Kinesin spindle protein (KSP) inhibitors. Part V: Discovery of 2-propylamino-2,4-diaryl-2,5-dihydropyrroles as potent, water-soluble KSP inhibitors, and modulation of their basicity by β-fluorination to overcome cellular efflux by P-glycoprotein

Article abstract of DOI:10.1016/j.bmcl.2007.03.006

Installation of a C2-aminopropyl side chain to the 2,4-diaryl-2,5-dihydropyrrole series of kinesin spindle protein (KSP) inhibitors results in potent, water soluble compounds, but the aminopropyl group induces susceptibility to cellular efflux by P-glycop

Full text of DOI:10.1016/j.bmcl.2007.03.006

Bioorganic & Medicinal Chemistry Letters 17 (2007) 2697–2702
Kinesin spindle protein (KSP) inhibitors. Part V: Discovery
of 2-propylamino-2,4-diaryl-2,5-dihydropyrroles as potent,
water-soluble KSP inhibitors, and modulation of their basicity
by b-fluorination to overcome cellular efflux by P-glycoprotein
Christopher D. Cox,a,* Michael J. Breslin,a David B. Whitman,a Paul J. Coleman,a
Robert M. Garbaccio,a Mark E. Fraley,a Matthew M. Zrada,a Carolyn A. Buser,b
Eileen S. Walsh,b Kelly Hamilton,b Robert B. Lobell,b Weikang Tao,b Marc T. Abrams,b
Vicki J. South,b Hans E. Huber,b Nancy E. Kohlb and George D. Hartmana
aDepartment of Medicinal Chemistry, Merck Research Laboratories, PO Box 4, Sumneytown Pike, West Point, PA 19486, USA
bDepartment of Cancer Research, Merck Research Laboratories, PO Box 4, Sumneytown Pike, West Point, PA 19486, USA
Received 9 February 2007; revised 28 February 2007; accepted 1 March 2007
Available online 6 March 2007
Abstract—Installation of a C2-aminopropyl side chain to the 2,4-diaryl-2,5-dihydropyrrole series of kinesin spindle protein (KSP)
inhibitors results in potent, water soluble compounds, but the aminopropyl group induces susceptibility to cellular efflux by P-gly-
coprotein (Pgp). We show that by carefully modulating the basicity of the amino group by b-fluorination, this series of inhibitors
maintains potency against KSP and has greatly improved efficacy in a Pgp-overexpressing cell line. The discovery that cellular efflux
by Pgp can be overcome by carefully modulating the basicity of an amine may be of general use to medicinal chemists attempting to
transform leading compounds into cancer cell- or CNS-penetrant drugs.
ꢀ 2007 Elsevier Ltd. All rights reserved.
The discovery and widespread use of the taxanes and
vinca alkaloids as effective therapies to combat cancer
has led to the recognition of the mitotic spindle as a
well-validated target for chemotherapeutic agents.1 As
effective as these antimitotic agents have been clinically,
they suffer from a number of liabilities, including: (1) the
development of resistance through both tubulin muta-
tions and P-glycoprotein (Pgp)-mediated cellular efflux;
(2) significant dose-limiting toxicities due to drug action
on post-mitotic neurons; and (3) inconvenient dosing
regimes including prophylactic treatment to prevent
vehicle-associated toxicities.2
required during mitosis. KSP is a microtubule motor
protein that is essential for the generation of a bipolar
mitotic spindle, and its inhibition leads to mitotic arrest
followed by apoptotic cell death.3 The anticipated clini-
cal benefits of KSP inhibitors relative to the taxanes and
vinca alkaloids include: (1) a different resistance profile
with the potential for activity against refractory tumors;
(2) lack of neuronal toxicity, since KSP is not expressed
in post-mitotic neurons; and (3) simplified dosing for-
mulations. Several companies are currently investigating
KSP inhibitors in the clinic to establish if these benefits
will be realized in practice.4
Small molecule inhibitors of kinesin spindle protein
(KSP, or HsEg5) are part of a new generation of antimi-
totic agents that do not interact directly with the mitotic
spindle, but instead target key regulatory enzymes
Dihydropyrazole 1 and dihydropyrrole 2 are small mol-
ecule inhibitors of KSP discovered at Merck that possess
moderate potency, but are not suitable for dosing in an
aqueous vehicle.5 We recently disclosed a structure-
based optimization of 1 that led to the installation of
an aminopropyl side chain to provide dihydropyrazole
3 with improved potency, pharmacokinetics, and good
aqueous solubility.6 In this communication, we describe
how the aminopropyl appendage was incorporated into
Keywords: Pgp; Antimitotics; Multidrug-resistance.
*
Corresponding author. Tel.: +1 215 652 2411; fax: +1 215 652
7310; e-mail: chris_cox@merck.com
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.03.006

2698
C. D. Cox et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2697–2702
the dihydropyrrole series to provide the potent KSP
inhibitor (KSPi) 4, but that this compound, as well as
its dihydropyrazole analog 3, is not potent in a multi-
drug-resistant (MDR) cell line due to efflux by Pgp.
We then go on to illustrate that careful modulation of
the basicity of the amine in 4 by b-fluorination provides
KSP inhibitors such as 18 that are effective in a Pgp-
overexpressing cell line, and have thus overcome cellular
efflux by Pgp.
The triflate of ketone 8 was generated regioselectively
with PhN(Tf)2 and NaHMDS,5b and converted to 9
via Suzuki reaction in 64% yield for two steps. The ox-
azolidinone was hydrolyzed with NaOH and the resulting
amino alcohol was treated with t-butyldimethylsilyl
chloride to provide 10 in nearly quantitative yield. As
noted in our previous reports, only the (S)-antipode of
compounds in this class of inhibitors exhibits KSP activ-
ity, so a chiral HPLC separation was carried out at this
stage to provide optically pure (S)-10. Pretreatment of
this intermediate with triphosgene, followed by the addi-
tion of dimethylamine and subsequent deprotection, led
to 11 in 90% yield from (S)-10.
Design and synthesis. Our previous studies have estab-
lished that the preferred substitution patterns in the
dihydropyrazole and dihydropyrrole series are very sim-
ilar.5,6 Additionally, X-ray crystallographic evidence
indicates that both series bind in a similar mode to the
same allosteric site of KSP, making effective use of both
hydrophobic and hydrophilic interactions. It was there-
fore a logical extension that aminopropyl substitution at
C2 of dihydropyrrole 2 would lead to a KSPi with a sim-
ilar increase in potency as seen in going from 1 to 3, and
thus 4 became the target of our synthetic efforts (Fig. 1).
Evaluation of the literature revealed no general way to
access 2,2-disubstituted-2,5-dihydropyrroles, so we em-
braced the challenge of designing a route to these tar-
gets. As illustrated in Scheme 1, the methodology we
developed to reach 4 has the advantage of allowing
straightforward SAR investigation of the C4-aryl group,
the C2-alkyl substituent, and the acyl group at N1.
Installation of the aminopropyl side chain commenced
with Dess–Martin oxidation of 11 to the corresponding
aldehyde in quantitative yield, followed by Horner–
Wadsworth–Emmons reaction with commercially avail-
able trimethyl phosphonoacetate to provide 12. Since
the nonconjugated double bond in the core of 12 was
reactive to a number of standard hydrogenation condi-
tions, a NiCl2-mediated NaBH4 conjugate reduction of
the a,b-unsaturated ester was carried out to provide
13. Following reduction of the ester to alcohol 14, mes-
ylate formation, azide displacement, and finally Stau-
dinger reduction led to the targeted amine 4.
Table 1 summarizes the key properties of dihydropyr-
azole 3 and the newly synthesized dihydropyrrole 4.
As expected, 4 has excellent potency, both against puri-
fied enzyme and in a cell-based assay, as well as good
solubility in pH 4 water. While 3 displays weak binding
affinity for the hERG potassium channel, 4 shows some-
what greater binding, consistent with earlier reports on
compounds in the dihydropyrrole series.5b Unfortu-
nately, both inhibitors possessed significantly reduced
potency in Pgp-overexpressing cells derived from a
KB-3-1 human epidermoid carcinoma cell line, as mea-
sured by the MDR ratio. This ratio provides a measure
of Pgp-mediated resistance to mitotic arrest, and entails
determining the IC50 for induction of a G2/M block by
the test compound in a cell line that highly overexpresses
Pgp, as well as the IC50 in the parental line that does not
express Pgp. The quotient of these numbers is reported
as the MDR ratio. For instance, a ratio of unity indi-
cates that the KSPi is not effluxed by Pgp because it is
equipotent in both cell lines. Taxol, by comparison,
has an MDR ratio of about 25,000. KSP inhibitors 3
and 4, with MDR ratios of 490 and 1200, respectively,
are therefore subject to Pgp-mediated cellular efflux.7
The benzyl imine of racemic phenylglycine methyl ester
(5) was first alkylated with methallyl dichloride under
phase transfer conditions to give the monoalkylation
product (not shown or isolated). The imine was subse-
quently hydrolyzed with aqueous hydrochloric acid,
and during workup a second, intramolecular alkylation
occurred to furnish pyrrolidine 6. Crude 6 was treated
with LAH to afford an amino alcohol, and reaction with
1,10-carbonyldiimidazole provided oxazolidinone
7
which was treated with ozone to generate bicyclic ketone
8 in an overall yield of 20% from 5.
F
F
F
F
OH
N
O
N
N
CH3
O
NMe2
1 KSP IC50 = 26 nM
2 KSP IC50 = 50 nM
Optimization of Pgp profile. Pgp is a protein encoded in
the human MDR1 gene that has been extensively stud-
ied as a major aspect of the MDR phenotype.8 Since
many cancer cells have been shown to overexpress
Pgp, and it is assumed that Pgp plays a role in clinical
resistance to Taxol and other important antimitotic
agents,9 good levels of potency in a Pgp-overexpressing
cell line was a selection criterion for advancing com-
pounds in our program. Several small molecule inhibi-
tors of KSP have been reported to retain efficacy in
Pgp-overexpressing cell lines,10 and 1 and 2 both have
an MDR ratio of 1. This knowledge, together with the
F
F
F
F
NH2
NH2
N
O
2
5
N
N
NMe2
O
NMe2
3 KSP IC50 = 2 nM
4 KSP IC50 = ???
Figure 1. Dihydropyrazole and dihydropyrrole inhibitors of KSP.

C. D. Cox et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2697–2702
2699
O
d,e
g,h
a,b,c
f
N
N
OMe
N
CO2Me
Ph
N
H
O
O
O
F
7
8
O
5
6
O
F
F
F
F
F
F
F
OH
l,m,n
k
i,j
OTBS
OTBS
N
N
N
N
H
H
O
O
NMe2
O
10
9
(S)-10
11
F
F
F
F
F
F
CO2Me
OH
CO2Me
q
s,t,u
r
o,p
4
N
N
N
O
O
O
NMe2
NMe2
NMe2
12
13
14
Scheme 1. Reagents and conditions: (a) ClCH2C(CH2)CH2Cl, Bu4NHSO4, 10 M NaOH, DCM; (b) 1MHC1; (c) Na2CO3/H2O; (d) LAH, THF, 0 ꢁC
to rt; (e) CDI, TEA, DCM; (f) O3, DCM, ꢀ78 ꢁC, then DMS (ꢁ20% yield from 5, trituration from acetone); (g) NaHMDS, THF, ꢀ78 ꢁC, then
PhN(Tf)2; (h) 2,5-difluoroboronic acid, Na2CO3, LiCl, Pd(PPh3)4, DME (64% for 2 steps); (i) 10 M NaOH, EtOH, 60 ꢁC; (j) TBSC1, imidazole, 5%
DMF in DCM (96% for 2 steps); (k) Chiralpak AD, 1% iPrOH in hexanes (0.1% DEA as modifier), first isomer to elute is the desired (S)-isomer; (1)
triphosgene, TEA, THF; (m) NHMe2; (n) HF-TEA, CH3CN (90% for 3 steps); (o) Dess–Martin Periodinane, DCM (quant); (p) trimethyl
phosphonoacetate, NaH, THF (67%); (q) NiCl2, NaBH4, MeOH, 0 ꢁC; (r) LiBH4, 3:1 THF/MeOH (55% from 12); (s) MsCl, TEA, DCM; (t) NaN3,
DMF, 50 ꢁC; (u) PPh3, THF, 55 ꢁC, then H2O (43% for three steps).
Table 1. Comparison of the key properties of 3 and 4
cantly reduced, eliminating this favorable interaction.
A similar trend has been observed in the dihydropyrrole
F
F
F
F
series. In light of these observations, namely that KSP
potency is improved by a basic amine while the MDR
ratio is improved by a nonbasic amine (or lack of an
amine altogether), we sought to determine if a middle
ground could be established where both features are
maintained in an optimal range.
NH2
NH2
N
O
N
N
O
NMe2
NMe2
3
4
Property
Compound 3
Compound 4
To this end, we explored b-fluorination of the amine in 4
to carefully modulate its basicity in a step-wise manner
while imparting minimal steric effects.12 As revealed in
Table 2, substituting the amine with an ethyl group pro-
vides 15a with a slight reduction in potency relative to 4,
but it remains a good substrate for Pgp efflux. Monoflu-
orination of the ethyl group affords 15b with no loss in
KSP potency relative to 15a, but with a significant
improvement in the MDR ratio. Gratifyingly, difluori-
nation provides 15c that again retains good KSP po-
tency but has an MDR ratio of almost unity. Finally,
the nonbasic trifluoroethyl derivative 15d continues the
trend of reduced MDR ratio, but also suffers a substan-
tial reduction in KSP potency. The final two columns in
Table 2 illustrate the dramatic changes to pKa and logP
that b-fluorination imparts across this series of
inhibitors.
a
KSP IC50 (nM)
1.9 1.2
5.2 0.3
1.2
2.2 1.2
6.0 1.1
1.1
Cell potencya (nM)
logP
Solubilityb (mg/mL)
>12
19.3 0.7
490
>12
7.1 3.7
1200
c
hERG IC50 (lM)
MDR ratiod
a See Ref. 5b for assay details, average of n = 3 or greater.
b Solubility of free-base in 0.1 M citrate, initial pH 4.0.
c See Ref. 5b for assay details, average of n = 2 or greater.
d See text and Ref. 7 for discussion of this assay.
fact that basic amines are a pharmacophore often found
in Pgp substrates,11 implicated the aminopropyl side
chain for engendering Pgp susceptibility in 3 and 4.
Our previous investigation demonstrated that the pro-
tonated form of the basic amine in dihydropyrazoles
such as 3 is engaged in a hydrogen bonding interaction
with the amide backbone in the allosteric site of KSP.6
Consistent with this observation, a decrease in potency
was noted when the basicity of the amine was signifi-
The data presented in Table 2 indicate that when the
amine has a pKa of about 7, KSP activity is maintained
while Pgp-mediated efflux is greatly attenuated.

2700
C. D. Cox et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2697–2702
Table 2. Effect of b-fluorination on key properties of C2-propylamino
dihydropyrroles
F
F
F
F
F
F
CO2Me
O
F
F
a,b
c,d
R
N
H
N
N
O
NMe2
O
NMe2
N
13
16
O
NMe2
15
F
F
F
F
F
F
Compounda
R
H
KSP IC50
(nM)
MDR pKa
ratioc
logP
b
d
F
F
NHBn
NH2
3'
e
4
2.2 1.2
10.2 2.0
1200
>135
32
10.3
1.2
1.6
N
N
15a
15b
15c
15d
CH2CH3
10.7
8.8
7.0
5.2
O
NMe2
O
NMe2
CH2CH2F 10.2 1.9
CH2CHF2 12.1 2.5
2.6
3.4
3
17
18
CH2CF3
110 47
1
>3.2
a The synthesis of analogs 15a-d was performed via Na(OAc)3BH-
mediated reductive amination of the appropriate amine with the
aldehyde generated by Dess–Martin oxidation of 14.
b See Ref. 5b for assay details, average of n = 3 or greater.
c See text and Ref. 7 for discussion of this assay.
d Values were determined with a Sirius GL pKa titrator, average of
n = 3 determinations.
Scheme 2. Reagents and conditions: (a) LDA, F2CHP(O)(OEt)2,
ꢀ78 ꢁC, THF, then 13; (b) NaOMe, MeOH (77% for two steps); (c)
TiCl4, BnNH2, TEA, DCE, stir 24 h, then NaCNBH3 in MeOH; (d)
flash chromatography (SiO2, EtOAc/hexanes), first isomer to elute
(45% from 16); (e) 1,4-cyclohexadiene, 10% Pd/C, AcOH, 50 ꢁC (86%).
F
F
F
KSP IC50 = 5.2 0.8 nM
Cell Potency = 22.5 5.6 nM
logP = 3.2; pKa = 7.0
solubility @ pH 4: > 10 mg/mL
hERG = 15.9 1.7 μM
MDR ratio = 5
F
Extensive literature exists on the rational design of com-
pounds to evade Pgp recognition; however, most pub-
lished examples of altering an amino functionality to
circumvent Pgp efflux do so at the expense of adding
or subtracting hydrogen bond donors or acceptors,
and/or greatly altering the steric environment of the
amine.13 To the best of our knowledge, this is the first
demonstration that b-fluorination of an aliphatic amine
can overcome Pgp-mediated efflux while maintaining
good activity on the target of interest.14
NH2
N
O
NMe2
18
Figure 2. Important properties of 18.
As expected, the MDR ratio of 18 is very similar to that
observed for 15c. It is significant to note that a standard
monolayer transport assay also indicated 18 is much less
susceptible to Pgp efflux than 4. In an LLC-PK1 cell line
transfected with the human MDR1 gene, 18 has a
B ꢀ A/A ꢀ B ratio of 2.5 and a passive permeability of
38 · 10ꢀ6 cm/s. By comparison, 4 has a B ꢀ A/A ꢀ B
ratio of 18.5 and a much lower passive permeability of
4 · 10ꢀ6 cm/s.17 Thus, 18 represents an optimized KSPi
with good potency, aqueous solubility, and promising
activity in an MDR cell line that highly overexpresses
Pgp.18
Though 15c possesses a good balance of properties, a
fivefold reduction in KSP potency relative to 4 was ob-
served in the process of enlarging the substituent on the
amine. Several alternative b-difluorination strategies
were therefore undertaken in an attempt to increase po-
tency, the most successful of which resulted in the inte-
rior a-difluoromethyl substituted primary amine 18.
Scheme 2 illustrates how 13 was transformed into 18,
beginning with the addition of diethyl (difluoromethyl)
phosphonate to produce ketone 16, followed by TiCl4-
mediated reductive amination with benzylamine to gen-
erate 17. Conveniently, the desired 30R-diastereomer
was easily separated from its 30S-isomer at the stage of
17 by flash chromatography.15 Transfer hydrogenation
in acetic acid cleanly unveiled the primary amine 18.
Importantly, we have found that modulating the pKa of
a basic nitrogen to the range of 6.5–8.0 to balance Pgp
efflux potential with KSP potency is a general trend that
holds for several diverse series of KSP inhibitors.19
Adjusting the pKa to approximately physiological pH
insures that significant concentrations of both the ion-
ized and neutral forms are available for pharmacological
activity. This may explain the relationship found herein
between pKa and inhibitor activity, because the ability of
the amine to achieve the ionized form is required for
good KSP potency. On the other hand, increasing lipo-
philicity and favoring the neutral form of the amine by
halogenation increases membrane permeability, and
compounds with very good passive permeability can
override the effect of efflux by the Pgp transporter.20
Figure 2 shows the key properties of 18. Both enzymatic
and cell-based potency have been improved relative to
15c. The logP and pKa are dramatically affected by
b-fluorination, but solubility in pH 4 water remains
excellent. Additionally, the hERG binding of 18 is
reduced relative to the starting point 4. Consistent with
this observation, it has been proposed that decreasing
the pKa of a basic nitrogen disrupts p-cation interactions
in the hERG channel binding site due to a reduction in
the percentage of molecules in the ionized form at
physiological pH.16

C. D. Cox et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2697–2702
2701
G. D. Bioorg. Med. Chem. Lett. 2005, 15, 2041; (b) Fraley,
M. E.; Garbaccio, R. M.; Arrington, K. L.; Hoffman, W.
F.; Tasber, E. S.; Coleman, P. J.; Buser, C. A.; Walsh, E.
S.; Hamilton, K.; Fernandez, C.; Schaber, M. D.; Lobell,
R. B.; Tao, W.; South, V. J.; Yan, Y.; Kuo, L. C.;
Prueksaritanont, T.; Shu, C.; Torrent, M.; Heimbrook, D.
C.; Kohl, N. E.; Huber, H. E.; Hartman, G. D. Bioorg.
Med. Chem. Lett. 2006, 16, 1775; (c) Garbaccio, R. M.;
Fraley, M. E.; Tasber, E. S.; Olson, C. M.; Hoffman, W.
F.; Arrington, K. L.; Torrent, M.; Buser, C. A.; Walsh, E.
S.; Hamilton, K.; Schaber, M. D.; Fernandez, C.; Lobell,
R. B.; Tao, W.; South, V. J.; Yan, Y.; Kuo, L. C.;
Prueksaritanont, T.; Slaughter, D. E.; Shu, C.; Heim-
brook, D. C.; Kohl, N. E.; Huber, H. E.; Hartman, G. D.
Bioorg. Med. Chem. Lett. 2006, 16, 1780.
However, we cannot exclude the possibility that pKa is
simply a surrogate marker for Pgp efflux potential that
is correlated with other physicochemical changes, such
as the increased lipophilicity revealed by logP values,
that occur due to halogenation. While more detailed
studies are required to show how the interaction of
our inhibitors with the Pgp transporter changes as the
basicity of the amine is modified,21 that does not limit
the usefulness of the relationship we have established be-
tween pKa and activity in Pgp-overexpressing cells. Sig-
nificantly, the monolayer transport data for 18
corroborates the information provided by the MDR ra-
tio, and suggest that the correlations established herein
may be of general use to those involved not only in
oncology research, but in CNS drug discovery as well.
6. Cox, C. D.; Torrent, M.; Breslin, M. J.; Mariano, B. J.;
Whitman, D. B.; Coleman, P. J.; Buser, C. A.; Walsh, E.
S.; Hamilton, K.; Schaber, M. D.; Lobell, R. B.; Tao, W.;
South, V. J.; Kohl, N. E.; Yan, Y.; Kuo, L. C.;
Prueksaritanont, T.; Slaughter, D. E.; Li, C.; Mahan, E.;
Lu, B.; Hartman, G. D. Bioorg. Med. Chem. Lett. 2006,
16, 3175.
In conclusion, we described how addition of a propylami-
no side chain to the 2,4-diaryl-2,5-dihydropyrrole series
provided potent, water-soluble inhibitors of KSP.
Additionally, we illustrated that the efficacy of these
compounds against a Pgp over-expressing cell line can
be dramatically increased by carefully tuning the basicity
of the amine by b-fluorination. In a forthcoming paper,
we will demonstrate how we applied these lessons learned
in balancing Pgp efflux potential with KSP potency to
identify an optimal clinical candidate for the treatment
of cancer.
7. Relative to the parental KB-3-1 cells, KB-V-1 cells,
originally derived by culturing KB-3-1 cells in the presence
of the Pgp substrate vinblastine (J. Biol. Chem. 1986, 261,
7762), express >500-fold higher levels of Pgp mRNA and
protein. The compound potency (IC50) for induction of
mitotic arrest was determined by evaluating the levels of
the mitotic marker phospho-nucleolin after a 16-h incu-
bation with the test compound in an 11-point half-log
dilution series. The ratio of IC50 obtained in KB-V-1 cells
vs. that in KB-3-1 cells is defined as the MDR ratio. For
instance, compound 3 has an IC50 of 2159 nM in the KB-
V-1 line and 4.4 nM in the KB-3-1 line. The quotient of
2159/4.4 = 490, the MDR ratio of 3. As a general
guideline, we considered compounds with MDR ratios
<10 to be of interest for their ability to enter and kill cells
that overexpress Pgp. Verapamil, a competitive inhibitor
of Pgp, restores the activity of Taxol and our KSP
inhibitors in the KB-V-1 cell line to nearly that observed in
the parental KB-3-1 line, confirming that Pgp efflux is
responsible for the observed resistance to drug-mediated
mitotic arrest.
Acknowledgments
The authors thank Dr. David Dubost for determining
the solubility of 3 and 4, Dr. Cathy Shu, Ms. Nicole
Pudvah, Ms. Bing Li, and Dr. Masayo Yamazaki for
performing the monolayer efflux assays on 4 and 18,
and Dr. Chuck Ross and Ms. Joan Murphy for high res-
olution mass spectral analyses.
References and notes
8. For reviews of Pgp’s relevance to drug discovery, see: (a)
Hochman, J. H.; Yamazaki, M.; Ohe, T.; Lin, J. H. Curr.
Drug Metab. 2002, 3, 257; (b) Lin, J. H.; Yamazaki, M.
Drug Metab. Rev. 2003, 35, 417.
1. Wood, K. W.; Cornwell, W. D.; Jackson, J. R. Curr. Opin.
Pharmacol. 2001, 370.
2. Dutcher, J. P.; Novik, Y.; O’Boyle, K.; Marcoullis, G.;
Secco, C.; Wiernik, P. H. J. Clin. Pharmacol. 2000, 40,
1079.
9. Szakacs, G.; Paterson, J. K.; Ludwig, J. A.; Booth-
Genthe, C.; Gottesman, M. M. Nat. Rev. Drug Disc 2006,
5, 219.
3. (a) Mayer, T. U.; Kapoor, T. M.; Haggarty, S. J.; King, R.
W.; Schreiber, S. L.; Mitchison, T. J. Science 1999, 286,
971; (b) Sakowicz, R.; Finer, J. T.; Beraud, C.; Crompton,
A.; Lewis, E.; Fritsch, A.; Lee, Y.; Mak, J.; Moody, R.;
Turincio, R.; Chabala, J. C.; Gonzales, P.; Roth, S.;
Weitman, S.; Wood, K. W. Cancer Res. 2004, 64, 3276; (c)
Wood, K. W.; Bergenes, G. Annu. Rep. Med. Chem. 2004,
39, 173; (d) Bergenes, G.; Brejc, K.; Belmont, L. Curr.
Top. Med. Chem. 2005, 5, 127; (e) Jiang, C.; You, Q.; Li,
Z.; Guo, Q. Expert Opin. Ther. Patents 2006, 16, 1517, and
references therein.
10. (a) Marcus, A. I.; Peters, U.; Thomas, S. L.; Garrett, S.;
Zelnak, A.; Kapoor, T. M.; Giannakakou, P. J. Biol.
Chem. 2005, 280, 11569; (b) Peters, T.; Lindenmaier, H.;
Haefeli, W. E.; Weiss, J. Arch. Pharmacol. 2006, 372, 291;
(c) Ref. 3b.
11. (a) Salerno, M.; Przewloka, T.; Fokt, I.; Priebe, W.;
Garnier-Suillerot, A. Biochem. Pharmacol. 2002, 63, 1471;
(b) Kaiser, D.; Smiesko, M.; Kopp, S.; Chiba, P.; Ecker,
G. F. Med. Chem. 2005, 1, 431, and references therein.
12. For a series of articles relating to the use of fluorination in
drug design and development, see: Curr. Top. Med. Chem.
2006, 6.
4. (a) Miglarese, M. R.; Carlson, R. O. Expert Opin. Investig.
´
Drugs 2006, 15, 1411; (b) Sorbera, L. A.; Bolos, N.;
Serradell, N.; Bayes, M. Drugs Future 2006, 31, 778; (c)
Jackson, J. R.; Patrick, D. R.; Dar, M. M.; Huang, P. S.
Nat. Rev. Cancer 2007, 7, 107.
13. For a review on the circumvention of Pgp recognition by
rational design, see: (a) Raub, T. J. Mol. Pharmaceutics
2006, 3, 3; A recent study examined the effect of modifying
the local environment of a basic amine in a cytotoxic agent
to maintain activity in an MDR cell line; see: (b)
Ruchelman, A. L.; Houghton, P. J.; Zhou, N.; Liu, A.;
Liu, L. F.; LaVoie, E. J. J. Med. Chem. 2005, 48, 792.
´
5. (a) 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,

2702
C. D. Cox et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2697–2702
14. Recent publications have utilized b-fluorination of ali-
phatic amines to provide improvements in pharmacoki-
netics (ref a) or selectivity (ref b) while maintaining good
potency against the target of interest; see: (a) van Niel, M.
B.; Collins, I.; Beer, M. S.; Broughton, H. B.; Cheng, S. K.
F.; Goodacre, S. C.; Heald, A.; Locker, K. L.; MacLeod,
A. M.; Morrison, D.; Moyes, C. R.; O’Connor, D. O.;
Pike, A.; Rowley, M.; Russell, M. G. N.; Sohal, B.;
Stanton, J. A.; Thomas, S.; Verrier, H.; Watt, A. P.;
Castro, J. L. J. Med. Chem. 1999, 42, 2087; (b) Grune-
wald, G. L.; Seim, M. R.; Lu, J.; Makboul, M.; Criscione,
K. R. J. Med. Chem. 2006, 49, 2939.
15. The stereochemistry at the 30-center was determined by
carrying both diastereomers from the reductive amination
through the final step of the sequence described in Scheme
2, and studying the more potent diastereomer, 18, by
X-ray crystallography. The structure of KSP-18-ADP was
determined as previously described for other inhibitors in
our program (see Ref. 6 and references therein), and thus
also confirmed the (S)-stereochemistry at C2.
17. Compounds with a B ꢀ A/A ꢀ B ratio less than three are
considered weak Pgp substrates, and thus have the
potential for good brain penetration. For a discussion
on the use of the monolayer transport assay in drug
design, see Ref. 8a.
18. In the KB-3-1 (parental) cell line, 18 has an IC50 = 24 nM,
whereas in the KB-V-1 (Pgp-overexpressing) line the
IC50 = 123 nM. By way of comparison, the values for 4
are 23 nM and 28 lM, respectively.
19. In forthcoming publications we will illustrate that this
trend holds for structurally diverse KSP inhibitors where
basicity is controlled not only by b-fluorination, but by
other means as well.
20. Marbeuf-Gueye, C.; Ettori, D.; Priebe, W.; Kozlowski,
H.; Garnier-Suillerot, A. Biochim. Biophys. Acta 1999,
1450, 374.
21. Several studies have examined the way amine basicity
alters recognition of drugs by the Pgp transporter, see: (a)
Ecker, G.; Huber, M.; Schmid, D.; Chiba, P. Mol.
Pharmacol. 1999, 56, 791; (b) Frezard, F.; Pereira-Maia,
E.; Quidu, P.; Priebe, W.; Garnier-Suillerot, A. Eur. J.
Biochem. 2001, 268, 1561.
16. Jamieson, C.; Moir, E. M.; Rankovic, Z.; Wishart, G.
J. Med. Chem. 2006, 49, 5029.