Discovery of an in vitro and in vivo potent resorcylic lactone analog of LL-Z1640-2 as anti-inflammatory lead, II

Article abstract of DOI:10.1016/j.bmcl.2010.03.119

The potent in vitro lead compound, ER-803064 (2), a MEK1 and MEKK1 inhibitor inspired from natural product LL-Z1640-2 (f152A1), was further optimized to improve in vitro and in vivo potency. The modifications on C14 position led to discovery of the lead compounds 28 and 29, which regained full in vitro potency of f152A1 and showed higher in vivo potency by iv administration.

Full text of DOI:10.1016/j.bmcl.2010.03.119

Bioorganic & Medicinal Chemistry Letters 20 (2010) 3047–3049
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of an in vitro and in vivo potent resorcylic lactone analog
of LL-Z1640-2 as anti-inflammatory lead, II
Yongchun Shen a, Hong Du a, Makoto Kotake b, Tomohiro Matsushima b, Masaki Goto b, Hiroshi Shirota a,
Fabian Gusovsky a, Xiangyi Li a, Yimin Jiang a, Shawn Schiller a, Mark Spyvee a, Heather Davis a, Zhiyi Zhang a,
Robert Pelletier a, Megumi Ikemori-Kawada b, Yoshiyuki Kawakami b, Atsushi Inoue b, Yuan Wang a,
*
a Eisai Inc., 4 Corporate Drive, Andover, MA 01810, USA
b Eisai Tsukuba Research Laboratories, 1-3, Tokodai 5-chome, Tsukuba-shi, Ibaraki 300-2635, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The potent in vitro lead compound, ER-803064 (2), a MEK1 and MEKK1 inhibitor inspired from natural
product LL-Z1640-2 (f152A1), was further optimized to improve in vitro and in vivo potency. The mod-
ifications on C14 position led to discovery of the lead compounds 28 and 29, which regained full in vitro
potency of f152A1 and showed higher in vivo potency by iv administration.
Received 18 January 2010
Revised 26 March 2010
Accepted 31 March 2010
Available online 9 April 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Resorcylic lactone
LL-Z1640-2
f152A1
Anti-inflammatory
In an earlier Letter,1 we described the effort in bioassay-guided
fractionation through phenotypic screens in identifying a natural
product f152A1 (1), also known as LL-Z1640-2,2 from the fermenta-
tion broth of a fungus, Curvularia verruculosa, in the mid 1990’s.
f152A1 (1) was found to exhibit potent anti-inflammatory activities
in vitro and was rapidly inactivated in human or mouse micro-
somes and plasma.1 Shown in our last report, chemical modification
on C4 position led to ER-803064 (2) with increased metabolic sta-
bility and reduced potency (Fig. 1).1 ER-803064 is active in vivo,
(Fig. 2).3 The C14-ethyl analog 6 was slightly more potent than 2
(IC50: 100 nM). According to this result, we focused our efforts on
C14 modification.
Due to the labile nature of enone functionality, C14 analogs
could not be synthesized from ER-803064 (2) directly. To facili-
tate this modification, a practical synthetic route was developed
based on these criteria: (1) access to advanced intermediates
that could be readily derivatized to a wide range of analogs;
but the ED50 value (13.2 mg/kg, iv) was fairly high in regard to TNF
a
suppression. Thus, we pursued further optimization on the struc-
ture toward a lead compound that would maintain the metabolic
stability with improved potency in vitro and in vivo. In this Letter,
we described the discovery of lead compounds 28 and 29 through
optimization of the C14 position.
Having identified C4-methyl as the critical element for meta-
bolic stability, and explored structure modifications from C3 to
C8 positions, we turned our focus to study the SAR in C11–C12
and aromatic positions while keeping a methyl group in the C4 po-
OH
O
OH
O
1
O
4
O
4
O
O
O
O
14
OH
OH
10
OH
OH
2
1
: ER-803064
: f152A1 = LL,Z-1640-2
OH
O
sition. The assays used for lead optimization were TNFa-PLAP and
O
O
Actin-PLAP reporter assays described previously.1 The C11–C12
analogs and C13-Br analog were not as potent as 2 (IC50s:
136 nM for 2, 1683 nM for 3, 293 nM for 4 and 804 nM for 5)
O
S
28
29
R=
R=
R
O
O
N
N
OH
OH
* Corresponding author. Tel.: +1 978 837 4859; fax: +1 978 837 4863.
E-mail address: yuan_wang@eisai.com (Y. Wang).
Figure 1. Structures of f152A1 (1), ER-803064 (2), 28 and 29.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.03.119

3048
Y. Shen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3047–3049
OH
O
OH
O
OH O
O
O
O
O
O
O
O
O
O
O
NH
14
13
Br
OH
OH
OH
O
11
OH
OH
OH
2
3
4
5
6
TNF-PLAP (nM)
ACT-PLAP (nM)
136
1683
293
804
100
>10000
>10000
>10000
>10000
>10000
Figure 2. Structures and in vitro activities of C11–C14 analogs.
O
O
BzO
O
MOMO
O
O
OH
O
a
TMS TBS
I
O
O
TMS
HO
b
7
8
SePh
MOM
MOM
O
O
O
O
O
O
BzO
O
BzO
d
c
TMS
HO
O
O
10
9
O
O
MOM
O
OH
O
O
O
O
e
O
HO
O
R
O
R
OH
O
OH
O
11
C14-analogs
Scheme 1. Synthetic route towards C14 analogs. Reagents and conditions: (a) (i) HOCH2CH2TMS, DEAD, Ph3P, 55%; (ii) MOM-Cl, DBU, 74%; (iii) LDA, Ph2Se2; (iv) 2.5 N NaOH,
EtOH, 92% for two steps; (v) HOCH2CH2TMS, DEAD, Ph3P, 77%; (b) (i) LiHMDS, acyclic iodide; (ii) MCPBA, Et3N, 36% for two steps; (iii) TBAF; (iv) 2-Cl-1-Me pyridinium iodide,
79% for two steps; (c) TBAF, 58%; (d) (i) Mitsunobu; (ii) NaOH, 50% for two steps; (e) PCC OR Swern; (ii) HCl, 60% for two steps.
Table 1
(2) the enone functionality would be introduced at late stage to
Structures and activity data for selected C14 analogs
avoid the isomerization and decomposition. The synthesis started
from commercially available Methyl Orsellinate (7) (Scheme 1).
Under Mitsunobu condition, p-phenol group in 7 was selectively
protected as TMS ethylether (55% yield), which further went
through a four-step transformation sequence: MOM-Cl protection
(74% yield), selenination and saponification (92% yield for two
steps), and Mitsunobu esterification (77% yield) to afford aromatic
intermediate 8. Selenide 8 was coupled with acyclic iodide in the
presence of LiHMDS (acyclic iodide was synthesized from commer-
Analog
R
TNF
a
-PLAP IC50 (nM)
Actin-PLAP IC50 (nM)
2
6
MeO–
EtO–
HO–
n-PrO–
n-BuO–
i-PrO–
CF2HO–
BnO–
136
100
515
249
325
687
615
319
752
168
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
13
14
15
16
17
18
19
20
3,4-diCl-BnO–
4-MeO–BnO–
cially available 2-deoxy-D-ribose and (S)-(+)-3-hydroxy-N-butyric
acid methyl ester),3 followed by MCPBA oxidation (36% for two
steps), TBAF deprotection and lactone cyclization under the influ-
ence of Mukaiyama’s reagent (79% yield for two steps) to provide
lactone 9.4 Fully protected intermediate 9 was treated with TBAF
to remove TMS ethyl protection group and afford key advanced
intermediate 10 in 58% isolated yield. From phenol 10, C14 substi-
tutions were introduced through Mitsunobu reaction and benzoyl
group was then deprotected under basic condition to produce
allylic alcohol 11 in 50% yield over two steps. Alcohol 11 was oxi-
dized (PCC oxidation or Swern oxidation) into labile enone interme-
diate and MOM group was then removed in the presence of aq HCl
to furnish C14 analogs in 60% yield after purification over the final
two steps.
From the advanced intermediate 10, the C14 phenol group was
readily derivatized into analogs with simple alkyl and benzyl
groups.3 Biologically n-alkyl groups with 1–4 carbon length was
very well tolerated. The two-carbon alkyl substitution (analog 6)
demonstrated the best potency within this group (Table 1).
We explored the size of C14 binding pocket and whether addi-
tional interactions with the kinase around C14 position might im-
prove potency.5 Thus, analogs with C14 extended alkyl side chains
bearing N or O-functional groups were synthesized. Both ethylene
and propylene turned out to be good linkers between C14 oxygen

Y. Shen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3047–3049
3049
Table 2
synthesis of important C14 analogs. Our medicinal chemistry effort
led to the discovery of the analogs 28 and 29, fully synthetic ana-
logs of f152A1 (1). The analog 29 maintained the metabolic stabil-
ity, regained full in vitro potency similar to 1 and had significant
improvement in in vivo potency. These new leads were further
evaluated in various animal models to validate the anti-inflamma-
tory application.
Structures and activity data for selected analogs
Analog C14 R group
TNFa-PLAP IC50
(nM)
Actin-PLAP IC50
(nM)
21
22
23
24
25
26
27
28
29
Me2N–CH2CH2–O–
597
383
113
59
65
43
31
32
15
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
HO–CH2–(HO)CH–CH2–O–
2-Pyridine-NH–(CH2)2–O–
Morpholine-CH2CH2–O–
CH3–CO–NHCH2CH2–O–
1-Triazole-(CH2)2–O–
Ac–NH–CH2–CO–CH2–O–
MeSO2–(CH2)3–O–
1-Imidazole-(CH2)2–O–
Acknowledgments
We thank Dr. Yoshito Kishi of Eisai Scientific Advisory Board for
insightful advices in the long duration of this program. We
acknowledge contribution from researchers who supported this ef-
fort, Drs. Kenichi Chiba, Naoki Yoneda, Yoshihito Eguchi, Hideki
Sakurai, Hatsue Ito-Igarashi, Akifumi Kimura, Yoshikazu Kuboi,
Yoshiharu Mizui, Isao Tanaka, and Takatoshi Kawai. We also ex-
press appreciation to Drs. Seiichi Kobayashi (Eisai, Japan), Ieharu
Hishinuma and Nancy Wong (Eisai Inc.) for helpful discussion.
Table 3
In vivo data for selected analogs
Analog C14 R group
TNFa
-PLAP IC50
ED50, mg/kg (iv)
TNFa
(nM)
1
2
6
28
29
—
11
136
100
32
No dose response
13.2
12
9.8
6.5
References and notes
CH3–O–
CH3–CH2–O–
MeSO2–(CH2)3–O–
1-Imidazole-(CH2)2–O–
1. Du, H.; Matsushima, T.; Spyvee, M.; Goto, M.; Shirota, H.; Gusovsky, F.; Chiba, K.;
Kotake, M.; Yoneda, N.; Eguchi, Y.; DiPietro, L.; Harmange, J.-C.; Gilbert, S.; Li, X.-
Y.; Davis, H.; Jiang, Y.; Zhang, Z.; Pelletier, R.; Wong, N.; Sakurai, H.; Yang, H.; Ito-
Igarashi, H.; Kimura, A.; Kuboi, Y.; Mizui, Y.; Tanaka, I.; Ikemori-Kawada, M.;
Kawakami, Y.; Inoue, A.; Kawai, T.; Kishi, Y.; Wang, Y. Bioorg. Med. Chem. Lett.
2009, 19, 6196.
15
and the new functional groups. Sulfone analog 28 (R = MeSO2–
(CH2)3–O) and imidazole analog 29 (R = 1-imidazole-(CH2)2–O)
had significantly improved potency over ER-803064 (2) (IC50s:
32 nM for 28, 15 nM for 29 vs 136 nM for 2) (Table 2). The
in vitro potency of 29 was comparable to f152A1 (IC50: 11 nM).
We found that both 28 and 29 were more potent than 2 in vivo
(Table 3).6 In particular, 29 significantly improved in vivo potency
over ER-803064 (2) (ED50: 6.5 mg/kg for 29 vs 13.2 mg/kg for 2)
(Table 3).
In conclusion, we have studied the SAR of f152A1 analogs with
synthetic modifications on C11–C12 and aromatic positions. Mod-
ifications on C14 aromatic position resulted in a series of interest-
ing analogs. A practical synthetic route was developed for the
2. Ellestad, G. A.; Lovell, F. M.; Perkinson, N. A.; Hargreaves, R. T.; McGahren, W. J. J.
Org. Chem. 1978, 43, 2339.
3. All routes for analog preparations are described in this publication: Boivin, R.;
Chiba, K.; Chow, J.; Du, H.; Eguchi, Y.; Fujita, M.; Goto, M.; Gusovsky, F.;
Harmange, J. C.; Inoue, A.; Jiang, Y.; Kawada, M.; Kawai, T.; Kawakami, Y.;
Kimura, A.; Kotake, M.; Kuboi, Y.; Lemelin, C.; Li, X. Y.; Matsushima, T.; Mizui, Y.;
Muramoto, K.; Sakurai, H.; Shen, Y.; Shirota, H.; Spyvee, M.; Tanaka, I.; Wang, Y
(J).; Yamamoto, S.; Yoneda, N.; PCT/US03/07377, WO 2003076424 A1 20030918.
4. Mukaiyama’s reagent as activating agent for carboxylic acids: (a) Mukaiyama,
T.; Usui, M.; Saigo, K. Chem. Lett. 1976, 49; (b) Ando, W.; Furuhata, T.; Tsumaki,
H.; Akira Sekiguchi, A. Chem. Lett. 1978, 885.
5. For in silico analysis of f152A1 and analogs see: (a) Ikemori-Kawada, M.; Kawai,
T.; Goto, M.; Wang, Y. J.; Kawakami, Y. J. Chem. Inf. Model. 2009, 49, 2650; (b)
Ohori, M.; Kinoshita, T.; Yoshimura, S.; Warizaya, M.; Nakajima, H.; Miyake, H.
Biochem. Biophys. Res. Commun. 2007, 353, 633.
6. Conditions for various assays see Supplementary data in Ref. 1.