Novel second generation analogs of eribulin. Part II: Orally available and active against resistant tumors in vivo

Article abstract of DOI:10.1016/j.bmcl.2011.01.097

Eribulin mesylate is a newly approved treatment for locally advanced and metastatic breast cancer. We targeted oral bioavailability and efficacy against multidrug resistant (MDR) tumors for further work. The design, synthesis and evaluation of novel amine-containing analogs of eribulin mesylate are described in this part. Attenuation of basicity of the amino group(s) in the C32 side-chain region led to compounds with low susceptibility to PgP-mediated drug efflux. These compounds were active against MDR tumor cell lines in vitro and in xenograft models in vivo, in addition to being orally bioavailable.

Full text of DOI:10.1016/j.bmcl.2011.01.097

Bioorganic & Medicinal Chemistry Letters 21 (2011) 1634–1638
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Novel second generation analogs of eribulin. Part II: Orally available
and active against resistant tumors in vivo
⇑
Sridhar Narayan , Eric M. Carlson, Hongsheng Cheng, Krista Condon, Hong Du, Sean Eckley, Yongbo Hu,
Yimin Jiang, Vipul Kumar, Bryan M. Lewis, Philip Saxton, Edgar Schuck, Boris M. Seletsky, Karen Tendyke,
Huiming Zhang, Wanjun Zheng, Bruce A. Littlefield , Murray J. Towle, Melvin J. Yu
Eisai Product Creation Systems, Eisai Inc., 4 Corporate Dr., Andover MA 01810, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Eribulin mesylate is a newly approved treatment for locally advanced and metastatic breast cancer. We
targeted oral bioavailability and efficacy against multidrug resistant (MDR) tumors for further work. The
design, synthesis and evaluation of novel amine-containing analogs of eribulin mesylate are described in
this part. Attenuation of basicity of the amino group(s) in the C32 side-chain region led to compounds
with low susceptibility to PgP-mediated drug efflux. These compounds were active against MDR tumor
cell lines in vitro and in xenograft models in vivo, in addition to being orally bioavailable.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 16 December 2010
Accepted 20 January 2011
Available online 25 January 2011
Keywords:
Antitumor agents
Microtubule
Multi-drug resistance
P-Glycoprotein
Oral bioavailability
Xenograft models
Eribulin mesylate (Halaven™, Fig. 1) is a novel anti-tubulin
agent under development by Eisai as a treatment for breast cancer
MeO
OH
1
,2
H
2
N
H
O
and other solid tumors. A recent Phase 3 study showed that erib-
ulin significantly improved overall survival in patients with locally
advanced and metastatic breast cancer.3 Based on these results it
recently won approval in the United States, while it is currently
being reviewed by regulatory agencies in Europe and Japan. By vir-
tue of its unique mechanism of action targeting microtubule
O
O
H
O
MsOH
O
O
O
O
O
4
–8
dynamics as compared to taxanes and the vinca alkaloids,
as
well as its wide therapeutic window in preclinical studies, eribulin
represents a significant advance in the arena of chemotherapeutic
agents. The current studies represent the second of a series of
Figure 1. Eribulin mesylate (Halaven™).
9
three articles describing our continuing work on this class of mol-
ecules, in particular focusing on development of compounds with
oral bioavailability, brain penetration and activity against multi-
drug resistant (MDR) tumors.
this Letter, we describe our efforts aimed at optimizing these two
properties within the eribulin core structure. P-Glycoprotein (P-
gp)-mediated drug efflux is one of the primary mechanisms of drug
1
0,11
resistance in a number of cancer cell types.
Moreover, the role
A preclinical profile comprising of activity against a variety of
MDR cell lines was selected based on our expectation that such
properties could be expected to lead to better clinical outcomes.
In addition, although many chemotherapeutic drugs including erib-
ulin are routinely administered as intravenous infusions, an orally
available compound may be desirable in certain circumstances. In
of P-gp in limiting the intestinal absorption of drugs is also well
1
2–14
established.
Therefore, we selected amelioration of cellular ef-
flux by P-gp as a key element of our chosen preclinical profile.
We hypothesized that the basic primary amine moiety in the C32
side chain of eribulin played a key role in its P-gp liability, and thus
focused on modifying this part of the molecule in our second gener-
ation work. In part I, we described the replacement of the C32 amino
alcohol side chain of eribulin with functionalities that are neutral at
physiologic pH. Since amine basicity is often correlated with P-gp
⇑
Corresponding author. Tel.: +1 978 837 4743; fax: +1 978 837 4863.
E-mail address: Sridhar_Narayan@eisai.com (S. Narayan).
1
5,16
susceptibility,
we also designed analogs that would still contain
Present address: Harvard Medical School, BCMP, 240 Longwood Ave., Boston, MA
02115, USA.
an amine, but with attenuated basicity of the nitrogen substituent.
0
960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.01.097

S. Narayan et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1634–1638
1635
b
R1
OHC
I
N
R
2
a
2
3
MeO
OH
R1
N
R1
N
OH
O
e
32
R2
R2
II
HO
H
O
O
O
H
O
c, d
f - h
O
3
5
4
5
O
O
O
OMe
R1 OMe
N
O
i, j
HO
III
1
R2
6
7
k
b
OMe
IV
OHC
8
1
2
Figure 2. General methods for the synthesis of C34/C35 amine containing analogs. Reagents: (a) NaIO
4
, THFÁH
2
O, 23 °C; (b) R R NH, Na(AcO)
3
BH, AcOH, ClCH
2
CH
2
Cl, 23 °C;
2
O, 40 °C;
1
2
(
c) Ts
h) Bu
3 °C.
2
O, DMAP, CH
2
Cl
2 3
, 0–23 °C; (d) R R NH, Et N, MeOH; 23–60 °C; (e) (COCl)
2 3
, DMSO, Et N, CH
Cl
2 2
, À78 °C; (f) TBDPS-Cl, imidazole, DMF, 0 °C; (g) MeI, Ag
2
O, Et
1
2
(
4 2 3
NF-imidazole.HCl, THF, 23 °C; (i) MeSO Cl, Et N, CH
2
Cl
2
, 0 °C or TsCl, pyridine, CH
2
2
Cl , 23 °C; (j) R R NH, Et
3
N, CH
3
2 2
CN, 60 °C; (k) Dess–Martin periodinane, CH Cl ,
2
OH
F
F
N
O
N
N
H
H N
H N
2
2
eribulin
9
10
11
12
O
N
N
O
OH
N
EtO
N
OMe
N
F3C
N
1
1
3
8
14
15
16
17
Me
N
Me
N
O
O
Me OMe
N
OMe
OMe
OMe
N
N
19
20
21
22
Me
Ts
O
OMe
OMe
OMe
N
OMe
N
OMe
N
N
N
N
2
3
24
25
26
27
Me OMe
N
F3C
O
H
OMe
Me OMe
N
OMe
N
OMe
N
N
O
N
N
2
8
29
30
31
32
Figure 3. Partial structures (C32 side-chain) of eribulin analogs synthesized in this study.
We report herein a series of eribulin analogs that have low PgP sus-
ceptibility (FR 6 25), and show low- to sub-nM antiproliferative po-
tency against resistant tumor cell lines in vitro. These compounds,
containing an amine functionality at either C34 or C35, were synthe-
sized starting from a common intermediate. Further exploration led
to the identification of compounds that show in vivo efficacy against
MDR tumors in xenograft models. Oral efficacy was also demon-
strated with some of these compounds.
The compounds in this study were synthesized from previously
1
7
described diol 1 employing one of four general routes as summa-

1
636
S. Narayan et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1634–1638
Table 1
rized in Figure 2. C34 amines were synthesized by oxidative
cleavage of 1, followed by reductive amination of the resulting
aldehyde 2 (route I). C35 amino derivatives were obtained by alkyl-
ation of the appropriate amine with the 1° mono-tosylate deriva-
In vitro antiproliferative potency against four human cancer cell lines, and calculated
fold-resistance ratio (FR) for compounds synthesized in this study
Compd
MES-SA
IC50
MES-SA/Dx5-
Rx1
IC50 (nM)
FRa
DLD-1
IC50
(nM)
HCT-15
IC50
(nM)
tive of
1 (route II). C35 amines bearing a C34 methoxy
(nM)
substituent were accessed from alcohol 6, which in turn was pre-
pared from 1 in a three-step sequence comprising TBDPS protec-
tion of the primary alcohol, methylation at the C34 hydroxyl, and
TBDPS deprotection (route III). In the case of certain 2° amines
which proved recalcitrant to alkylation, route IV involving reduc-
tive amination of methoxy aldehyde 8 was employed. These trans-
formations provided access to a variety of analogs with diverse
structures at the C32 side chain, and a wide spectrum of physico-
chemical properties.
Eribulin
9
1.66
0.17
0.12
1.91
0.23
0.11
0.17
0.69
0.15
0.08
0.25
0.07
0.13
0.13
0.13
0.16
3.16
0.16
0.23
0.17
0.14
0.28
0.18
2.17
0.13
3058
47.63
2.52
31.86
2.89
1842
288.17
22.35
16.81
12.35
6.90
6.57
5.65
49.82
38.10 nt
12.37
29.40
21.59
19.91
18.26
18.04
12.35
11.82
11.87
6.26
20
204
4.67
0.23
3.16
0.61
0.22
0.31
1.42
0.31
8.57
0.16
3.11
0.46
0.14
0.21
0.89
0.36
nt
0.46
0.22
0.21
0.40
0.67
0.78
0.62
0.20
0.34
0.27
0.15
0.37
0.19
5.40
0.17
153.95
nt
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
0.75
1.15
3.99
7.26
3.05
3.12
1.95
2.41
2.55
2.23
2.79
39.10
1.91
2.73
1.01
0.83
1.60
1
8
0.70
0.23
0.23
0.29
0.47
0.53
0.99
0.23
0.51
0.44
0.17
0.58
0.31
6.90
0.28
nt
All compounds were tested for cell growth inhibitory activity
against four human cancer cell lines (Table 1). In addition to the
human sarcoma cell line MES-SA, the primary assay system for this
study, compounds were also tested against the doxorubicin-resis-
tant subline MES-SA/Dx5-Rx1. The ratio of IC50 values between
these two cell lines, termed fold-resistance ratio (FR), was used
19
as a measure of Pgp susceptibility. Additionally, compounds were
evaluated in the human colon cancer cell lines DLD-1 and HCT-15,
both of which also express the MDR phenotype.
6.03
5.83
5.19
4.66
In agreement with our prediction, compounds with attenuated
basicity of substituents in the C32 side-chain region did indeed
lead to significantly reduced FR, while maintaining excellent
in vitro potency. The majority of compounds synthesized had FR
<20 and low- to sub-nM potency. Amines containing unsaturated
alkyl groups or electron-withdrawing substituents generally dis-
played lower FR, as compared to alkyl substituted or unsubstituted
amines. The same decrease in FR was also obtained by introducing
other heteroatoms (morpholines, acyl or sulfonyl piperazines), or
by way of ring strain (aziridines). Where present, conversion of
the C34 hydroxyl to the methyl ether was found to lower the FR.
In short, increasing the overall lipophilicity of the molecule led to
0.95
10.11
0.53
3.99
Doxorubicin 26.40
Paclitaxel
Vinblastine
2702.33
>1000
439.73
134.00
>338
224.18 nt
2.96
2.28
nt
28.50
IC50 values are means of at least two measurements. Partial structures (C32 side-
chain) of compounds 9–32 are shown in Figure 3.
nt = not tested.
Defined as ratio of antiproliferative IC50 against MES-SA/Dx5-Rx1 cells to IC50
against MES-SA cells.
a
Figure 4. Correlation of fold-resistance (FR) with log D7.4 values for the analogs synthesized in this study (compound 9 and eribulin not shown).

S. Narayan et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1634–1638
1637
lower FR within this series. This trend can be readily seen in Figure
, which shows the relationship between FR and the calculated
log D at pH 7.4 of the compounds synthesized in this study. An in-
crease in log D from 1.4 to 3 led to a significant lowering of FR from
8
00
4
control
Cmpd 25 3.0 mg/kg
Cmpd 25 2.25 mg/kg
Cmpd 25 1.7 mg/kg
Cmpd 25 1.3mg/kg
Cmpd 25 0.95 mg/kg
paclitaxel 12.5 mg/kg
700
5
0 to 5, although log D values beyond three did not lead to any fur-
ther decrease in FR. The observed dependence of FR on amine ba-
sicity is in line with observations reported by the Merck group on
KSP inhibitors, where P-gp susceptibility was decreased by intro-
6
00
2
0
500
400
ducing electron withdrawing groups proximal to an amine.
The most potent compounds exhibited sub-nM potency in all
four cell lines, making available a large pool of candidates for fur-
ther evaluation. We next turned to measuring the pharmacokinet-
ics of compounds from this series in mice. Our aim was to
determine if sufficient drug exposure would permit once daily dos-
ing in vivo. Table 2 summarizes pharmacokinetic parameters for
eribulin and compounds 11, 18, 20, and 25. Gratifyingly, for all
compounds tested, drug levels were found to be significantly high-
er than in vitro IC50 values over several hours post-administration.
In addition, and in contrast to eribulin, several analogs from this
study exhibited moderate levels of oral bioavailability. Thus, our
premise that optimizing for P-gp susceptibility might lead to im-
proved oral bioavailability appeared to hold true.
3
2
00
00
100
0
7
14
21
28
35
Day
Figure 6. Effect of compound 25 at different doses on the growth of subcutaneously
implanted LOX xenografts in athymic mice. Compound 25 was dosed orally, once
Table 2
daily, 5 days
intravenously.
a week, for two weeks starting day 6. Paclitaxel was dosed
Pharmacokinetic parameters for selected compounds in BALB-c mice following a
single oral dose
ssa(L/kg)
CL (L/kg/h)
a
F (%)
Compd
AUC0–inf (ng h/mL)
t
1/2 (h)
V
Eribulin
161
63
236
229
1094
15.4
4.1
1.7
4.7
0.9
4.95
4.78
3.64
12.78
0.92
1.19
2.91
2.82
4
3.2
11.3
13.3
18.3
14.9
Based on cell growth inhibition data, in vitro distribution,
metabolism and pharmacokinetic (DMPK) properties, and in vivo
pharmacokinetics, several candidates for were prioritized for effi-
cacy testing in in vivo tumor xenograft models. We chose the
LOX human melanoma xenograft as a prototypical drug-sensitive
model, while DLD-1 human colon cancer served as a typical MDR
tumor model. For purposes of direct comparison, all compounds
were tested in similar dosing regimens. In addition, all compounds
were administered intravenously so that their relative in vivo
activities could be assessed whether or not they exhibited oral bio-
availability. Figure 5 shows the effect of compound 25 on tumor
growth in a DLD-1 xenograft model in athymic mice. Compound
25 inhibited tumor growth in this model in a dose dependent man-
ner at doses well below 5 mg/kg. After a 2 week initial delay, tumor
stasis was observed as long as compound administration was con-
tinued, with tumor regrowth occurring after dosing was stopped.
Oral efficacy of compound 25 was evaluated in LOX xenograft
models (Fig. 6). Long term tumor regression was observed in this
model following oral dosing. Taken together, the data reported
herein suggest that compound 25 and its analogs might be active
against a wide range of human cancers, in addition to being active
against MDR tumors and amenable to oral dosing regimens.
In summary, introduction of amines with low basicity in the
C32 side chain, or otherwise increasing lipophilicity in this portion
of eribulin, led to the identification of analogs with greatly reduced
susceptibility to P-gp mediated drug efflux. These compounds
showed moderate oral bioavailability and were orally active
in vivo. In addition, this series afforded compounds with activity
against resistant tumor cell lines such as DLD-1 in xenograft mod-
els. These novel compounds may thus provide candidates for the
treatment of a wider variety of human cancers.
1
1
2
2
1
8
0
5
0.68
t
max = 5 min in all cases. Eribulin was administered at 10 mg/kg; all other com-
pounds were administered at 5 mg/kg.
a
Measured from a parallel iv experiment.
7
6
5
4
3
2
1
00
00
00
00
00
00
00
0
Cmpd 25 3.0 mg/kg
Cmpd 25 2.0 mg/kg
Cmpd 25 1.3 mg/kg
CPT-11 50 mg/kg
Saline Control
1
4
21
28
35
42
49
56
Day
Acknowledgments
Figure 5. Effect of compound 25 at different doses and CPT-11 on subcutaneously
implanted DLD-1 xenografts in athymic mice. Compounds were administered
intravenously once daily, on Mondays, Wednesdays and Fridays, for three weeks
starting day 13.
We thank our colleagues in Andover’s section of Pharmaceutical
Science and Technology Core Function Unit (formerly Andover Pro-
cess Research and Chemical Development Departments) for the

1
638
S. Narayan et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1634–1638
supply of eribulin and other intermediates and many helpful
discussions.
8. Dumontet, C.; Jordan, M. A. Nat. Rev. Drug Discovery 2010, 9, 790.
Yu, M. J.; Kishi, Y.; Littlefield, B. A. Anticancer Agents Nat. Prod. 2005, 241.
9.
1
1
1
1
1
0. Lin, J. H.; Yamazaki, M. Drug Metab. Rev. 2003, 35, 417.
1. Goda, K.; Bacso, Z.; Szabo, G. Curr. Cancer Drug Targets 2009, 9, 281.
2. Glavinas, H.; Krajcsi, P.; Cserepes, J.; Sarkadi, B. Curr. Drug Delivery 2004, 1, 27.
3. Chan, L. M.; Lowes, S.; Hirst, B. H. Eur. J. Pharm. Sci. 2004, 21, 25.
4. Tanigawara, Y. Ther. Drug Monit. 2000, 22, 137.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.01.097. These data
include MOL files and InChiKeys of the most important compounds
described in this article.
15. Salerno, M.; Przewloka, T.; Fokt, I.; Priebe, W.; Garnier-Suillerot, A. Biochem.
Pharmacol. 2002, 63, 1471.
16. Frezard, F.; Pereira-Maia, E.; Quidu, P.; Priebe, W.; Garnier-Suillerot, A. Eur. J.
Biochem. 2001, 268, 1561.
17. Zheng, W.; Seletsky, B. M.; Palme, M. H.; Lydon, P. J.; Singer, L. A.; Chase, C. E.;
Lemelin, C. A.; Shen, Y.; Davis, H.; Tremblay, L.; Towle, M. J.; Salvato, K. A.;
Wels, B. F.; Aalfs, K. K.; Kishi, Y.; Littlefield, B. A.; Yu, M. J. Bioorg. Med. Chem.
Lett. 2004, 14, 5551.
References and notes
1
8. Typically 1–5 mg of each analog was synthesized. Compounds were purified by
1
2
3
.
.
.
Dean-Colomb, W.; Esteva, F. J. Semin. Oncol. 2008, 35, S31.
Newman, S. Curr. Opin. Invest. Drugs 2007, 8, 1057.
Twelves, C.; Cortes, J.; Vahdat, L. T.; Wanders, J.; Akerele, C.; Kaufman, P. A. Clin.
Breast Cancer 2010, 10, 160.
Smith, J. A.; Wilson, L.; Azarenko, O.; Zhu, X.; Lewis, B. M.; Littlefield, B. A.;
Jordan, M. A. Biochemistry 2010, 49, 1331.
Jordan, M. A.; Kamath, K.; Manna, T.; Okouneva, T.; Miller, H. P.; Davis, C.;
Littlefield, B. A.; Wilson, L. Mol. Cancer Ther. 2005, 4, 1086.
Kuznetsov, G.; Towle, M. J.; Cheng, H.; Kawamura, T.; TenDyke, K.; Liu, D.; Kishi,
Y.; Yu, M. J.; Littlefield, B. A. Cancer Res. 2004, 64, 5760.
Okouneva, T.; Azarenko, O.; Wilson, L.; Littlefield, B. A.; Jordan, M. A. Mol.
Cancer Ther. 2008, 7, 2003.
HPLC, and characterized by 1H NMR and MS.
1
9. Liu, J.; Towle, M. J.; Cheng, H.; Saxton, P.; Reardon, C.; Wu, J.; Murphy, E. A.;
Kuznetsov, G.; Johannes, C. W.; Tremblay, M. R.; Zhao, H.; Pesant, M.; Fang, F.
G.; Vermeulen, M. W.; Gallagher, B. M., Jr.; Littlefield, B. A. Anticancer Res. 2007,
4
5
6
7
.
.
.
.
2
7, 1509.
2
0. Cox, C. D.; Breslin, M. J.; Whitman, D. B.; Coleman, P. J.; Garbaccio, R. M.; Fraley,
M. E.; Zrada, M. M.; Buser, C. A.; Walsh, E. S.; Hamilton, K.; Lobell, R. B.; Tao, W.;
Abrams, M. T.; South, V. J.; Huber, H. E.; Kohl, N. E.; Hartman, G. D. Bioorg. Med.
Chem. Lett. 2007, 17, 2697.