF, rt (12b); (c) bis(pinacolato)diboron, NaOAc, Pd(dppf)Cl2·DCM, DMF,
microwave, 100 °C; (d) 5-bromo-3-iodopyridin-2-amine, NaHCO3, Pd(dppf)Cl2·DCM, DMF/water, microwave, 100 °C; (e) RB(OR)2 or
RB(OH)2, Na2CO3, Pd(PPh3)4, DMF/water or DME/water, microwave, 110−135 °C (16a−i); (f) 1-(5-bromothiophen-3-yl)-N,N-
dimethylmethanamine, Na2CO3, Pd(PPh3)4, DME/water, microwave, 110 °C (16j−l); (g) NH3, MeOH, 70−80 °C; (h) TFA, DCM, 0 °C; (i)
( )-5,5,5-trifluoropent-3-yn-2-ol, di-tert-butyl azodicarboxylate, PPh3, DCM, 0 °C to rt; (j) H2, Lindlar’s catalyst, MeOH, rt.
pyridine 16l was deprotected with trifluoroacetic acid to afford
phenol 18, which was derivatized with ( )-5,5,5-trifluoropent-
3-yn-2-ol14 under Mitsunobu conditions to give ether 19.
Target amide rac-21 was obtained after hydrogenation with
Lindlar catalyst to afford ester 20 followed by ammonolysis
(Scheme 2).
candidate for the benzimidazole moiety was the amino-
pyrazine-type scaffold, which in our studies consistently showed
better LE compared to the benzimidazoles. Indeed, super-
position of cocrystal structures of both series indicated that
hypothetical hybrids of these series can address all the
pharmacophoric features of benzimidazole (R)-3 (Figure 2A).
This prompted us to prepare a series of compounds exploring
this design principle and comprising the following key features:
the presence of a suitably placed amino-substituted aromatic
group, an aminopyrazine-type binding scaffold, and a phenyl
carboxyamide substituted with a benzyl ether (Figure 1).
SAR of Benzyl Ether-Substituted Hybrid Compounds.
The initial series of compounds were built by fusing the hinge-
binding fragment of aminopyrazine 2 with the benzyloxy-
substituted benzamide moiety of benzimidazole (R)-3 (Figure
1). In order to place a suitable basic residue mimicking the
piperidine ring of compound (R)-3, a p-N,N-dimethylamino-
methylphenyl group was used instead of the trimethoxyphenyl
substituent. The resulting hybrid 11a showed a Nek2 IC50 of
RESULTS AND DISCUSSION
■
Hybrid Compounds Design. As mentioned above, our
previous studies of the benzimidazole series led to permeable
and selective inhibitors that suffered from insufficient potency
and high lipophilicity (compound (R)-3, Figure 1).6 Our results
indicated that the poor ligand efficiency (LE) and modest
potency of these compounds were at least in part due to
inefficient binding of the phenyl substituted benzimidazole core
4 (IC50 = 62 μM, LE = 0.32, Figure 1). We speculated that a
more potent compound could be obtained if the substituents of
benzimidazole 3 were to be grafted on a core that binds more
efficiently to the hinge region of Nek2. A replacement
3230
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
Figure 2. Structures of aminopyridines rac-17j and rac-21 bound to Nek2 as determined by X-ray crystallography. (A) The rational basis of the
hybrid design is illustrated by the superposition of the structure of Nek2 (purple) bound to rac-3 (carbons colored yellow), extracted from PDB
code 2XNP, on the structure of Nek2 (not shown) bound to an aminopyrazine inhibitor (carbons colored green), PDB code 2XKF. (B) View of the
ATP binding site of Nek2 (carbons colored teal) occupied by aminopyridine rac-17j (carbons colored orange). Wire-mesh represents the 2mFo −
DFc electron density map in the vicinity of the compound, contoured at 1.0σ. Black dashed lines represent hydrogen bonds formed between the
protein and the compounds, magenta dashed lines represent the orthogonal interaction between Ile14 and the trifluoromethyl group. (C) Structure
viewed in a second orientation, showing the hydrophobic contacts between the compound and protein (double-headed arrows) and the hydrogen
bonds between the amide group of the compound and Asp159 of the Nek2 DFG motif (black dashed lines). (D) Superposition of structures of
compound rac-17j bound to Nek2 (teal) and compound rac-3 (PDB code 2XNP, carbons colored yellow) bound to Nek2 (not shown), viewed
from above. (E) Structure of rac-21 (carbons colored magenta) bound to Nek2 (light gray). An ordered water molecule in the active site is shown as
a red sphere. Wire-mesh represents the 2mFo − DFc electron density map in the vicinity of the compound and the water molecule, contoured at 1.0σ.
(F) Superposition of the structures of Nek2 (light gray) bound to rac-21 with the structure of Nek2 (not shown) bound to rac-17j.
0.79 μM (Table 1) and proved to be approximately equipotent
to the parent compounds (0.23 and 0.36 μM, respectively, for
pyrazine 2 and benzimidazole (R)-3, Figure 1). Moving the
amino group attachment point to the meta position resulted in
aminopyrazine 11b which demonstrated reduced Nek2
inhibition (IC50 of 2.47 μM, Table 1). During the course of
previous studies on the aminopyrazine series, it was found that
switching to an aminopyridine hinge-binding scaffold caused a
marked increase in Nek2 inhibition.15,12 For this reason,
aminopyridines 17a and 17b were prepared, featuring a para-
and a meta-substituted phenyl ring moiety, respectively.
Gratifyingly, both compounds showed improved activity against
Nek2 with IC50 values of 0.12 and 0.21 μM, respectively (Table
1). Overall, these hybrid compounds showed moderate activity
and LE as well as high predicted lipophilicity (Table 1).
aminopyridines 17d and 17e are 0.047 and 0.058 μM,
respectively (Table 3).
Next, we prepared two additional compounds bearing
alterations of the basic residue: compound 17f features a
cyclic, more rigid amino group, while compound 17g features a
less basic morpholine ring. The former did show only a slight
decrease in Nek2 inhibition (IC50 for 17f is 0.093 μM, Table 3),
suggesting that the shape of the aminomethyl substituent is not
crucial for activity. Conversely, the reduced basicity of
aminopyridine 17g had a more pronounced effect, with the
IC50 increasing more than 7-fold to 0.318 μM (compare
compound 17c and 17g, Tables 2 and 3). These observations
are in line with previous findings on the role of amino
substituents in benzimidazole-based Nek2 inhibitors and stress
the importance of a suitably placed basic residue with the ability
to engage in charge−charge type interactions.6
In order to improve the potency of this series against Nek2,
the aminomethyl-substituted phenyl ring was replaced by an
isosteric thiophene. Whereas in a previous series this
modification had a generally positive effect on Nek2 activity,5
these results showed that the outcome strongly depends on
hinge-binding motif (Table 2). Thus, introduction of an α′- or
β′-aminomethyl substituted α-thiophene on the aminopyrazine
scaffold resulted in a drop in Nek2 inhibition, with compounds
11c and 11d showing IC50 values of 2.68 and 2.63 μM,
respectively (Table 2; compare with aminopyrazines 11a and
11b, Table 1). Conversely, when the same change was applied
on the aminopyridine core, a significant improvement in
activity toward Nek2 was observed: the IC50 values for
compounds 17c and rac-17j, featuring an α′- or β′-aminomethyl
substituent, are 0.047 and 0.059 μM, respectively (Table 2).
Further decoration of the thiophene group of aminopyridine
17c with methyl substituents at either the β′ or β position
resulted in virtually equipotent compounds: the IC50 values for
Since this series was initially inspired by Plk1 inhibitors16 and
still maintained the original substituted benzyl ether moiety, we
investigated the Plk1 inhibition for selected compounds. While
all showed a preference for Nek2, the window was relatively
small: Plk1 IC50 values were 3.90, 1.56, and 1.32 μM,
respectively, for amides 11a, 17a, and rac-17j, which translated
to 10- to 25-fold reduced activity when compared to Nek2
(Tables 1 and 2). Since we were interested in developing tool
compounds and Plk1 has been suggested to play a role
upstream of Nek2,3b an improved selectivity window was
desirable. In addition, we were still concerned about the high
lipophilicity of the described compounds.
Cocrystal Structure of Aminopyrydine rac-17j Bound
to Nek2. In order to address these issues, we solved the
cocrystal structure of rac-17j bound to Nek2. As expected from
our design hypothesis leading to this hybrid series, the binding
mode shares many features with that of rac-3, including a DFG-
out conformation of the kinase.6 Here, we will focus on key
3231
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
a
Table 1. Phenyl Substituted Hybrids
a
Results are the mean SD for n ≥ 3 or the mean of two independent determinations with individual determinations in parentheses or samples run
n = 1. Compounds are racemic unless otherwise stated.
elements. The aminopyridine forms two hydrogen bonds with
the hinge region at Glu87 and Cys89 (Figure 2B and Figure
2C), compared to the single hydrogen bond formed between
the imidazole ring of rac-3 and Cys89. The phenyl ring of the
substituted benzamide is sandwiched between the gatekeeper
Met86 and Phe148, while the amide group engages in two
hydrogen bonds with the backbone NH and the side chain
carbonyl group of Asp159 of the DFG motif. As observed for
rac-3, the substituted benzyl group is well ordered with the
trifluoromethyl group binding to a hydrophobic pocket formed
by part of the glycine rich loop, and one of the fluorine atoms is
engaged in an orthogonal interaction with the carbonyl group
of Ile14 (3.6 Å). The position of the methyl group clearly
reveals the R configuration at the benzylic stereocenter of rac-
17j even though the racemic mixture was used for the soaking
process. This confirmed that the R isomer is the most potent
enantiomer for Nek2 binding as observed for the benzimidazole
series.6 The thiophene ring engages in hydrophobic contacts
with Ile14 and Gly92. The dimethylamino group is clearly
resolved and superimposes well with the basic center of the
piperidine ring of rac-3 (Figure 2D). We have shown that the
correct positioning of the piperidine ring of rac-3 is crucial for
high selectivity of this compound against Plk1 most likely by
causing a clash with the Arg136 residue in Plk1.6 We
hypothesize that the N,N-dimethylamino group of rac-17j
plays a similar role, leading to the 25-fold selectivity with
respect to Plk1 inhibition. The observation that rac-17j shows a
somewhat smaller selectivity window might be explained by the
smaller size of the dimethylamino group compared to the
piperidine group, leading to reduced steric interaction with
Arg136 in Plk1.
The binding mode revealed by the crystal structure thus
confirmed our design hypothesis. However, we were keen to
further improve the selectivity against Plk1 and to reduce the
lipophilicity of our compounds. Our attention turned to the
role of the benzyl ether moiety. Close examination of the X-ray
cocrystal structures of rac-17j and rac-3 revealed that this
aromatic region is solvent exposed and has few contacts with
Nek2 (Figure 2B and Figure 2D). Its main role appears to be as
a scaffold for the trifluoromethyl group that binds to the small
hydrophobic groove formed partially by the Gly rich region. As
we have previously shown, the benzylic methyl group is
3232
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
a
Table 2. Thiophene Substituted Hybrids
a
Results are the mean SD for n ≥ 3 or the mean of two independent determinations with individual determinations in parentheses or samples run
n = 1. Compounds are racemic unless otherwise stated.
important for activity, since it stabilizes the bioactive
conformation in which the fluorinated moiety points toward
the groove through minimization of the 1,3-allylic strain.6 We
hypothesized that the phenyl group is not needed for potent
Nek2 inhibition and can be replaced by a smaller moiety that
leads to comparable positioning of the trifluoromethyl group.
Replacement of the aromatic residue would have the benefit of
significantly reducing the lipophilicity and molecular weight. In
addition, we speculated that this change might be significantly
less tolerated for Plk1 and thus lead to an improved selectivity
window.
Synthesis and Profiling of Truncated Inhibitors. To
explore this design rationale, we prepared a compound in which
the benzyl ether moiety was replaced by an allylic ether bearing
the desired Z configuration. Pleasingly, aminopyridine rac-21
was roughly equipotent compared to the parent compound: the
Nek2 IC50 was found to be 0.073 μM, and the LE value was
0.29 with a CLogP reduced to 4.37 (Table 4). These values
compared well with those associated with benzyl ether rac-17j
and demonstrate that replacing the aromatic ring with a smaller
allyl ether is tolerated and afforded a compound with better LE
and lower CLogP (aminopyridine rac-21, Table 4). The crystal
structure of of rac-21 bound to Nek2 confirmed that the
binding mode was as expected (Figure 2E and Figure 2F). The
reduced bulk of the allylic ether compared to the benzyl ether
allowed the binding of an ordered water molecule within the
active site, although this apparently had little impact on the
potency of inhibition. Our previous studies on benzimidazole-
based Nek2 inhibitors showed that the R configured
enantiomer is significantly more potent than the S enantiomer.6
We therefore separated the two enantiomers of aminopyridine
rac-21 through chiral HPLC.17 The faster eluting enantiomer
showed a Nek2 IC50 of 1.80 μM, more than 70 times weaker
than the slower eluting enantiomer, which had an IC50 of 0.022
μM. Since we have previously shown that the R configuration is
preferred for Nek2 inhibition, we assigned the R configuration
to the slower eluting enantiomer, whereas the faster eluting
compound was associated with the S configuration (Table 4).
It is worth noting that the high activity against Nek2 is
accompanied by substantial improvement of lipophilicity and
LE (compare compounds rac-17j, rac-21, and (R)-21, Table
3233
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
a
Table 3. SAR around the Thiophene Core
a
Results are the mean SD for n ≥ 3 or the mean of two independent determinations with individual determinations in parentheses or samples run
n = 1. Compounds are racemic unless otherwise stated.
4). To our knowledge, compound (R)-21 is the most potent
reversible Nek2 inhibitor reported to date.
21 fulfills our requirements in terms of both potency and
selectivity (IC50 of 0.022 μM, Gini coefficient18 of 0.514 (1
μM), Tables 5 and 6).
Kinase Selectivity. An assessment of the selectivity profile
for key compounds was then carried out on four kinases
involved in the mitotic machinery: Plk1, MPS1, AurA, and
CDK2 (Table 5). For comparison purposes, aminopyridine 17j
was prepared as a single enantiomer in the R configuration
(compound (R)-17j, Table 5). In general, all the compounds
showed a good level of selectivity. However, as anticipated, the
truncated compound (R)-21 showed significantly improved
selectivity against Plk1 (compare aminopyridines 17a and (R)-
17j and (R)-21, Table 5). Selectivity data on MPS1 inhibition
showed a similar trend, with the alkene compounds showing
higher IC50 values (Table 5). Finally, all the compounds
demonstrated low activity toward Aurora A and CDK2 with the
exception of the weak Nek2 inhibitor (S)-21, which inhibited
CDK2 with an IC50 of 0.53 μM (Table 5). Further profiling of
compound (R)-21 was performed against a panel of 24 kinases
using a ProfilerPro 1 kit. Again, at least a 100-fold window of
selectivity was observed in most cases even though a few cell
cycle unrelated kinases showed a more pronounced inhibition
(Table 6). Despite this, we concluded that aminopyridine (R)-
Cellular Profiling of rac-3, (R)-17j, (R)-21, and (S)-21.
Next, we evaluated the ability of aminopyridines (R)-17j, (R)-
21, and (S)-21 to inhibit Nek2 in cells by analyzing the effect of
these compounds on phosphoryation of C-Nap1. As described
in the Introduction, the centrosomal protein C-Nap1 has been
shown to be a substrate of Nek2, and S-2179 was identified by
mass spectrometry as a site that was phosphorylated by Nek2 in
vitro. We therefore raised rabbit polyclonal antibodies against
this phosophorylation site that could act as a direct marker of
Nek2 activity in cells. The specificity of this pC-Nap1 antibody
was confirmed using S2179A mutants and by RNAi depletion
of either C-Nap1 or Nek2 in cells.19 Briefly, osteosarcoma
U2OS cells were synchronized using aphidicolin for 16−18 h,
as Nek2 activity is elevated in S and G2 phases, and shortly
after release treated with compounds at different concen-
trations. Phosphorylation of C-Nap1 was determined by
measuring the intensity of centrosome staining following
immunofluorescence microscopy with the pC-Nap1 antibody.
Centrosomes, detected with γ-tubulin antibodies, were scored
3234
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
a
Table 4. Alkene Series
a
Results are the mean SD for n ≥ 3 or the mean of two independent determinations with individual determinations in parentheses or samples run
n = 1. Compounds are racemic unless otherwise stated.
as positive for pC-Nap1 staining on an IN Cell Analyzer 1000 if
the fluorescence intensity was above a defined threshold. The
IN Cell Investigator Developer Toolbox (GE Healthcare) was
used to analyze the data. All compounds were tested at least
twice.20
however, other factors, such as inhibition of upstream kinases,
might also play a role. The flat concentration dependency could
also be explained by the hypothesis that the weak effect of
aminopyridine (S)-21 in this assay is due to modulation of
different targets, very likely with different IC50 values. Overall,
these experiments showed that (R)-17j and particularly (R)-21
inhibit the phosphorylation of C-Nap1, a cellular substrate of
Nek2, in a concentration dependent manner. Next, we tested
the effect of benzimidazole rac-3 and of aminopyridines (R)-
17j, (S)-21, and (R)-21 on the growth of a panel of cancer cell
lines. We determined the concentration required to inhibit the
growth by 50% (GI50) after 96 h. However, we did not observe
any clear trends between biochemical and C-Nap1 assay on one
hand and antiproliferative effect on the other hand (Table 7).
These results suggest, at least in the case of the weak Nek2
inhibitor (S)-21, that inhibition of other targets is responsible
for the toxicity against the cancer cell lines after long-term
exposure. Similar off-target inhibition might also contribute to
the antiproliferative effects of aminopyridines (R)-17j and (R)-
21, albeit to a lesser degree.
Representative results are shown in Figure 3A. Pleasingly,
aminopyridine (R)-21 indeed showed concentration-dependent
down-regulation of C-Nap1 phosphoryation with an IC50 of 0.8
μM (Figure 3B). The approximately 40-fold drop of the IC50
values compared to the biochemical assays can be attributed to
the higher ATP concentrations in cells and is not uncommon
for ATP competitive kinase inhibitors. The corresponding
benzyl ether compound (R)-17j also showed an effect in this
assay albeit at a slightly higher concentration leading to an IC50
of 2.7 μM (Figure 3). The somewhat reduced activity might be
due to the more lipophilic character of (R)-17j, causing binding
to other components and so reducing the effective concen-
tration. It is worth noting that benzimidazole rac-3 did not
show any significant modulation of C-Nap1 phosphorylation in
this assay even at high concentrations (Figure 3A). Finally we
tested aminopyridine (S)-21, a much weaker biochemical
inhibitor of Nek2, in this assay. Indeed, this compound showed
significantly less suppression of C-Nap1 phosphorylation and a
flatter concentration dependency (Figure 3). We attributed the
weak effect of (S)-21 in this assay to residual Nek2 activity;
The data collected in the C-Nap1 phosphorylation assay
combined with the profiling against other prominent cell cycle
kinases suggest that aminopyridine (R)-21 can serve as a
valuable tool for investigating the role of Nek2 in the cell cycle
and its function in mitotic organization. It will be particularly
3235
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
a
Table 5. Selectivity Profile
a
Results are the mean SD for n ≥ 3 or the mean of two independent determinations with individual determinations in parentheses or samples run
n = 1. Compounds are racemic unless otherwise stated.
useful in short-term experiments using synchronized cell
populations aiming to identify additional substrates and
functions of Nek2.
Lastly, racemic aminopyridines rac-17j and rac-21 showed
medium to high permeability at physiological pH in a parallel
artificial membrane assay (PAMPA, 15.2 × 10−6 and 27.6 ×
10−6 cm/s, respectively) as well as high metabolic degrada-
tion.21
(compare core compound 17i with aminopyridine 17h or 17k,
IC50 values of 10.5, 9.0, and 2.26 μM, respectively, Table 8).
Instead, a large jump in activity (∼200-fold) takes place when
both substituents are present in rac-17j, as previously observed
in the benzimidazole series. These results confirm that the
nonlinear SAR observed for benzimidazoles translates into the
aminopyridine series.
CONCLUSIONS
Confirmation of Nonlinear SAR. Finally, we were
interested in investigating whether the nonadditive SAR
observed in the benzimidazole series is also translated in the
aminopyridine series. As described above, we previously found
that in order to attain a high level of Nek2 inhibition, the two
key pharmacophoric elements (namely, the amino group and
the substituted benzylic ether; see Figure 1) had to be present
at the same time.6 Thus, three truncated analogues of
aminopyridine rac-17j were prepared, lacking the amino
group, the benzylic ether, or both groups at the same time.
The results show that loss of either of these moieties leads to a
sharp decrease in activity to a level observed for the core
compound 17i (Table 8). This confirms that neither of the two
substituents significantly improves Nek2 activity on its own
■
Starting from two different sets of Nek2 ligands, we designed a
series of hybrid inhibitors. Using structure-based design, we
optimized Nek2 potency, kinase selectivity, lipophilicity, and
ligand efficiency to the desired range, ultimately culminating in
aminopyridine (R)-21. To our knowledge (R)-21 is the most
potent reversible inhibitor of Nek2 known to date. In our
cellular assay, compound (R)-21 showed a concentration
dependent inhibition of phosphorylation of C-Nap1, a known
substrate of Nek2. Gratifyingly, aminopyridine (R)-21 also
showed sufficient selectivity against the most relevant cell cycle
kinases and will therefore make for a useful tool compound to
probe the role of Nek2 in cell cycle control. The results of these
experiments will be reported in due course.
3236
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
appropriate internal nondeuterated solvent peak. The data are given as
follows: chemical shift (δ) in ppm, multiplicity (where applicable),
coupling constants (J) in Hz (where applicable), and integration
(where applicable). LCMS analyses were performed on a Micromass
LCT/Waters Alliance 2795 separations module HPLC system with a
Merck Chromolith SpeedROD RP-18e 50 mm × 4.6 mm column at
22 °C. The following solvent system, at a flow rate of 2 mL/min, was
used: solvent A, methanol; solvent B, 0.1% formic acid in water.
Gradient elution was as follows: 1:9 (A/B) to 9:1 (A/B) over 2.25
min, 9:1 (A/B) for 0.75 min, then reversion back to 1:9 (A/B) over
0.3 min, 1:9 (A:B) for 0.2 min. Detection was carried out with a
Waters 2487 dual wavelength absorbance detector (detecting at 254
nm), and ionization was electrospray (ESI). Some LCMS and all
HRMS analyses were performed on a Agilent 1200 series HPLC
system with a Merck Chromolith SpeedROD RP-18e 50 mm × 4.6
mm column at 22 °C. The following solvent system, at a flow rate of 2
mL/min, was used: solvent A, methanol; solvent B, 0.1% formic acid in
water. Gradient elution was as follows: 1:9 (A/B) to 9:1 (A/B) over
2.5 min, 9:1 (A/B) for 1 min, then reversion back to 1:9 (A/B) over
0.3 min, 1:9 (A/B) for 0.2 min. This was connected to a Agilent 6200
time of flight (ToF) mass spectrometer (simultaneous ESI and APCI
or ESI only) with detection at 254 nm. The following reference masses
were used for HRMS analysis: caffeine [M + H]+ = 195.087 652,
Table 6. Selectivity Data for (R)-21 against a Panel of
Kinases
a
kinase
% inhibition at 1 μM
IC50 (μM)
MAPKAPK2
AurA
6
19
12
29
7
PKCz
RSK1
PRAK
Erk1
2
PKD2
CH1d
CHK1
ABL
77
67
11
46
57
43
32
46
89
51
93
11
12
0
0.33
0.43
1.11
0.75
1.36
2.00
1.17
0.11
0.84
0.07
FYN
LYN
CHK2
MET
LCK
SRC
GSK3b
Erk2
reserpine [M
+ = 609.280 657 and (1H,1H,3H-
H]+
tetrafluoropentoxy)phosphazene [M + H]+ = 922.009 798. The purity
of final compounds was determined by HPLC as described above and
is ≥95% unless specified otherwise.
PKA
AKT2
INSR
p38a
0
Methyl 4-Bromo-2-hydroxybenzoate 6. A solution of 4-bromo-
2-hydroxybenzoic acid 5 (5.00 g, 23.04 mmol) in MeOH (20 mL) was
treated with sulfuric acid (1.7 mL). The mixture was refluxed for 24 h,
poured onto ice−water, and extracted with DCM. The combined
organics were washed with saturated aqueous NaHCO3, dried
0
AKT1
MSK1
3
25
a
1 μM (R)-21, 30 min preincubation of compound + enzyme, assay
run at apparent ATP Km for each enzyme.
1
(Na2SO4), and concentrated to give ester 6 (3.97 g, 75%). H NMR
(500 MHz, CDCl3) δ 10.82 (s, 1H), 7.70 (d, J = 8.5 Hz, 1H), 7.20 (d,
J = 1.9 Hz, 1H), 7.04 (dd, J = 8.5, 1.9 Hz, 1H), 3.97 (s, 3H).
Methyl 4-Bromo-2-((4-methoxybenzyl)oxy)benzoate 12a. A
solution of phenol 6 (1.70 g, 7.36 mmol), 4-methoxybenzyl alcohol
(1.29 g, 9.30 mmol), and triphenylphosphine (2.81 g, 10.71 mmol) in
DCM (35 mL) was cooled at 0 °C and treated with di-tert-butyl
azodicarboxylate (2.45 g, 10.65 mmol). The mixture was allowed to
reach room temperature and stirred overnight. The mixture was
diluted with DCM and quenched with water. The organic layer was
separated and extracted with DCM. The combined organics were
dried (Na2SO4), concentrated, and purified by Biotage column
EXPERIMENTAL SECTION
■
General Chemistry Information. Starting materials, reagents,
and solvents for reactions were reagent grade and used as purchased.
Chromatography solvents were HPLC grade and were used without
further purification. Thin layer chromatography (TLC) analysis was
performed using Merck silica gel 60 F-254 thin layer plates. Flash
column chromatography was carried out using columns prepacked
with 40−63 μm silica. NMR spectra were recorded on a Bruker
Avance 500 MHz spectrometer, and samples were referenced to the
Figure 3. C-Nap1 phosphorylation inhibition activity for rac-3, (R)-17j, (S)-21, and (R)-21. (A) C-Nap1 phosphorylation experiments (results are
expressed as the percent of cells with pC-Nap1 spot(s) as a percent of DMSO (1%) control). Error bars represent the standard deviation of six
replicates obtained in two separate experiments. (B) IC50 values for C-Nap1 phosphorylation inhibition calculated for compounds achieving more
than 40% inhibition.
3237
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
a
Table 7. GI50 Values for Compounds rac-3, (R)-17j, (S)-21, and (R)-21
GI50 (μM)
compd
U2OS
MDA-MB-231
HeLa
MCF7
rac-3
2.96 (3.73, 2.17)
0.48 (0.51, 0.45)
2.17 (2.17, 2.16)
1.82 (2.22, 1.41)
4.95 (4.81, 5.08)
0.95 (0.87, 1.02)
7.25 (6.40, 8.09)
4.16 (3.94, 4.38)
1.20 0.23
0.14 0.06
0.44 0.13
0.59 0.06
2.03 (1.68, 2.38)
0.22 (0.21, 0.23)
5.42 (4.72, 6.12)
2.17 (2.03, 2.31)
(R)-17j
(S)-21
(R)-21
a
Results are mean SD for n ≥ 3 or the mean of two independent determinations with individual determinations in parentheses or samples run n =
1. Compounds are racemic unless otherwise stated.
a
Table 8. Nonadditive SAR
a
Results are the mean SD for n ≥ 3 or the mean of two independent determinations with individual determinations in parentheses or samples run
n = 1. Compounds are racemic unless otherwise stated.
chromatography (0−7% EtOAc/cyclohexane) to give ether 12a (2.04
bromo-3-iodopyridin-2-amine (850 mg, 2.84 mmol) in DMF/water
(8/1, 15 mL) was stirred at 100 °C under microwave irradiation for 1
h 30 min. The reaction was quenched with brine and extracted with
EtOAc. The combined organics were washed with brine, dried
(Na2SO4), concentrated, and purified by Biotage column chromatog-
raphy (0−30% EtOAc/cyclohexane) to give bromopyridine 14b (1.02
g, 81% over two steps). HRMS (ESI) m/z calcd for C21H20BrN2O4 (M
+ H) 443.0601, found 443.0617. 1H NMR (500 MHz, CDCl3) δ 8.13
(d, J = 2.4 Hz, 1H), 7.90 (d, J = 7.8 Hz, 1H), 7.45 (d, J = 2.4 Hz, 1H),
7.43−7.40 (m, 2H), 7.07−7.04 (m, 2H), 6.95−6.92 (m, 2H), 5.17 (s,
2H), 4.57 (br s, 2H), 3.94 (s, 3H), 3.83 (s, 3H).
Methyl 4-(2-Amino-5-(4-((dimethylamino)methyl)thiophen-
2-yl)pyridin-3-yl)-2-((4-methoxybenzyl)oxy)benzoate 16l. A
solution of bromide 14b (1.01 g, 2.28 mmol), bis(pinacolato)diboron
(870 mg, 3.43 mmol), potassium acetate (680 mg, 6.94 mmol), and
1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II)·DCM
(200 mg, 0.25 mmol) in DMF (11 mL) was stirred at 100 °C under
microwave irradiation for 1 h 30 min. The reaction was quenched with
1
g, 79%). LCMS (ESI) m/z 373 (M + Na). H NMR (500 MHz,
CDCl3) δ 7.70 (d, J = 8.3 Hz, 1H), 7.43−7.40 (m, 2H), 7.20 (d, J = 1.8
Hz, 1H), 7.15 (dd, J = 8.3, 1.8 Hz, 1H), 6.96−6.93 (m, 2H), 5.11 (s,
2H), 3.89 (s, 3H), 3.83 (s, 3H).
Methyl 4-(2-Amino-5-bromopyridin-3-yl)-2-((4-
methoxybenzyl)oxy)benzoate 14b. A solution of bromide 12a
(1.10 g, 3.13 mmol), bis(pinacolato)diboron (1.20 g, 4.72 mmol),
potassium acetate (925 mg, 9.44 mmol), and 1,1′-bis-
(diphenylphosphino)ferrocene]dichloropalladium(II)·DCM (130 mg,
0.16 mmol) in DMF (15 mL) was stirred at 100 °C under microwave
irradiation for 1 h 30 min. The reaction was quenched with brine and
extracted with AcOEt. The combined organics were washed with
brine, dried (Na2SO4), and concentrated to afford the crude boronic
ester 13a.
A solution of crude boronic ester 13a (∼3.13 mmol), sodium
bicarbonate (480 mg, 5.71 mmol), 1,1′-bis(diphenylphosphino)-
ferrocene]dichloropalladium(II)·DCM (125 mg, 0.15 mmol), and 5-
3238
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
brine and extracted with AcOEt. The combined organics were washed
with brine, dried (Na2SO4), and concentrated to afford the crude
boronic acid 15b.
7.65 (br s, 1H), 7.57 (br s, 1H), 7.53 (d, J = 2.4 Hz, 1H), 7.26 (d, J =
1.5 Hz, 1H), 7.22 (dd, J = 8.0, 1.5 Hz, 1H), 7.18 (s, 1H), 7.09 (d, J =
1.5 Hz, 1H), 6.41 (dd, J = 12.1, 8.5 Hz, 1H), 6.09−6.04 (m, 1H), 5.98
(s, 2H), 5.58−5.54 (m, 1H), 3.36 (s, 2H), 2.15 (s, 6H), 1.53 (d, J = 6.3
Hz, 3H).
A solution of crude boronic acid 15b (∼2.28 mmol), 1-(5-
bromothiophen-3-yl)-N,N-dimethylmethanamine13 (1.0 g, 4.54
mmol), tetrakis(triphenylphosphine)palladium(0) (250 mg, 0.22
mmol), and sodium carbonate (485 mg, 4.58 mmol) in DME/water
(8/1, 15 mL) was heated to 110 °C under microwave irradiation for 1
h. The reaction was quenched with brine and extracted with EtOAc.
The combined organics were washed with brine, dried (Na2SO4),
concentrated, and purified by Biotage column chromatography (0−
25% MeOH/DCM) to give amine 16l (569 mg, 50% over two steps).
LCMS (ESI) m/z 504 (M + H). 1H NMR (500 MHz, CDCl3) δ 8.36
(s, 1H), 7.92 (d, J = 7.9 Hz, 1H), 7.56 (d, J = 2.3 Hz, 1H), 7.44−7.41
(m, 2H), 7.24 (s, 1H), 7.14−7.10 (m, 3H), 6.97−6.91 (m, 2H), 5.19
(s, 2H), 4.59 (br s, 2H), 3.94 (s, 3H), 3.82 (s, 3H), 3.54 (s, 2H), 2.36
(s, 6H).
Aminopyridine rac-21 was separated into enantiomers by chiral
HPLC (LUX cellulose II, 90% acetonitrile/2-propanol, 1 mL/min):
Peak 1: retention time 12.3 min, (S)-(Z)-4-(2-Amino-5-(4-
((dimethylamino)methyl)thiophen-2-yl)pyridin-3-yl)-2-((5,5,5-tri-
fluoropent-3-en-2-yl)oxy)benzamide (S)-21. HRMS (ESI) m/z calcd
1
for C24H26F3N4O2S (M + H) 491.1723, found 491.1722. H NMR
(500 MHz, (CD3)2SO) δ 8.27 (d, J = 2.4 Hz, 1H), 7.88 (d, J = 8.0 Hz,
1H), 7.65 (br s, 1H), 7.57 (br s, 1H), 7.53 (d, J = 2.4 Hz, 1H), 7.26 (d,
J = 1.5 Hz, 1H), 7.22 (dd, J = 8.0, 1.5 Hz, 1H), 7.18 (s, 1H), 7.09 (d, J
= 1.5 Hz, 1H), 6.41 (dd, J = 12.1, 8.5 Hz, 1H), 6.09−6.04 (m, 1H),
5.98 (s, 2H), 5.58−5.54 (m, 1H), 3.36 (s, 2H), 2.15 (s, 6H), 1.53 (d, J
= 6.3 Hz, 3H).
Peak 2: retention time 15.2 min, (R)-(Z)-4-(2-Amino-5-(4-
((dimethylamino)methyl)thiophen-2-yl)pyridin-3-yl)-2-((5,5,5-tri-
fluoropent-3-en-2-yl)oxy)benzamide (R)-21. HRMS (ESI) m/z calcd
Methyl 4-(2-Amino-5-(4-((dimethylamino)methyl)thiophen-
2-yl)pyridin-3-yl)-2-hydroxybenzoate 18. A solution of phenol
ether 16l (560 mg, 1.11 mmol) in DCM (7 mL) was treated with
trifluoroacetic acid (800 μL, 10.81 mmol) at 0 °C. After 1 h 30 min the
mixture was brought to pH ≈ 5−6 with 1 M NaOH and 1 M HCl, the
aqueous layer separated and extracted with DCM. The combined
organic layers were concentrated and purified by Biotage column
chromatography (0−15% MeOH/DCM) to give phenol 18 (394 mg,
92%). HRMS (ESI) m/z calcd for C20H22N3O3S (M + H) 384.1376,
1
for C24H26F3N4O2S (M + H) 491.1723, found 491.1719. H NMR
(500 MHz, (CD3)2SO) δ 8.27 (d, J = 2.4 Hz, 1H), 7.88 (d, J = 8.0 Hz,
1H), 7.65 (br s, 1H), 7.57 (br s, 1H), 7.53 (d, J = 2.4 Hz, 1H), 7.26 (d,
J = 1.5 Hz, 1H), 7.22 (dd, J = 8.0, 1.5 Hz, 1H), 7.18 (s, 1H), 7.09 (d, J
= 1.5 Hz, 1H), 6.41 (dd, J = 12.1, 8.5 Hz, 1H), 6.09−6.04 (m, 1H),
5.98 (s, 2H), 5.58−5.54 (m, 1H), 3.36 (s, 2H), 2.15 (s, 6H), 1.53 (d, J
= 6.3 Hz, 3H).
1
found 384.1391. H NMR (500 MHz, MeOD) δ 8.28 (d, J = 2.4 Hz,
Biochemical Assays. Nek2 and Plk1 biochemical assays were
performed as reported previously. MPS1, AurA and CDK2 counter-
screen assays were carried out using similar procedures.5,6
1H), 7.99 (d, J = 8.2 Hz, 1H), 7.68 (d, J = 2.4 Hz, 1H), 7.58 (d, J = 1.4
Hz, 1H), 7.39 (d, J = 1.4 Hz, 1H), 7.10 (d, J = 1.7 Hz, 1H), 7.07 (dd, J
= 8.2, 1.7 Hz, 1H), 4.30 (s, 2H), 4.00 (s, 3H), 2.88 (s, 6H).
Cellular Assays. CellTiter-Blue Assay for Growth Inhibition.
U2OS human osteosarcoma cells (American Type Culture Collection,
Manassas, VA, U.S.) were grown in McCoy’s 5A medium
supplemented with 1.5 mM L-glutamine, 25 mM HEPES, 2%
penicillin/streptomycin (Invitrogen, Paisley, U.K.), and 10% fetal
bovine serum (Biosera, Ringmer, East Sussex, U.K.). Cells were
maintained in a humidified atmosphere of 5% CO2 at 37 °C. The
medium was aspirated, and the cells were washed with PBS
(Invitrogen, Paisley, U.K.), trypsinized (internal supply, 0.25% versene
trypsin with EDTA), neutralized, and counted. Cells were seeded into
384-well clear tissue culture treated microtiter plates (Corning B.V.
Life Sciences, Amsterdam, The Netherlands) at 200 cells per well in 45
μL of the respective medium. Columns 1 and 24 had no cells added
and were plated with 45 μL of medium alone. Cells were incubated at
37 °C/5% CO2. At 24 h after plating, test compounds, etoposide as
positive control (Sigma-Aldrich, Gillingham, Dorset, U.K.), or DMSO
at 1% v/v final concentration (Fisher Scientific, Loughborough,
Leicestershire, U.K.) were dispensed using an Echo liquid handling
system (Labcyte, Dublin, Ireland). After 92 h, 5 μL of CellTiter-Blue
Rreagent (Promega, Southampton, U.K.) was added to the cells using
a multidrop dispenser (Thermo Electron, Basingstoke, Hants, U.K.)
and incubated for 4 h in a humidified atmosphere of 5% CO2 at 37 °C.
After the incubation, the plates were placed at room temperature for
40 min before fluorescence was recorded (560Ex/590Em) on an
EnVision 2103 plate reader (PerkinElmer Life Sciences). Data were
plotted as percentage of DMSO control against compound
concentration using GraphPad Prism 5 software. The 50% growth
inhibition (GI50) was calculated as the compound concentration
required to reduce the cell number by 50% compared with the DMSO
control.
(
)-(Z)-Methyl 4-(2-Amino-5-(4-((dimethylamino)methyl)-
thiophen-2-yl)pyridin-3-yl)-2-((5,5,5-trifluoropent-3-en-2-yl)-
oxy)benzoate 20. A solution of phenol 18 (64 mg, 0.17 mmol),
( )-5,5,5-trifluoropent-3-yn-2-ol14 (20% w/w solution in EtOAc, 0.31
mmol), and triphenylphosphine (70 mg, 0.27 mmol) in DCM (1 mL)
was cooled at 0 °C and treated with di-tert-butyl azodicarboxylate (60
mg, 0.26 mmol). The mixture was allowed to reach room temperature
and stirred overnight. Additional batches of reagents were added as
required until the reaction reached completion. The mixture was
diluted with DCM and quenched with water. The organic layer was
separated and extracted with DCM. The combined organics were
dried (Na2SO4), concentrated, and purified by Biotage column
chromatography (0−15% MeOH/DCM) to give ether 19.
A solution of ether 19 (∼0.050 mmol) in MeOH (2 mL) was
treated with palladium on calcium carbonate (poisoned with lead
(Lindlar’s catalyst), 5% w/w, 5 mg, 2.4 μmol) and stirred in an
atmosphere of hydrogen overnight. Additional batches of palladium on
calcium carbonate (poisoned with lead, 5% w/w, 7 mg, 3.3 μmol) were
added, and the mixture was stirred in an atmosphere of hydrogen until
completion. The mixture was filtered over Celite, washing with
MeOH, the solvent removed under reduced pressure, and the residue
purified by semipreparative reverse phase HPLC (Phenomenex
Gemini C18 column; 15 min gradient 25−50% MeOH/water, 0.1%
formic acid; 5 mL/min) to give alkene 20 (12 mg, 14% over two
steps). HRMS (ESI) m/z calcd for C25H27F3N3O3S (M + H)
1
506.1720, found 506.1701. H NMR (500 MHz, MeOD) δ 8.29 (d, J
= 2.4 Hz, 1H), 7.88 (d, J = 8.0 Hz, 1H), 7.65 (d, J = 2.4 Hz, 1H), 7.56
(s, 1H), 7.37 (d, J = 1.5 Hz, 1H), 7.22 (dd, J = 8.0, 1.5 Hz, 1H), 7.17
(s, 1H), 6.27 (dd, J = 12.1, 8.7 Hz, 1H), 5.94−5.85 (m, 1H), 5.49−
5.45 (m, 1H), 4.24 (s, 2H), 3.92 (s, 3H), 2.83 (s, 6H), 1.55 (d, J = 6.4
Hz, 3H).
( )-(Z)-4-(2-Amino-5-(4-((dimethylamino)methyl)thiophen-
2-yl)pyridin-3-yl)-2-((5,5,5-trifluoropent-3-en-2-yl)oxy)-
benzamide rac-21. Ester 20 (16 mg, 0.032 mmol) was treated with
ammonia in methanol (7M, 4 mL) and heated to 75 °C in a closed-cap
vial for 4 days. The mixture was concentrated and the residue purified
by Biotage column chromatography (0−15% MeOH/DCM) to give
amide rac-21 (9 mg, 56%). HRMS (ESI) m/z calcd for
MDA-MB-231 cells (American Type Culture Collection, Manassas,
VA, U.S.) were grown in RPMI medium (Invitrogen, Paisley, U.K.)
supplemented with 2% penicillin/streptomycin and 10% fetal bovine
serum.
HeLa cells (American Type Culture Collection, Manassas, VA,
U.S.) were grown in Dulbecco’s modified Eagle medium (Invitrogen,
Paisley, U.K.) supplemented with 2% penicillin/streptomycin and 10%
fetal bovine serum.
1
C24H26F3N4O2S (M + H) 491.1723, found 491.1749. H NMR (500
MCF7 cells (American Type Culture Collection, Manassas, VA,
MHz, (CD3)2SO) δ 8.27 (d, J = 2.4 Hz, 1H), 7.88 (d, J = 8.0 Hz, 1H),
U.S.) were grown in Dulbecco’s modified Eagle medium supple-
3239
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
mented with 2% penicillin/streptomycin and 10% fetal bovine serum.
They were plated at 800 cells per well.
Accession Codes
†Atomic coordinates and structure factors for the crystal
structure of ligand bound Nek2 can be accessed using the
following PDB codes: 4A4X for rac-17j; 4AFE for rac-21.
Phosphorylated C-Nap1 in Cell Assay. U2OS human
osteosarcoma cells (American Type Culture Collection, Manassas,
VA, U.S.) were grown in McCoy’s 5A medium supplemented with 1.5
mM L-glutamine, 25 mM HEPES, 2% penicillin/streptomycin
(Invitrogen, Paisley, U.K.), and 10% fetal bovine serum (Biosera,
Ringmer, East Sussex, U.K.). Cells were maintained in a humidified
atmosphere of 5% CO2 at 37 °C. The medium was aspirated, and the
cells were washed with PBS (Invitrogen, Paisley, U.K.), trypsinized
(internal supply, 0.25% versene trypsin with EDTA), neutralized, and
counted. Cells were seeded into 96-well black clear bottom tissue
culture treated microtiter plates (PerkinElmer Life Sciences, Waltham,
MA, U.S.) at 10 000 cells per well in 180 μL of medium. Cells were
incubated for 24 h at 37 °C/5% CO2, after which aphidicolin (Sigma-
Aldrich, Gillingham, Dorset, U.K.) was added to all wells for cell
synchronization at a final concentration of 2 μg/mL for 16−18 h. The
cells were then released into 180 μL of fresh medium. Simultaneously,
the cells were treated in triplicate wells with test compound in DMSO
at 1% v/v final concentration (Fisher Scientific, Loughborough,
Leicestershire, U.K.) and incubated for 3 h. 1% DMSO final
concentration was used as a negative control. The cells were then
fixed for 15 min at 2−8 °C with 50 μL of cold 100% methanol. The
plates were washed once with 100 μL of PBS before blocking
nonspecific binding with 50 μL of 5% w/v bovine serum albumin
(Sigma-Aldrich, Gillingham, Dorset, U.K.) in PBS for 1 h at room
temperature with gentle agitation. A rabbit polyclonal antibody raised
to phosphorylated C-Nap1 (generated by Peptide Specialty Labo-
ratories, GmbH, Heidelberg, Germany) was added at a dilution of
1:750, and a monoclonal antibody raised to the centrosomal marker γ-
tubulin (Sigma-Aldrich, Gillingham, Dorset, U.K.) was added at a
dilution of 1:500 for 1 h at room temperature with gentle agitation.
Following a further wash with PBS, Alexa fluor 488 goat anti-rabbit
IgG and Alexa fluor 568 goat anti-mouse IgG (Invitrogen, Paisley,
U.K.) were added at a final concentration of 4 μg/mL in 5% BSA/PBS
for 1 h at room temperature under gentle agitation. After another wash
with PBS, the nuclear stain DAPI (Invitrogen, Paisley, U.K.) was
added at a final concentration of 2.5 μg/mL in PBS for 10 min at room
temperature with gentle agitation. The plate was washed in PBS once
again and refrigerated until ready to image. The assay plates were read
on the IN Cell Analyzer 1000 using the Workstation 1000 acquisition
software (GE Healthcare, Amersham, U.K.). The instrument was
equipped with a 20× dry Nikon objective, a D360/40X DAPI
excitation filter, an HQ480/40X FITC excitation filter, an HQ565/
30X excitation filter, an HQ460/40M bandwidth emission filter, an
HQ535nm/50M bandwidth emission filter, and an HQ620nm/60M
bandwidth emission filter. The exposure times were consistently 200
ms in the DAPI channel, 400 ms in the FITC channel, and 1000 ms in
the red channel. Fifteen fields of view were imaged in all wells. IN Cell
Investigator Developer Toolbox (GE Healthcare) was used to analyze
the assay data. The algorithm was written to identify the DAPI-stained
nuclei and then segment the cells, setting a small collar around the
nuclei as the centrosomal inclusion area. The γ-tubulin-identified
centrosomes and phosphorylated C-Nap1 spots were chosen based on
size and pixel intensity and were only counted if they were located
within the centrosomal inclusion area. A “one to one link” was written
into the protocol so that only phosphorylated C-Nap1 spots
colocalized with the centrosomal γ-tubulin were counted. Data were
plotted as percentage of DMSO control against compound
concentration using GraphPad Prism 5 software.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: +44 20 87224353. E-mail: swen.hoelder@icr.ac.uk.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge NHS funding to the NIHR Biomedical
Research Centre and funding from Cancer Research UK
Programme Grant C309/A8274. A.M.F. acknowledges support
from Cancer Research UK, The Wellcome Trust, and the
Association for International Cancer Research. R.B. acknowl-
edges support from Cancer Research UK (Grant C24461/
A10285 and infrastructure support for Structural Biology at the
ICR), a Royal Society Research Fellowship, and the Career
Development Programme of the ICR. We are thankful to Prof.
Julian Blagg, Prof. Roger Griffin, Prof. Herbie Newell, Dr. Ian
Collins, and Prof. Paul Workman for helpful discussions. We
also thank Dr. Amin Mirza, Meirion Richards, and Dr. Maggie
Liu for their help with HPLC, NMR, and mass spectrometry.
Jessica Schmitt is acknowledged for biochemical assays. We are
indebted to the staff of ESRF beamline ID29 as well as our
many colleagues of the Section of Structural Biology (ICR in
London) for their support during data collection.
ABBREVIATIONS USED
■
Abl, Abelson murine leukemia viral oncogene; ATP, adenosine
triphosphate; AurA, Aurora A; CDK2, cyclin-dependent kinase
2; C-Nap1, centrosomal Nek2-associated protein 1; DCE, 1,2-
dichloroethane; DCM, dichloromethane; DME, 1,2-dimethox-
yetane; DMF, N,N-dimethylformamide; DFG, Asp-Phe-Gly;
Hec1, highly expressed in cancer 1; DMSO, dimethyl sulfoxide;
HMGA2, high mobility group AT-hook 2; Lck, lymphocyte-
specific protein tyrosine kinase; LE, ligand efficiency; Mad,
mitotic arrest deficient-like; MPS-1, human monopolar spindle
1; Nek, NIMA (never in mitosis gene a)-related kinase; Nlp,
ninein-like protein; PAMPA, parallel artificial membrane assay;
Plk, polo-like kinase; PMB, p-methoxybenzyl; SAC, spindle
assembly checkpoint; SAR, structure−activity relationship; SD,
standard deviation; TFA, trifluoroacetic acid; THF, tetrahy-
drofuran
REFERENCES
■
(1) O’Regan, L.; Blot, J.; Fry, A. M. Mitotic regulation by NIMA-
related kinases. Cell Div. 2007, 2, 25−36.
(2) (a) Lou, Y.; Yao, J.; Zereshki, A.; Dou, Z.; Ahmed, K.; Wang, H.;
Hu, J.; Wang, Y.; Yao, X. NEK2A interacts with MAD1 and possibly
functions as a novel integrator of the spindle checkpoint signaling. J.
Biol. Chem. 2004, 279, 20049−20057. (b) Chen, Y.; Riley, D. J.;
Zheng, L.; Chen, P. L.; Lee, W. H. Phosphorylation of the mitotic
regulator protein Hec1 by Nek2 kinase is essential for faithful
chromosome segregation. J. Biol. Chem. 2002, 277, 49408−49416.
(3) (a) Henise, J. C.; Taunton, J. Irreversible Nek2 kinase inhibitors
with cellular activity. J. Med. Chem. 2011, 54, 4133−4146. (b) Mardin,
B. R.; Lange, C. L.; Baxter, J. E.; Hardy, T.; Scholz, S. R.; Fry, A. M.;
Schiebel, E. Components of the Hippo pathway cooperate with Nek2
kinase to regulate centrosome disjunction. Nat. Cell Biol. 2010, 12,
1166−1176.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures, analytical data for final compounds
11, 17 and intermediates, a summary of crystallographic
analysis results of aminopyridine rac-17j and rac-21, HPLC
traces for compounds 21, C-Nap1 antibody characterization,
and pictures of assay plates. This material is available free of
charge via the Internet at http://pubs.acs.org.
3240
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241

Journal of Medicinal Chemistry
Article
(4) (a) Fletcher, L.; Cerniglia, G. J.; Nigg, E. A.; Yend, T. J.; Muschel,
R. J. Inhibition of centrosome separation after DNA damage: a role for
Nek2. Radiat. Res. 2004, 162, 128−135. (b) Kokuryo, T.; Senga, T.;
Yokoyama, Y.; Nagino, M.; Nimura, Y.; Hamaguchi, M. Nek2 as an
effective target for inhibition of tumorigenic growth and peritoneal
dissemination of cholangiocarcinoma. Cancer Res. 2007, 67, 9637−
9642. (c) Tsunoda, N.; Kokuryo, T.; Oda, K.; Senga, T.; Yokoyama,
Y.; Nagino, M.; Nimura, Y.; Hamaguchi, M. Nek2 as a novel molecular
target for the treatment of breast carcinoma. Cancer Sci. 2009, 100,
111−116.
(5) Whelligan, D. K.; Solanki, S.; Taylor, D.; Thomson, D. W.;
Cheung, K. M. J.; Boxall, K.; Mas-Droux, C.; Barillari, C.; Burns, S.;
Grummitt, C. G.; Collins, I.; van Montfort, R. L. M.; Aherne, G. W.;
Bayliss, R.; Hoelder, S. Aminopyrazine inhibitors binding to an
unusual inactive conformation of the mitotic kinase Nek2: SAR and
structural characterization. J. Med. Chem. 2010, 53, 7682−7698.
(6) Solanki, S.; Innocenti, P.; Mas-Droux, C.; Boxall, K.; Barillari, C.;
van Montfort, R. L. M.; Aherne, G. W.; Bayliss, R.; Hoelder, S.
Benzimidazole inhibitors induce a DFG-out conformation of never in
mitosis gene A-related kinase 2 (Nek2) without binding to the back
pocket and reveal a nonlinear structure−activity relationship. J. Med.
Chem. 2011, 54, 1626−1639.
(17) See Experimental Section and Supporting Information for
details.
(18) Graczyk, P. P. Gini coefficient: a new way to express selectivity
of kinase inhibitors against a family of kinases. J. Med. Chem. 2007, 50,
5773−5779.
(19) See Figure S2 in Supporting Information for more details.
(20) See Figure S3 in the Supporting Information for samples of the
assay readout.
(21) 87% and 95%, respectively, of aminopyridines rac-17j and rac-
21 were metabolized in MLM after 30 min. Data for the benzimidazole
series have been previously published (see ref 6).
(7) Mitsunobu, O. The use of diethyl azodicarboxylate and
triphenylphosphine in synthesis and transformation of natural-
products. Synthesis 1981, 1−28.
(8) Ishiyama, T.; Murata, M.; Miyaura, N. Palladium(0)-catalyzed
cross-coupling reaction of alkoxydiboron with haloarenes: a direct
procedure for arylboronic esters. J. Org. Chem. 1995, 60, 7508−7510.
(9) (a) Kundo, M.; Khairatkar-Joshi, N.; Nadkarni, S. M.; Pansare, R.
M.; Karnik, P. V. Bicyclic Heteroaryl Derivatives as Cannabinoid
Receptor Modulators. Patent WO 2007/096764, 2007; Glenmark
Pharmaceuticals SA. (b) Binch, H.; Robinson, D.; Miller, A.; Fraysse,
D. Pyrrolopyrazines and Pyrazolopyrazines Useful as Inhibitors of
Protein Kinases. Patent WO 2006/058074, 2006; Vertex Pharma.
(10) Miyaura, N.; Suzuki, A. Palladium-catalyzed cross-coupling
reactions of organoboron compounds. Chem. Rev. 1995, 95, 2457−
2483.
(11) See Supporting Information for detailed procedures.
(12) Hilton, S.; Naud, S.; Caldwell, J. C.; Boxall, K.; Burns, S.;
Anderson, V. E.; Antoni, L.; Allen, C. A.; Pearl, L. H.; Oliver, A. W.;
Aherne, G. W.; Garrett, M. D.; Collins, I. Identification and
characterisation of 2-aminopyridine inhibitors of checkpoint kinase
2. Bioorg. Med. Chem. 2010, 18, 707−718.
(13) 1-(5-Bromothiophen-3-yl)-N,N-dimethylmethanamine was pre-
pared by reductive amination of the known corresponding aldehyde:
Deng, J.; Kerns, J. K.; Jin, Q.; Lin, G.; Lin, X.; Lindenmuth, M.; Neipp,
C. E.; Nie, H.; Thomas, S. M.; Widdowson, K. L. Chemical
Compounds. Patent WO 2007/005534, 2007; Smithkline Beecham
Corporation. See Supporting Information for experimental details.
(14) ( )-5,5,5-Trifluoropent-3-yn-2-ol was prepared according to a
modified literature method: Yamazaki, T.; Mizutani, K.; Kitazume, T.
Modified preparation of trifluoromethylated propargylic alcohols and
its application to chiral 2,6-dideoxy-6,6,6-trifluorosugars. J. Org. Chem.
1995, 60, 6046−6056 . See Supporting Information for experimental
details.
(15) Whelligan, D. K.; Solanki, S.; Taylor, D.; Thomson, D. W.;
Cheung, K. M. J.; Hoelder, S. Unpublished results.
(16) (a) Rapley, J.; Baxter, J. E.; Blot, J.; Wattam, S. L.; Casenghi, M.;
Meraldi, P.; Nigg, E. A.; Fry, A. M. Coordinate regulation of the
mother centriole component Nlp by Nek2 and Plk1 protein kinases.
Mol. Cell. Biol. 2005, 25, 1309−1324. (b) For a study on Plk1
inhibitors, see the following: Emmitte, K. A.; Adjebang, G. M.;
Andrews, C. W.; Alberti, J. G.; Bambal, R.; Chamberlain, S. D.; Davis-
Ward, R. G.; Dickson, H. D.; Hassler, D. F.; Hornberger, K. R.;
Jackson, J. R.; Kuntz, K. W.; Lansing, T. J.; Mook, R. A. Jr.; Nailor, K.
E.; Pobanz, M. A.; Smith, S. C.; Sung, C. M.; Cheung, M. Design of
potent thiophene inhibitors of polo-like kinase 1 with improved
solubility and reduced protein binding. Bioorg. Med. Chem. Lett. 2009,
19, 1694−1697.
3241
dx.doi.org/10.1021/jm201683b | J. Med. Chem. 2012, 55, 3228−3241