Inhibitors of human mitotic kinesin Eg5: Characterization of the 4-phenyl-tetrahydroisoquinoline lead series

Article abstract of DOI:10.1016/j.bmcl.2006.01.056

In a high-throughput screening effort, a series of tetrahydroisoquinolines was identified as modest inhibitors of human Eg5. A medicinal chemistry optimization effort led to the identification of R-4-(3-hydroxyphenyl)-N,N-7,8- tetramethyl-3,4-dihydroisoqu

Full text of DOI:10.1016/j.bmcl.2006.01.056

Bioorganic & Medicinal Chemistry Letters 16 (2006) 2095–2100
Inhibitors of human mitotic kinesin Eg5: Characterization of the
4-phenyl-tetrahydroisoquinoline lead series
Christine M. Tarby,* Robert F. Kaltenbach, III, Tram Huynh, Andrew Pudzianowski,
Henry Shen, Marie Ortega-Nanos, Steven Sheriff, John A. Newitt, Patricia A. McDonnell,
Neil Burford, Craig R. Fairchild, Wayne Vaccaro, Zhong Chen, Robert M. Borzilleri,
Joseph Naglich, Louis J. Lombardo, Marco Gottardis,
George L. Trainor and Deborah L. Roussell
Bristol-Myers Squibb Company, Route 206 and Province Line Road, Princeton, NJ 08543, USA
Received 14 December 2005; revised 13 January 2006; accepted 17 January 2006
Available online 3 February 2006
Abstract—In a high-throughput screening effort, a series of tetrahydroisoquinolines was identified as modest inhibitors of human
Eg5. A medicinal chemistry optimization effort led to the identification of R-4-(3-hydroxyphenyl)-N,N-7,8-tetramethyl-3,4-dihydro-
isoquinoline-2(1H)-carboxamide (32a) as a potent inhibitor of human Eg5 (ATPase IC50 104 nM) with good anti-proliferative activ-
ity in A2780 cells (IC50 234 nM).
ꢀ 2006 Elsevier Ltd. All rights reserved.
Compounds that target mitosis (e.g., taxanes, vinca
alkaloids)1 have found clinical utility as chemotherapeu-
tic agents. While effective against many cancer types,
these agents can be compromised by acquired resistance
and undesirable toxicities (neuropathy, neutropenia,
and GI toxicity) linked to mechanism of action. As such,
a need exists for small molecule agents which target
mitosis by novel mechanisms as alternatives to the exist-
ing therapies.
treatment with reported human Eg5 inhibitors monas-
trol (1)6 and HR22C16 (2a)7 (Fig. 1) has been shown
to induce cell death via apoptosis.8
Since agents which target Eg5 selectively would be
expected to act only on cells undergoing cell division,
we9 and others10,11 have undertaken efforts to identify
small molecule inhibitors of this kinesin.
In a high-throughput screening effort, we identified a
series of tetrahydroisoquinolines as inhibitors of human
Kinesins are a superfamily of motor proteins which
hydrolyze ATP in order to transport cargo across micro-
tubule networks. The microtubule associated kinesin
Eg5 (also known as kinesin spindle protein (KSP)) is re-
quired for assembly and function of the mitotic spindle
apparatus.2 More specifically, Eg5 functions to bundle
microtubules to form the bipolar mitotic spindle.2–4
Modulation of Eg5 activity, by immunodepletion of
Eg5 protein or knockdown of Eg5 mRNA by siRNA,
has been shown to cause aberrant mitotic spindle forma-
tion, cell cycle arrest in mitosis, and inhibition of prolif-
eration of tumor cells in culture.4,5 Additionally,
OH
O
OH
O
N
R
H
N
HN
S
OEt
CH3
N
N
H
H
O
1
2a
Monastrol
R=CH3, HR22C16
N
CH
O
3
Keywords: Eg5 inhibitor; Kinesin inhibitor; Kinesin spindle protein;
KSP; Eg5.
3
*
Figure 1. Eg5 inhibitors: Monastrol (1), HR22C16 (2a), and tetrahy-
droisoquinoline screening hit (3).
Corresponding author. Tel.: +1 609 252 4669; fax: +1 609 252
7410; e-mail: christine.tarby@bms.com
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.01.056

2096
C. M. Tarby et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2095–2100
R2
Eg5. A representative member of the series, compound 3
(Fig. 1), displayed micromolar activity in both the Eg5
ATPase assay (IC50 9.7 lM)12 and an anti-proliferation
assay in A2780 ovarian carcinoma cells (IC50 2.4 lM).13
Thus, a medicinal chemistry program was initiated to
further optimize the screening hit.
HO
+
a
CHO
NH2
OH
NH
R1
H2N
R1
R2
R2
1
NMR experiments including H/15N chemical shift per-
4
c, d
turbation,14 saturation transfer difference (STD),15 and
T2 relaxation studies16 were performed on HTS com-
pounds of interest to verify direct binding to Eg5. Com-
petition experiments using STD with a known allosteric
inhibitor demonstrated that compound 3 bound to the
human Eg5 motor domain in an allosteric site previous-
ly described for other Eg5 inhibitors11,17 and not present
in the apo structure of the enzyme. Guided by co-crystal
structures of Monastrol17 (1) and HR22C16 analog (2b,
R = NH2),18 with the Eg5 motor domain, a binding
model of compound 3 was constructed. The van der
Waals affinity of the 2b allosteric site for a methyl probe
O
X
+
b
R1
R2
NH
R1
5
Scheme 1. Synthesis of the tetrahydroisoquinoline scaffold. Reagents
and conditions: (a) NaBH4, Et3N, MeOH, 0–25 ꢁC; (b) CH3SO3H,
CH2Cl2, 0–25 ꢁC; (c) 1 M K2CO3, EtOAc; (d) NaBH4, MeOH, 0–
25 ꢁC.
was mapped19 at a grid spacing of 0.5 A. Compound 3
˚
was docked into the site with the aid of this map, and
the entire Eg5/ligand complex was allowed to relax
through 1000 steps of energy minimization with the Tri-
pos force field.20 The resulting model and contour were
used to guide SAR development (Fig. 2).
R2
R2
R3
The tetrahydroisoquinoline (THIQ) scaffold was con-
structed from readily available starting materials using
the synthetic route outlined in Scheme 1. Reductive ami-
nation of substituted benzaldehydes with ethanolamines
gave the benzyl-substituted amino-alcohols 4. Acid cat-
alyzed cyclization of 4 with methanesulfonic acid in
CH2Cl2 provided the THIQ core 5 in excellent yields.
Alternatively, the cyclization precursor 4 could be ob-
tained through NaBH4-mediated reduction of the prod-
uct from the alkylation of benzylamines and a-
haloacetophenones. This latter sequence typically pro-
vided the THIQ core 5 in moderate yields.
R2
N
SO2R3
N
R1
R1
a
c
8
6
NH
d
b
R1
R2
R2
5
N
NR3R4
N
R3
R1
R1
O
O
9
7
Scheme 2. Reagents and conditions: (a) R3X, K2CO3, CH3CN, 70 ꢁC;
(b) R3COCl, 1 M K2CO3, EtOAc or R3CO2H, EDCI-HOBT, DMF;
(c) R3SO2Cl, 1 M K2CO3, EtOAc; (d) R3NCO or R3R4NCOCl, 1 M
K2CO3, EtOAc.
Elaboration of the intermediate N-unsubstituted THIQ
core is illustrated in Scheme 2. Alkylation with alkyl ha-
lides in the presence of potassium carbonate at 70 ꢁC
gave the N-alkyl analogs 6. The acylated analogs 7 were
prepared either by addition of an acid chloride to a two
phase mixture of 5 in 1 M K2CO3/EtOAc (1:1) or
through the coupling of the carboxylic acid with
EDCI/HOBT. The sulfonated analogs 8 and the ureas
9 were prepared from the corresponding sulfonyl chlo-
rides, isocyanates, or carbamoyl chlorides using stan-
dard procedures.
An initial scan of R1 substitutions on the THIQ aryl ring
yielded the 7,8-dimethyl groups as suitable replacements
for the dihydrofuran ring of 3 (data not shown). Using
the 7,8-dimethyl THIQ (10), the nitrogen substituent
(R3) was optimized (Table 1). Incorporation of small al-
kyl groups on nitrogen provided compounds with
Figure 2. Compound 3 modeled into the human Eg5 motor domain:2b
(magenta) crystal structure. MOL2MAP methyl probe affinity surface
(van der Waals energy) is shown in yellow, contoured at ꢀ1.5 kcal/
mol.

C. M. Tarby et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2095–2100
2097
Table 1. 4-Phenyl-tetrahydroisoquinoline: optimization of the N-
substituent
Table 2. 4-Phenyl-tetrahydroisoquinoline: effect of substituents on the
4-phenyl ring
R2
N
N
N
H3C
R3
H3C
CH3
CH3
O
Compound R3
ATPase
a
A2780
IC50 (lM) cytotoxicity
a
a
Compound
R2
H
ATPase IC50
(lM)
Cytotoxicity IC50
(lM)
a
IC50 (lM)
28
32
33
34
35
2.75
1.20
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
–H
–CH3
–Et
8.33
18.4
25.2
11
9.04
7.30
5.30
6.28
ndb
2.31
2.42
2.37
nd
3-OH
4-OH
3-CF3
3-Cl
0.306 (0.080)
>25
13.2
0.376 (0.022)
ndb
nd
–iPr
–(CH2)2Ph
3.48
3.02
>25
2.27
3.06
4.65
>25
8.88
5.36
17.9
16.4
>25
14.4
>25
>25
>25
2.75
22.3
12.8
a IC50 values were derived from dose–response curves generated from
triplicate data points. Values with standard deviations (in parenthe-
ses) are means of three independent experiments.
b nd, not determined.
–(CH2)2NH2
–(CH2)3NH2
–(CH2)4NH2
–(CH2)2N-pyrrole
–(CH2)3NMe2
–(CH2)4NMe2
–COMe
3.24
1.86
nd
located in the pocket predicted to accommodate the 4-
phenyl group. Both monastrol and HR22C16 analog
2a donate a hydrogen bond through a phenol function-
ality to one of these moieties. Incorporation of a 30-hy-
droxy group onto the 4-phenyl ring provided compound
32, which exhibited a 7.7-fold increase in activity in the
ATPase assay and 3-fold increase in anti-proliferative
activity over 28 (Table 2). Moving the phenol substitu-
ent from the 3- to the 4-position (33) eliminated all
detectable activity. Replacing the hydrogen-bond donat-
ing substituent with a 3-chloro or 3-trifluoromethyl pro-
vided compounds with micromolar activity.
–COPh
–SO2Me
7.77
nd
–CO2Me
–CO2Et
–CONH2
nd
nd
nd
nd
–CONHMe
–CON(Me)2
–CO-morpholine
–CO-piperidine
1.20
nd
2.83
1.21
–CON(Me)(CH2)3NMe2 5.44
a IC50 values were derived from dose–response curves generated from
triplicate data points.
b nd, not determined.
With the 30-hydroxy group incorporated, we turned our
attention to optimization of the substituents on the
THIQ aryl ring (R1, Table 3). Initial SAR exploration
of simple substituents at all four positions on the aryl
ring quickly indicated a requirement for a substituent
at the C7 or C8 position. The unsubstituted THIQ 36
and the C5 or C6 substituted compounds (data not
shown) have IC50s greater than 25 lM. Electron-with-
drawing halogen groups incorporated at C7 provided
compounds with modest to no activity in the ATPase as-
say (37, 38), however, as the size of the C7 substituent
increased and the electron-withdrawing capacity de-
creased, submicromolar activity was attained (39, 40).
Incorporation of both the 7-methyl and 7-methoxy com-
pounds also provided compounds with potent submicro-
molar activities (41, 42). At C8, halogens were not
tolerated and substitution of a methyl (46) or methoxy
(48) provided compound with micromolar activity.
However, the 8-ethyl compound 47 exhibited equipotent
activity to 32 in both assays. Substitution with groups at
both the 7- and 8-positions provided mixed results.
While 7,8-dimethyl compound 32 exhibited improved
activity in the ATPase and antiproliferation assays over
the singly substituted compounds (41, 46), preparation
of the 7,8-dimethoxy compound 49 resulted in substan-
tial loss of activity over the singly substituted com-
pounds (42, 48). Submicromolar activities were
observed for compounds with a ring appended in the
modest activity in the human Eg5 ATPase assay (11-13),
while use of larger N-alkyl substituents (>4C; e.g., 14,
18) led to loss of activity. However, when the N-alkyl
chain contained a basic nitrogen 2–4 carbons away from
the THIQ core (15–17, 19–20), activity was retained or
improved over the parent compound in both the ATP-
ase and antiproliferation assays. From our computer
model, we hypothesized that the charged side chain
extended toward solvent.
Acylation of the THIQ nitrogen to provide amides, sul-
fonamindes, carbamates or primary ureas afforded com-
pounds with little to no activity. However, the N,N-
dimethyl urea, 28, was equipotent with the compounds
containing a basic amine in the side chain (15–17). Larg-
er disubstituted ureas (29, 30) were less active than 28.
Combination of the disubstituted urea functionality
with the basic amine containing side chain also provided
micromolar compounds (e.g., 31). Since we expected
that compounds containing a basic nitrogen would have
a greater potential for off-target activities, we chose to
further explore SAR development in the N,N-dimethyl
urea series (Table 2).
Our binding model of compound 3 with Eg5 indicated
two carbonyls, from residues Trp127 and Glu118, were

2098
C. M. Tarby et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2095–2100
Table 3. 4-Phenyl-tetrahydroisoquinoline: effect of substitution on the
tertrahydroisoquinoline aryl ring
OH
5
6
N
N
R1
8
O
R1
ATPase IC50
(lM)
A2780
Cytotoxicity
a
Compound
a
IC50 (lM)
36
37
38
39
40
41
42
43
44
45
46
47
48
32
49
50
51
52
H
7-F
>25
>25
nd
ndb
1.19
7-Cl
2.49
7-Br
7-I
0.808 (0.089)
0.317 (0.080)
0.595
0.879 (0.143)
0.363 (0.015)
0.851
7-Me
7-OMe
8-F
Figure 4. The binding site for compound 32a on human Eg5. The
compound is shown with thick rods and the protein is shown with
thinner rods. The initial 2Fo ꢀ Fc electron density map contoured at 1r
is shown in cyan mesh. Figure created using PyMol.24
0.617 (0.019)
>25
>25
0.402 (0.032)
nd
nd
8-Cl
8-Br
8-Me
8-Et
>25
4.71
nd
2.05
0.344
1.78
0.186
0.332
ture of the protein.23 The binding site for compound 32a
is shown in Figure 4. The phenol moiety of compound
32a binds in a hydrophobic sub-pocket surrounded by
side chains of Arg 119, Trp 127, Ala 133, Pro 137, Tyr
211, and Leu 214 with additional van der Waals interac-
tions contributed by the main chain of Gly 117 and Glu
118, and the side chain of Asp 130. Additionally, the
phenol forms hydrogen bonds to the Glu 118 carbonyl
oxygen and the Arg 119 side chain. The 7,8-dimethyl
substituted aryl ring of the THIQ is sandwiched between
the side chains of Glu 116 and Leu 214 with additional
van der Waals interactions with the side chains of Ile
136, Pro 137, Leu 160, Ala 218, and Phe 239, and the
main chain of Glu 116. The piperidine moiety of the
THIQ forms van der Waals contacts with the side chain
of Ala 218 and the carbonyl oxygens of Glu 116 and Gly
117. The dimethyl urea moiety, which is pointed out of
the binding pocket, forms van der Waals contacts with
both the main and side chains of Leu 214 and Glu
215, and the side chains of Tyr 211, Arg 119, and Ala
218.
8-OMe
7,8-Me2
0.306 (0.080)
12.8
1.27
0.376 (0.022)
2.92
3.44
7,8-(OMe)2
7,8-Cl2
7,8-naphthyl
0.477
0.639
0.373
0.235
7,8-(–(CH2)2O–)
a IC50 values were derived from dose–response curves generated from
triplicate data points. Values with standard deviations (in parenthe-
ses) are means of three independent experiments.
b nd, not determined.
7,8-position (51, 52). In general, close correlation be-
tween the data obtained in the ATPase and cell prolifer-
ation experiments was observed across all program
compounds.
Compound 32 was resolved into its enantiomers and the
active enantiomer 32a (Fig. 3, R-isomer21) was used for
co-crystallography experiments and further biological
evaluation.
The X-ray crystal structure of 32a in complex with Mg2+
Development of structure activity relationships during
optimization of the THIQ series of Eg5 inhibitors dem-
onstrated a correlation between the microtubule-depen-
dent Human Eg5 ATPase activity and A2780 ovarian
carcinoma cell proliferation. In order to confirm that
the cytotoxicity displayed by the THIQ series was a con-
sequence of Eg5 inhibition, we further characterized the
cellular activity of compound 32a.
22
˚
and ADP was determined to 2.3 A resolution. As pre-
dicted from our modeling studies, compound 32a binds
in the known allosteric pocket of Eg5.9,17 This allosteric
˚
binding site of 32a is approximately 13 A from the
nucleotide binding site and causes rearrangement of
backbone and side chains relative to the Mg-ADP struc-
OH
Cells with reduced Eg5 activity, for example by immun-
odepletion or small molecule inhibitors, display a char-
acteristic mitotic cell phenotype and arrest in
mitosis.4,6 As shown in the confocal fluorescent micros-
copy images of Figure 5, compound 32a induced the for-
mation of monopolar mitotic spindles in kangaroo rat
kidney epithelial cells (PtK2) after 4 h treatment.25 Mul-
tiple monopolar mitotic spindles but no bipolar mitotic
spindles were observed (Fig. 5A). Under higher magnifi-
N
N
O
32a
ATPase IC 50 0.104 (0.044) µM
A2780 Cytotoxicity IC50 0.234 (0.036) µM
Figure 3. Compound 32a.

C. M. Tarby et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2095–2100
2099
B
A
Difference Mean Log10(IC
)
50
3
2
1
0
-1
-2
-3
Cell Line
IC50 (μM)
MDA-MB-468
NCI-H446
PC-3
SHP-77
LNCaP
HCT116
MDA-MB-231
A2780/S
Du145
CCRF-CEM
A2780/TAX-R
0.350
0.360
0.223
0.261
0.807
0.259
0.480
0.267
0.484
0.282
0.215
C
M109
P388
SW620
0.259
0.460
0.466
Mean IC50 (μM)
Median IC50 (μM)
Min IC50 (μM)
Max IC50 (μM)
Range (Log Conc) 0.575
0.334
0.282
0.215
0.807
Figure 5. Confocal fluorescent microscopy images of PtK2 following
treatment with 8.6 lM compound 32a for 4 h. Monopolar mitotic
spindles (arrowheads) in a representative field of cells at increasing
magnification (10·, 63·, and 63· Zoom in panels A, B, and C,
respectively), microtubules (green), and chromosomal DNA (blue).
Figure 6. Compound 32a Eg5 inhibitor 32a demonstrated cytotoxic
activity against a panel of tumor cell lines.27 IC50s were determined
after 72 h exposure. The mean bar graph depicts the difference between
the log of individual cell line IC50 values relative to the mean log of all
the IC50 values. Relative sensitivities are shown; right-projecting bars
are more sensitive cell lines and left-projecting bars are less sensitive
cell lines.
cation (Fig. 5C), several DAPI-stained chromosomes
(blue) are detected in the focal plane with microtubules
of the mitotic spindle (green).
Compound 32a also arrested cells in mitosis. A2780
ovarian carcinoma cells treated with compound 32a
for 17 h, corresponding to approximately one cell cycle,
showed an increase in the mitotic index (percentage of
cells in M phase) relative to untreated cells, as quantified
by flow cytometry.13 The induction of mitotic arrest at
17 h is in agreement with the cytotoxicity observed in
A2780 cells at 72 h (Table 4). Further evaluation of
the cytotoxic activity of compound 32a revealed anti-
proliferative activity against multiple tumor cell lines
(breast, colorectal, lung, ovarian, and prostate). Cyto-
toxicity IC50 values in the different cell lines varied less
than 0.6 log of the mean, as graphically depicted in
the mean bar graph (Fig. 6). As expected for an inhibitor
of the essential protein Eg5, compound 32a was equipo-
tent in all cell lines tested, including both A2780S
taxol-sensitive (IC50 0.267 lM) and A2780-TAXR tax-
ol-resistant (IC50 0.215 lM) cell lines.26
Modeling-guided optimization of the tetrahydroisoquin-
oline screening hit 3 led to the identification of com-
pound 32a, a sub-micromolar human Eg5 inhibitor.
Biological characterization demonstrated that com-
pound 32a displays cellular activities in secondary as-
says consistent with an Eg5 mechanism of action.
Compounds from the tetrahydroisoquinoline series are
novel human Eg5 inhibitors with promising preclinical
activity profiles and represent significant new leads for
further optimization.
Table 4. Cell cycle analysis
Compound
Concentration (lM)
A2780 cell cycle profile
% S Phase
Mitotic index
% M Phase
% G1 Phase
% G2 + M Phase
32a
32a
0.257a
0.328b
0.08%c
0.1d
9.1
3.1
11.8
5.9
73.2
84
39.3
52.2
3.2
DMSO
Nocodazole
41.4
12
37.5
15.7
14.8
62.1
23.7
Experimental details described in Ref. 13.
a A2780 IC50
.
b A2780 IC90
.
c neg. control.
d pos. control.

2100
C. M. Tarby et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2095–2100
J. Med. Chem. 2002, 45, 3905(b) Roussell, D. et al.
(Manuscript in preparation).
Acknowledgments
14. Shuker, S. B.; Hajduk, P. J.; Meadows, R. P.; Fesik, S. W.
Science 1996, 274, 1531.
15. Mayer, M.; Meyer, B. Angew. Chem. Int. Ed. 1999, 38,
1784.
16. Navon, G.; Lanier, A. J. Magn. Reson. 1972, 8, 14.
17. Yan, Y. S. V.; Xu, B.; Homnick, C.; Halczenko, W.;
Buser, C. A.; Schaber, M.; Hartman, G. D.; Huber, H. E.;
Kuo, L. C. J. Mol. Biol. 2004, 335, 547.
The authors acknowledge Luciano Mueller for his assis-
tance with the NMR experiments; Timothy Spicer and
Jonathan O’Connell for high-throughput screening
experiments; Gretchen Schroeder for her assistance
examining the HTS data; Atsu Apedo for the CD–UV
spectrum of compound 32a; Rolf Rysek for cloning Eg5
DNA sequence; Gerald Duke for the preparation of
Eg5 protein used in the Eg5 enzymatic assays; Russell
Peterson and Kathy Johnston for cell panel cytotoxicity
studies; Anne Lewin for the confocal microscopy images;
and Mike Poss for program discussions around the lead
series.
18. Newitt, J. Sheriff, S. Unpublished results.
19. Goodford, P. J. J. Med. Chem. 1985, 28, 849.
20. Clark, M. C. R. D., III; Van Opdenbosch, N. J. Comp.
Chem. 1989, 10, 982.
21. R-configuration determined through CD–UV and com-
putational experiments. Structural assignments subse-
quently confirmed by co-crystallography in Eg5 binding
pocket.
References and notes
22. (a) PDB Deposition number is 2FME.; (b) Newitt, J. et al.
(Manuscript in preparation).
1. Dutcher, J. P.; Novik, Y.; O’Boyle, K.; Marcoullis, G.;
Secco, C., et al. J. Clin. Pharmacol. 2000, 40, 1079.
2. Sawin, K. E.; LeGuellec, K.; Philippe, M.; Mitchison, T. J.
Nature 1992, 359, 540.
3. Walczak, C. E.; Vernos, I.; Mitchison, T. J.; Karsenti, E.;
Heald, R. Curr. Biol. 1998, 8, 903.
23. Turner, J.; Anderson, R.; Guo, J.; Beraud, C.; Fletterick,
R., et al. J. Biol. Chem. 2001, 276, 25496.
24. DeLano, W. L. The PyMol Molecular graphics System
(2002). DeLano Scientific, San Carlos, CA, USA. http:/
www.pymol.org.
25. PtK2 cells were grown in minimal essential medium
(MEM) supplemented with 10% fetal calf serum in an
incubator at 37 ꢁC, 5% CO2. Cells were plated onto
chamber slides at a density of 10,000 cells/cm2. Treatment
was initiated 24 h after plating with the addition of
compound. After 4 h of treatment, compound was
removed , methanol was added for 2 min to fix and
permealize cells, and 3% bovine serum albumin (BSA) in
phosphate-buffered saline (PBS) was added to block.
DNA was stained with 50 nM DAPI and microtubules
were stained with Monoclonal Anti-tubulin (Sigma) fol-
lowed by FITC-conjugated goat anti-mouse (ICN/Cappell
Fluorescein conjugate F(AB0)2FRAGMENT to Mouse
IGG whole molecule).
4. Blangy, A.; Lane, H. A.; d’Herin, P.; Harper, M.; Kress,
M., et al. Cell 1995, 83, 1159.
5. Weil, D.; Garcon, L.; Harper, M.; Dumenil, D.; Dautry,
F., et al. Biotechniques 2002, 33, 1244.
6. Mayer, T. U.; Kapoor, T. M.; Haggarty, S. J.; King, R.
W.; Schreiber, S. L., et al. Science 1999, 286, 971.
7. Hotha, S.; Yarrow, J. C.; Yang, J. G.; Garrett, S.;
Renduchintala, K. V.; Mayer, T. U.; Kapoor, T. M.
Angew. Chem. Int. Ed. Engl. 2003, 42, 2379.
8. (a) Marcus, A. I.; Peters, U.; Thomas, S. L.; Garrett, S.;
Zelnak, A., et al. J. Biol. Chem. 2005, 280, 11569; (b)
Leizerman, I.; Avunie-Masala, R.; Elkabets, M.; Fich, A.;
Gheber, L. Cell Mol. Life Sci. 2004, 61, 2060.
9. Kim, K. et al. (Manuscript in preparation).
26. Giannakakou, P.; Sackett, D. L.; Kang, Y. K.; Zhan, Z.;
Buters, J. T., et al. J. Biol. Chem. 1997, 272, 17118.
27. (a) MDA-MB-468: Human breast carcinoma; NCI-H446:
Human small cell lung carcinoma; PC-3: Human prostate
carcinoma; SHP-77: Human small cell lung carcinoma;
LNCaP: Human prostate carcinoma; HCT-116: Human
colon carcinoma; MDA-MB-231: Human breast carci-
noma; A2780S: Human ovarian carcinoma, taxol sensi-
tive; Du145: Human prostate carcinoma; CCRF-CEM:
Human leukemia; A2780/TAX-R: Human ovarian carci-
noma, taxol resistant; M109: Murine lung carcinoma;
P388: Murine leukemia; SW620: Human colorectal carci-
noma; (b) Lee, F. Y. F.; Borzilleri, R.; Fairchild, C.; Kim,
S. H.; Long, B. H.; Raventos-Suarez, C.; Vite, G. D.;
Rose, W. C.; Kramer, R. A. Clin. Cancer Res. 2001, 7,
1429.
10. (a) Sakowicz, R.; Finer, J. T.; Beraud, C.; Crompton, A.;
Lewis, E., et al. Cancer Res. 2004, 64, 3276; (b) Sunder-
Plassmann, N.; Sarli, V.; Gartner, M.; Utz, M.; Seiler, J.,
et al. Bioorg. Med. Chem. 2005, 13, 6094; (c) Bergnes, G.;
Brejc, K.; Belmont, L. Curr. Top. Med. Chem. 2005, 5, 127.
11. Cox, C. D.; Breslin, M. J.; Mariano, B. J.; Coleman, P. J.;
Buser, C. A., et al. Bioorg. Med. Chem. Lett. 2005, 15,
2041.
12. Burford, N. et al. ATPase assays with Eg5 motor domain
protein were performed with Kinase Glo Kit as described
by the manufacturer (Promega) (Manuscript in
preparation).
13. Human tumor cell line maintenance and cell-based assays
were performed as described in (a) Kim, K. S.; Kimball, S.
D.; Misra, R. N.; Rawlins, D. B.; Hunt, J. T., et al.