Kinesin spindle protein (KSP) inhibitors. Part 2: The design, synthesis, and characterization of 2,4-diaryl-2,5-dihydropyrrole inhibitors of the mitotic kinesin KSP

Article abstract of DOI:10.1016/j.bmcl.2006.01.030

The evolution of 2,4-diaryl-2,5-dihydropyrroles as inhibitors of KSP is described. Introduction of basic amide and urea moieties to the dihydropyrrole nucleus enhanced potency and aqueous solubility, simultaneously, and provided compounds that caused mito

Full text of DOI:10.1016/j.bmcl.2006.01.030

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1775–1779
Kinesin spindle protein (KSP) inhibitors. Part 2: The design,
synthesis, and characterization of 2,4-diaryl-2,5-dihydropyrrole
inhibitors of the mitotic kinesin KSP
Mark E. Fraley,a,* Robert M. Garbaccio,a Kenneth L. Arrington,a William F. Hoffman,a
Edward S. Tasber,a Paul J. Coleman,a Carolyn A. Buser,b Eileen S. Walsh,b
Kelly Hamilton,b Christine Fernandes,b Michael D. Schaber,b Robert B. Lobell,b
Weikang Tao,b Victoria J. South,b Youwei Yan,c Lawrence C. Kuo,c
Thomayant Prueksaritanont,d Cathy Shu,d Maricel Torrent,a David C. Heimbrook,b
Nancy E. Kohl,b Hans E. Huberb and George D. Hartmana
aDepartment of Medicinal Chemistry, Merck Research Laboratories, West Point, PA 19486, USA
bDepartment of Cancer Research, Merck Research Laboratories, West Point, PA 19486, USA
cDepartment of Structural Biology Merck Research Laboratories, West Point, PA 19486, USA
dDepartment of Drug Metabolism, Merck Research Laboratories, West Point, PA 19486, USA
Received 4 November 2005; revised 5 January 2006; accepted 6 January 2006
Available online 24 January 2006
Abstract—The evolution of 2,4-diaryl-2,5-dihydropyrroles as inhibitors of KSP is described. Introduction of basic amide and urea
moieties to the dihydropyrrole nucleus enhanced potency and aqueous solubility, simultaneously, and provided compounds that
caused mitotic arrest of A2780 human ovarian carcinoma cells with EC50s < 10 nM. Ancillary hERG activity was evaluated for this
series of inhibitors.
Ó 2006 Elsevier Ltd. All rights reserved.
Mitotic kinesins are microtubule-associated motor pro-
teins that utilize the energy from ATP hydrolysis for
bipolar spindle assembly, maintenance and elongation,
chromosome alignment and segregation, and microtu-
bule depolymerization, among other functions during
cell division.1 KSP (kinesin spindle protein, HsEg5) is
a mitotic kinesin that causes centrosome separation, a
process required for bipolar spindle formation and
maintenance.2 Inhibition of KSP with small molecules
results in collapse of bipolar spindles and gives rise to
monopolar spindle arrays, referred to as monoasters.3
Importantly, this phenotype is accompanied by cell cycle
arrest in mitosis, and prolonged mitotic arrest in tumor
cells leads to apoptosis.4,5 KSP inhibitors therefore have
potential as general antiproliferative agents useful for
the treatment of cancer. Additionally, KSP inhibition
represents a novel and specific mechanism to target the
mitotic spindle that may be devoid of the neuropathy-
associated, mechanism-based side effects common to
the taxanes and other natural products that target
microtubules. KSP inhibitors may also offer advantages
over these agents with regard to formulation and tumor
resistance. Efforts to establish proof-of-concept with
ispinesib (SB-715992), a KSP inhibitor of the quinazoli-
none class,6 in Phase II clinical trials are currently ongo-
ing. In this communication, we describe the design,
synthesis, and evaluation of 2,4-diaryl-2,5-dihydropyr-
roles as inhibitors of KSP.
We recently reported the development of 3,5-diaryl-4,5-
dihydropyrazoles as potent and cell-active KSP inhibi-
tors.7 These compounds demonstrated high selectivity
for KSP versus other kinesins, a consequence of their allo-
steric mode of binding. Part of our efforts to expand the
scope of this series centered on modifications to the core
structure designed to improve reactivity in amide bond-
forming reactions and to preserve the orientation of the
two aryl groups to maintain potency. We rationalized
Keywords: KSP; hERG; Dihydropyrroles; Kinesin spindle protein;
Antimitotics; Antiproliferative; Anticancer.
*
Corresponding author. Tel.: +1 215 652 6937; fax: +1 215 652
7310; e-mail: mark_fraley@merck.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.01.030

1776
M. E. Fraley et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1775–1779
Table 1. KSP inhibitory activity of compounds 2–7
that a 2,5-dihydropyrrole nucleus might serve as a viable
replacement for the 4,5-dihydropyrazole core in that the
hybridization of the aryl-bearing carbon atoms would
be maintained, allowing for a similar presentation of the
diaryl array (Fig. 1). To test this hypothesis, a synthetic
route to racemic 1 was developed based on sequential
Heck coupling reactions of aryl diazonium intermediates
reported by Carpes and Correia (Fig. 2).8 The substitu-
tion pattern selected for the diaryl rings of 1 was found
to be optimal for the dihydropyrazole series, and thus
was maintained for our exploratory studies in this new
series.
F
F
F
F
N
R
N
N
R
O
O
core A
core B
Compound
Core
R
KSP IC50 (nM)
2 (+/À)
3 (+/À)
4 (+/À)
5 (+/À)
6 (À, S)
7 (+/À)
A
B
A
B
A
B
Me
Me
4000
94
t-Bu
t-Bu
NMe2
NMe2
85
Acylation of (+/À)-1 with acetyl, pivaloyl or dimethyl-
carbamyl chloride provided dihydropyrrole amides 2
and 4 and urea 6 (following resolution by chiral chroma-
tography), respectively.9 The KSP inhibitory activity10
of this initial set of compounds was compared to the
corresponding series of dihydropyrazoles, 3, 5, and 7
(Table 1). While compound 2 showed lower potency rel-
ative to 3, compounds 4 and 6 compared favorably to
their counterparts 5 and 7, suggesting that branched acyl
groups were necessary for potency in the dihydropyrrole
series. In addition, the S configuration of 6 was deter-
mined to be a critical feature of potency,11 an observa-
tion consistent with the stereochemical requirement for
potency in the dihydropyrazole series. Having estab-
lished the dihydropyrrole ring system as a suitable alter-
native to the dihydropyrazole core, we next devised an
113
38
84
asymmetric sequence to (S)-1 to support our optimiza-
tion efforts.
As illustrated in Figure 3, the route began with a one-
pot preparation of alcohol 9 from nitrile 812 by the
method of Maeda et al.13 Swern oxidation of 9 followed
by regioselective enolization of ketone 10 with NaH-
MDS14 and trapping of the resultant enolate with
PhNTf2 furnished enol triflate 11. Suzuki cross-coupling
reaction of 11 with 2,5-difluorophenylboronic acid pro-
ceeded smoothly to afford 12, and removal of the Boc
group then provided (S)-1.
sp3
sp3
F
F
F
F
OTBS
CN
TBSCN
O
sp2
Cl
sp2
Cl
N
O
KCN, 18-C-6
95%
N
N
8
(S)-(+)
O
R
R
4,5-Dihydropyrazole
2,5-Dihydropyrrole
TBSO
N
2. H2, Pt/C
1. PhMgBr
Figure 1. Design strategy: maintain hybridization of aryl-substituted
carbon atoms.
MTBE, then THF
0 to 23 ˚C
3. Boc2O
4. HF• Et3N
33% (4 steps)
O
HO
H
F
-
H
BF4
+
(COCl)2/ DMSO
F
F
F
N2
N
N
CH2Cl2
-78 to 0 ˚C
85%
Boc
Boc
N
Pd(OAc)2
CCl4/H2O, 23 ˚C
OH
N2
Boc
N
10
9
Boc
F
TfO
-
NaHMDS;
PhNTf2
BF4
+
H
F
B(OH)2
F
F
TFAA
N
THF
-78 to 0 ˚C
Pd(PPh3)4
Na2CO3
dioxane, reflux
73% (2 steps)
Boc
2,6-lutidine, toluene
0 ˚C to reflux
Pd2(dba)3
CH3CN, 23 ˚C
62%
N
11
Boc
52% (2 steps)
F
F
F
F
F
F
TFA
F
F
TFA
H
H
CH2Cl2
100%
CH2Cl2
100%
N
N
H
N
N
H
Boc
Boc
(+/-)-1
12
Figure 3. Synthesis of (S)-1.
(S)-1
Figure 2. Synthesis of (+/À)-1.

M. E. Fraley et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1775–1779
1777
Table 2. Properties of dihydropyrroles 13–22
With (S)-A in hand, a wide range of branched basic
amides and ureas were prepared with the goal of identi-
fying compounds with improved potency and aqueous
solubility, important for iv formulation, over leads 4
and 6. The KSP inhibitory profiles of a select set of com-
pounds, 13–22, prepared by the methods illustrated in
Figure 4, are presented in Table 2.15 Included are the
EC50 values for mitotic arrest (Cell EC50)16 in A2780
human ovarian carcinoma cells and binding data against
the IKr potassium channel hERG (human Ether-a-go-go
Related Gene).17 We closely monitored hERG binding
in that blockade of this ion channel has been implicated
in drug-induced prolongation of the heart-rate corrected
QT interval of the ECG, a side effect that can potentiate
cardiac arrhythmia and lead to torsades de pointes or
even sudden death in extreme cases.18
F
F
H
N
R
O
Compound
R
KSP IC50 Cell EC50 HERG
(nM)
(nM)
IC50 (nM)
HN
13
14
15
16
17
2.6
6.8
1300
N
Me
AcN
50
69
44
11
12
>10,000
1300
N
Me
In general, the data for this select set of compounds indi-
cated that incorporation of basic unsymmetrical ureas
or branched a- or b-amino amides provided low molec-
ular weight inhibitors with excellent potency and solu-
bility properties. Specifically, in the urea series, the
4-piperidinyl group of urea 13 proved optimal in terms
of potency, cellular activity, and aqueous solubility
(>10 mg/mL, pH 5). However, the basicity of 13 con-
ferred binding affinity to hERG, as determined by com-
parison to the non-basic acylated derivative 14. In the
amide series, a number of highly potent and water solu-
ble inhibitors were identified. Interestingly, in the case of
the a-amino amides, only the S configuration of the
a-stereocenter imparted a high level of potency (com-
pare diastereomers 17 and 18). These basic amides, like
the basic ureas, demonstrated moderate levels of hERG
binding. The hERG IC50 value for 19 was corroborated
by cellular electrophysiology using a standard whole-cell
voltage clamp assay in CHO cells that stably expressed
hERG (IC50 = 6500 nM). As a partial solution to this
potential liability, we found that neutral acyl groups,
such as those incorporated in 21 and 22, ameliorated
hERG activity without major loss of potency, but at
the expense of aqueous solubility. The following paper
details strategies to solubilize charge-neutral inhibitors
from this series.
N
Me
16
HN
Me
Me
5.2
3.6
2200
NH2
Me
2400
Me
(S)
NH2
Me
18
19
3100
2.0
ND
8.6
ND
Me
(R)
NH2
3500
NH2
Me
Me
20
2.7
7.1
1500
Me
NH2
Me
Me
Me
21
22
7.4
11
13
20
25,000
11,000
HN
O
O
OH
OH
ND, not determined.
1. CDI
THF, 70 ˚C
2. MeI
F
F
F
Comparative analysis of the X-ray crystal structures of
compounds 13 and 19 with 6 provided possible explana-
tions for their differences in KSP inhibitory activity
(Fig. 5).19 In the structure of 13, the piperidinyl group
projects away from the hydrophobic regions of the allo-
steric pocket to place the basic nitrogen at the polar sol-
vent front, minimizing the energy cost of desolvation.
Interestingly, this preference gives rise to a rotamer con-
formation for the urea moiety that is transposed relative
to urea 6. Adoption of this rotamer conformation di-
rects the methyl group of 13 into a hydrophobic region
just below the plane of the dihydropyrrole core, further
increasing the binding energy.20 Similarly with 19, posi-
tioning of the primary amino group at the solvent
interface has the effect of directing the lipophilic
cyclopropyl sidechain into this hydrophobic region,
and thus offers an explanation for the requirement of
F
CH3CN, 60 ˚C
H
H
3. NHMe
HN
N
N
H
N
O
Me
(S)-1
N
13
Boc
4. TFA, 50% yield
1.
BocHN CO2H
F
F
F
F
H
H
PyBOP, Et3N
DMF, 23 ˚C
2. TFA/CH2Cl2
68%
N
H
N
O
(S)-1
NH2
19
Figure 4. Synthesis of 13 and 19 from (S)-1.

1778
M. E. Fraley et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1775–1779
Figure 5. Crystal structures of 6, 13 and 19 bound to the allosteric site of KSP. Protein shown in solid ribbon. Green patch indicates location of a
hydrophobic region in the binding site.
´
2. Blangy, A.; Lane, H. A.; d’Herin, P.; Harper, M.; Kress,
M.; Nigg, E. A. Cell 1995, 83, 1159.
Table 3. Pharmacokinetic profile of 19 following iv dosing to animals
F
3. Mayer, T. U.; Kapoor, T. M.; Haggarty, S. J.; King, R.
W.; Schreiber, S. L.; Mitchison, T. J. Science 1999, 286,
971.
F
H
4. Marcus, A. I.; Peters, U.; Thomas, S. L.; Garrett, S.;
Zelnak, A.; Kapoor, T. M.; Giannakakou, P. J. Biol.
Chem. 2005, 280, 11569.
5. Tao, W.; South, V. J.; Zhang, Y.; Davide, J. P.; Farrell, L.;
Kohl, N. E.; Sepp-Lorenzino, L.; Lobell, R. B. Cancer
Cell 2005, 8, 49.
N
O
NH2
19
Species
Dose
(mg/kg)
Cl
(mL/min/kg)
Vdss
t1/2
6. For reviews, see: (a) Duhl, D. M.; Renhowe, P. A. Curr.
Opin. Drug Discov. Dev. 2005, 8, 431; (b) Coleman, P. J.;
Fraley, M. E. Expert Opin. Ther. Patents 2004, 14, 1659.
7. Cox, C. D.; Breslin, M. J.; Mariano, B. J.; Coleman, P. J.;
Buser, C. A.; Walsh, E. S.; Hamilton, K.; Huber, H. E.;
Kohl, N. E.; Torrent, M.; Yan, Y.; Kuo, L. C.; Hartman,
G. D. Bioorg. Med. Chem. Lett. 2005, 15, 2041.
8. Carpes, M. J. S.; Correia, C. R. D. Synlett 2000, 9, 1037.
(L/kg)
(h)
Rat (n = 4)
Dog (4)
Monkey (3)
1.0
0.4
0.4
40 5.8
23 4.5
21 2.5
4.4 1.5
3.8 0.3
7.4 0.8
1.4 0.4
2.3 0.3
4.5 0.8
the S stereochemical configuration of the a-amino
amides for optimal potency.
1
9. All compounds were characterized by H NMR and high
resolution mass spectrometry. For detailed experimental
procedures, see WO03105855.
The pharmacokinetic profile of compound 19 in rat, dog,
and monkey following iv administration is presented in
Table 3 and was representative of the series. Plasma clear-
ance was moderate in these species with half-lives ranging
from 1 to 4 h, respectively.
10. KSP inhibitory activity was assessed by measuring the
release of inorganic phosphate from ATP hydrolysis
through absorbance detection of a quinaldine red-phos-
phate complex. For assay details, see: Cogan, E. B.; Birell,
G. B.; Griffith, O. H. Anal. Biochem. 1999, 271, 29, IC50
values are reported as the averages of at least two
independent determinations; standard deviations are with-
in 25-50% of IC50 values.
In summary, we have described the development of
dihydropyrroles as novel and potent KSP inhibitors
through core modification of dihydropyrazole-based
lead structures. Potency and aqueous solubility were
gained simultaneously through the introduction of basic
amides and ureas, a key finding from our optimization
efforts. In addition, it was discovered that the moderate
level of hERG binding demonstrated by the basic inhib-
itors could be attenuated through the use of charge-
neutral acyl groups.
11. The KSP IC50 for the (+, R) enantiomer of 6 was
determined to be >50,000 nM.
12. Compound 8 was prepared from (S)-epichlorohydrin by
modification of the method of Sassaman, M. B.; Prakash,
G. K. S.; Olah, G. A. J. Org. Chem. 1990, 55, 2016.
13. Maeda, K.; Yamamoto, Y.; Tomimoto, K.; Mase, T.
Synlett 2001, 1808.
14. Interestingly, use of LiHMDS under the same reaction
conditions produced a 2:1 mixture of enol triflates,
favoring 11.
15. As with the dihydropyrazoles, the 2,5-difluorophenyl ring
was found to be optimal in the dihydropyrrole series (data
not shown).
Acknowledgments
16. Mitotic arrest was measured by assessing the mitosis-
specific phosphorylation of nucleolin using an antibody-
coated, bead-based assay. In this assay, total nucleolin
We thank Dr. Harry Ramjit for high resolution mass
spectral analysis, Carl Homnick for chiral separation
to give 6, and Dr. Laszlo Kiss for cellular electrophysi-
ological evaluation of 19 on hERG.
is captured on
a streptavidin-coated parameagnetic
bead coupled with biotinylated nucleolin monoclonal
IgG1 antibody 4E2 (Research Diagnostics, Inc.). Nucle-
olin phosphorylation is detected by an antibody com-
plex consisting of a phospho-specific nucleolin IgM
monoclonal antibody, TG3 (Applied NeuroSolutions,
References and notes
Inc.) and
a goat anti-mouse IgM labeled with a
1. For reviews, see: (a) Heald, R. Cell 2000, 102, 399; (b)
Sharp, D. J.; Rogers, G. C.; Scholey, J. M. Nature 2000,
407, 41; (c) Endow, S. A. Eur. J. Biochem. 1999, 262, 12.
ruthenium Tris-bipyridyl complex (BV-TAG Technolo-
gy, BioVeris Corp.). The complex is quantitated via

M. E. Fraley et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1775–1779
1779
electrochemiluminescence based on the excitation/emis-
sion properties of the Tris-bipyridyl complex using a
BioVeris M-series analyzer. EC50 for KSP-induced
nucleolin phosphorylation was determined after treat-
ment with a compound in a 7-point, half-log dilution
series for 16 h. The values of EC50 are reported as the
average of at least two independent determinations;
by radioligand binding competition experiments using
membrane preparations from human embryonic kidney
cells that stably express hERG. For assay details, see:
Bilodeau, M. T. et al. J. Med. Chem. 2004, 47, 6363, and
references therein.
18. Roden, D. M. N. Engl. J. Med. 2004, 350, 1013.
19. PDB deposition numbers for 6, 13, and 19: 2FL6, 2FKY,
and 2FL2, respectively.
20. Hydrophobic grids were calculated on an SGI and
generated using a Morse potential and a Lennard-Jones
repulsive VdW.
standard deviations are within
values.
25–50% of EC50
17. The hERG IC50 values are reported as the average of at
least two independent determinations and were acquired