Aleglitazar, a new, potent, and balanced dual PPARα/γ agonist for the treatment of type II diabetes

Article abstract of DOI:10.1016/j.bmcl.2009.03.036

Design, synthesis, and SAR of novel α-alkoxy-β-arylpropionic acids as potent and balanced PPARαγ coagonists are described. One representative thereof, Aleglitazar ((S)-2Aa), was chosen for clinical development. Its X-ray structure in complex with both receptors as well as its high efficacy in animal models of T2D and dyslipidemia are also presented.

Full text of DOI:10.1016/j.bmcl.2009.03.036

Bioorganic & Medicinal Chemistry Letters 19 (2009) 2468–2473
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Aleglitazar, a new, potent, and balanced dual PPAR
for the treatment of type II diabetes
a/c agonist
Agnes Bénardeau a, Jörg Benz a, Alfred Binggeli a, Denise Blum a, Markus Boehringer a, Uwe Grether a,
Hans Hilpert a, Bernd Kuhn a, Hans Peter Märki a, Markus Meyer b, Kurt Püntener a, Susanne Raab a,
Armin Ruf a, Daniel Schlatter a, Peter Mohr a,
*
a F. Hoffmann-La Roche Ltd, Pharma Research, CH-4070 Basel, Switzerland
b F. Hoffmann-La Roche Ltd, Pharma Development, CH-4070 Basel, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Design, synthesis, and SAR of novel a-alkoxy-b-arylpropionic acids as potent and balanced PPARac coag-
Received 12 February 2009
Revised 11 March 2009
Accepted 12 March 2009
Available online 14 March 2009
onists are described. One representative thereof, Aleglitazar ((S)-2Aa), was chosen for clinical develop-
ment. Its X-ray structure in complex with both receptors as well as its high efficacy in animal models
of T2D and dyslipidemia are also presented.
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
a-Alkoxypropionic acids
Nuclear hormone receptors
PPAR
X-ray crystallography
Diabetes
Dyslipidemia
Peroxisome Proliferator-Activated Receptors (PPAR) a, c, and d
are three of the 48 nuclear hormone receptors encoded in the hu-
man genome.1 As ligand-activated transcription factors they work
in concert with co-activators and co-repressors and regulate gene
Aromatic
Center
Cyclic
Tail
Acidic Head
Linker
Linker
expression.2 PPAR
a
is supposed to control mainly genes involved
in fatty acid b-oxidation and to play a pivotal role in energy
homeostasis,3 whereas PPAR
agonists seem to be mainly involved
Scheme 1. Simplified topology of typical synthetic PPAR agonists. The linkers can
be branched to access additional subpockets in the receptor.
c
in glucose homeostasis, insulin sensitivity, and lipid storage.4 Li-
gands for the former, albeit very weak ones, generally termed Fi-
brates, have been in clinical use since a long time to treat
dyslipidemia,5 whereas the glitazones (thiazolidinediones), ago-
nists of the latter, belong today to the standard repertoire for type
II diabetes treatment.6 The function of PPARd, on the other hand, is
much less understood to date, but some recent data enlighten its
role in the regulation of fatty acid oxidation in several tissues.7
Some time ago we reported about structure-based design and syn-
thesis of indole propionic acids as novel PPARac coagonists.8 In
this Letter we describe an extension of this work culminating in
the discovery of Aleglitazar, a balanced, potent coagonist which
provides new opportunities to address both hyperglycemia as well
as the enhanced cardiovascular risk of diabetic patients.9,10 Aleglit-
azar has successfully completed phase II clinical development.
Again, we started with a modular approach as schematically de-
picted in Scheme 1. We chose a-alkoxyacids, which are apt to form
up to 4 pivotal hydrogen bonds with serine, tyrosine and histidine
of the protein, as the acidic warhead; such a strong hydrogen
acceptor is indispensable for obtaining potent agonists. For the
cyclic tail, partly solvent exposed and in general quite tolerant with
respect to structural variations, we focused again on the aryloxaz-
ole-moiety, present and well-tried in countless PPAR ligands like
Farglitazar, Imiglitazar, or Muraglitazar, to name just a few.
Regarding the aromatic center, close inspection of several X-ray
structures revealed that a simple phenyl ring does not optimally
fit the cavity of the receptor. We therefore decided to explore the
potential of bicyclic cores, specifically naphthalene and benzothio-
phene analogues which had already proven in the past, albeit re-
stricted to pure PPAR
c agonists, to yield drug like antidiabetic
compounds like Edaglitazone 1. In the following, we describe the
synthesis, SAR, X-ray structures, and in vivo activity of PPARac
coagonists of generic structure 2, 3, and 4 (Fig. 1).
* Corresponding author. Tel.: +41 61 688 5691; fax: +41 61 688 64 59.
E-mail address: peter.mohr@roche.com (P. Mohr).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.03.036

A. Bénardeau et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2468–2473
2469
O
R2
O
O
N
O
N
OH
R1
S
R3
O
O
N
S
O
O
S
Edaglitazone 1
2
4
R2
R2
O
O
O
N
O
OH
R1
OH
R1
R3
R3
O
O
N
O
O
S
3
Figure 1. Chemical structure of Edaglitazone and PPAR coagonists 2–4.
The synthesis started with the preparation of aryl-oxazoles 8
following the established route8 from the appropriate benzalde-
hyde 5 and butane-2,3-dione mono-oxime 6 via 7 according to Go-
to’s method (Scheme 2).11 Elongation by one carbon atom to 9 was
achieved by treatment with NaCN in DMSO. Standard hydrolysis
and ensuing reduction of the carboxylic acid 10 with borane deliv-
ered the terminal alcohol 11. Mitsunobu condensation with build-
ing block 12 generated finally the key intermediate 13, albeit in
moderate yield. The former was synthesized in an expeditious
manner following the route described previously.12 Aldol conden-
sponding product 15 in excellent yield. In the cases where
alkylated glycolate esters 14 are not commercially available, they
were prepared by alkylation of glycolic ester with the appropriate
alkyl iodide in the presence of Ag2O.13 Ionic reduction with trieth-
ylsilane in TFA14 and basic hydrolysis completed eventually the
synthesis of 2. The corresponding naphthalene derivatives 3 and
the isomeric benzothiophenes 4 were prepared analogously, but
using in step (f) 4-hydroxy-naphthalene-1-carbaldehyde or 7-hy-
droxy-benzo[b]thiophene-4-carbaldehyde, respectively.
Homochiral PPAR ligands 2, 3, and 4, on the other hand, were
synthesized using Evans’ boron enolate methodology as depicted
sation with the Li-enolate of
a-alkoxyester 14 afforded the corre-
R2
R2
R2
O
H
R2
O
a
b
O
N+
O
N
O
N
c
+
R3
R3
R3
NOH
R3
O
8
5
6
7
9
Cl
N
d
R2
O
S
R2
O
e
O
+
R3
f
N
O
R3
N
11
12
O
10
O
O
O
R2
O
O
S
R2
g
O
O
N
O
N
O
+
O
R1
O
O
R3
R3
R1
O
O
S
14
13
15
h
O
R2
i
R2
O
O
N
O
O
O
R3
O
R1
O
S
R3
O
R1
N
O
S
2
16
Scheme 2. Reagents and conditions: (a) AcOH/HCl gas, 0 °C, 1 h, 60–90%; (b) POCl3/CH2Cl2, reflux, 14 h, 60–90%; (c) NaCN/DMSO, 35 °C, 2 h, 90–98%; (d) NaOH/EtOH/water,
65 °C, 14 h, 60–85%; (e) BH3/THF, rt, 14 h, followed by reflux in MeOH, 1 h, 80–95%; (f) Ph3P/DIAD/toluene, rt, 2 h, 25–50%; (g) LDA/THF/hexane, ꢀ78 °C, 30 min, 70–98%; (h)
Et3SiH (10 equiv)/TFA, 0 °C, 3–14 h, 60–85%; (i) NaOH/EtOH/THF/water, rt, 2 h, 70–95%.

2470
A. Bénardeau et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2468–2473
O
S
O
O
O
R2
O
O
S
R2
a
O
N
O
N
N
N
O
O
+
R3
O
O
R3
R1
R1
O
O
13
18
17
b
R2
O
O
c
R2
O
O
O
O
R3
N
O
O
R1
N
O
S
R3
O
R1
N
O
S
(S)-2
19
Scheme 3. Reagents and conditions: (a) nBu2BOTf, NEt3, CH2Cl2, ꢀ78 to 0 °C, 3 h, 30–90%; (b) Et3SiH, TFA, 14 h, 60–90%; (c) LiOH/THF, 35 °C, 2 h, 90–98%.
in Scheme 3.14,15 The control of absolute configuration at the
a
-po-
All compounds were screened in PPARa and c binding (IC50)
sition was close to perfect (routinely >95%). Similar ionic reduction
as above and ensuing mild hydrolysis afforded the desired target in
good yield.
and transactivation (EC50) assays; selected compounds were also
tested in PPARd assays.16 As can be seen from Table 1, most of
the molecules turned out to be potent agonists at both receptors.
Table 1
Binding affinities (IC50) and functional transactivation data (EC50) of 2, 3, and 4 on human PPARsa,b
R2
R3
R2
R2
R3
R3
O
O
O
O
N
O
N
O
N
R4
O
R4
R4
O
O
O
R1
O
O
O
R2'
R3'
R1
R1
S
O
O
R2'
R2'
R3'
R3'
S
2
3
4
0
0
Compound No.
R1
R2, R2
R3, R3
R4
IC50a
(lM)
IC50c
(l
M)
EC50
a
(
lM) (% effect)
EC50c (lM) (% effect)
rac-2Aa
rac-2Ba
rac-2Ca
rac-2Da
rac-2Ea
rac-2Ag
rac-3Aa
rac-3Ba
rac-3Ca
rac-3Ah
(S)-2Aa
(S)-2Ab
(S)-2Ac
(S)-2Ad
(S)-2Ae
(S)-2Ba
(S)-2Ca
(S)-2Fa
(S)-2Bb
(S)-2Ga
(S)-2Bf
(S)-2Gd
(R)-2Ba
(R)-2Aa
(S)-3Ba
(S)-3Bb
(S)-3Ff
CH3
H, H
H, H
H, H
H, H
H, H
OMe, H
H, H
H, H
H, H
Me, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H
H
H
H
H
H
H
H
H
H
H
CF3
H
iPr
H
H
H
H
CF3
H, H
Ph
iPr
H
H
H
CF3
Ph
H
0.028
0.022
0.175
0.239
0.120
0.294
0.061
n. d.
0.069
0.432
0.038
0.053
n. d.
0.045
0.048
0.053
0.393
0.060
0.063
0.021
0.194
0.037
2.370
4.610
0.198
0.905
0.429
0.918
0.028
0.333
>10
0.046
0.007
0.004
0.003
0.002
0.916
0.113
0.030
0.007
0.802
0.019
0.023
0.008
0.002
0.003
0.021
0.002
0.003
0.017
0.002
0.003
0.005
2.68
0.103 (168)
0.027 (164)
0.075 (123)
0.048 (181)
0.249 (115)
0.435 (101)
0.131 (159)
n. d.
0.087 (56)
0.123 (114)
0.050 (156)
0.002 (109)
0.043 (?)
0.020 (151)
0.061 (196)
0.027 (109)
0.013 (204)
0.555 (179)
0.089 (43)
0.027 (108)
0.167 (92)
0.016 (122)
1.91 (73)
0.358 (109)
0.011 (78)
0.011 (135)
0.987 (79)
0.65 (92)
0.059 (131)
0.018 (100)
0.012 (64)
0.004 (63)
0.004 (85)
1.615 (64)
0.123 (98)
n. d.
0.016 (81)
0.716 (78)
0.021 (67)
0.053 (132)
0.011 (61)
0.010 (53)
0.023 (44)
0.021 (63)
0.004 (25)
0.151 (119)
0.009 (89)
0.032 (79)
0.431 (61)
0.010 (56)
0.787 (81)
1.21 (65)
CH3CH2
CH3CH2CH2
CH2CH(CH3)2
n-Hexyl
CH3
CH3
CH3CH2
CH3CH2CH2
CH3
CH3
CH3
CH3
CH3
CH3
CH3CH2
CH3CH2CH2
CF3CH2
CH3CH2
CH2CHCH2CH2
CH3CH2
CH2CHCH2CH2
CH3CH2
CH3
CH3CH2
CH3CH2
CF3CH2
CH3
Me, Me
H, H
OMe, OMe
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
H, H
1.44
0.036
0.027
0.008
1.57
0.011
0.013
0.45
0.037 (72)
0.021 (156)
0.038 (96)
0.122 (31)
0.033 (159)
0.007 (40)
0.45 (85)
(R)-3Aa
(S)-4Aa
(S)-4Ba
Rosiglitazone (RG)
Farglitazar (GW2570)
Fenofibric acid
CH3
CH3CH2
H
H
0.157 (174)
0.058 (89)
>10
0.284 (100)
69 (11)
0.633
106
0.001
744
0.005 (252)
217 (2)
a
PPAR radioligand binding and functional transactivation (luciferase transcriptional reporter gene) assays were performed as described in Binggeli et al.29
Effects are reported in relation to 2-(S)-2-(2-benzoyl-phenylamino)-3-{4-[2-(5-methyl-2-phenyl-oxazol-4-yl)-ethoxy]-phenyl}-propionic acid (Farglitazar, GW 262570)
b
(100%, PPARa) and 5-{4-[2-(5-methyl-2-phenyl-oxazol-4-yl)-ethoxy]-benzo[b]thiophen-7-ylmethyl}-thiazolidine-2,4-dione (Edaglitazone) (100%, PPARc).

A. Bénardeau et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2468–2473
2471
that ortho-substitution, in particular o,o0-disubstitution, abolishes
activity completely.
Tyr314
To gain further insight into the binding mode we cocrystallized
the ternary complex of the human PPARa receptor ligand binding
His440
domain with compound (S)-2Aa and a receptor coactivator SRC-1
fragment.17 The structure was solved to a resolution of 2.2 Å and
showed a clear electron density for the bound ligand.18 In addition,
dual agonist (S)-2Aa was also cocrystallized with the ternary com-
Ser280
Tyr464
plex of the human PPARc receptor ligand binding domain with the
agonist and a receptor coactivator SRC-1 fragment.19 This structure
was solved to a resolution of 2.3 Å.18 The overall shape of this com-
plex is very reminiscent of previously published PPARa and c com-
plex structures with the AF2-helix in the agonist-type
conformation. Within the ligand binding sites shown in Figure 2,
the typical four strong hydrogen bonds between the ligand carbox-
ylate and the Ser, His and Tyr residues of PPARa and c, respec-
tively, can be identified (all H-bond distances 6 3.0 Å). Apart
from the polar head group recognition, both binding sites are com-
posed of predominantly hydrophobic side chains with limited sol-
vent access in the central and tail region. The central
benzothiophene linker designed to snugly fit into the protein cav-
ity does its job as anticipated.20
His323
His449
The two X-ray structures nicely allow rationalizing the ob-
served reverse SAR trend with respect to R1-extensions.While in
Ser289
PPAR
date longer R1 residues, the respective cleft in PPAR
ably smaller due to
c
a preformed pocket exists which can smoothly accommo-
Tyr473
a
is consider-
a
sequence difference (Phe363(g) ,
Ile354(a)). Very likely, the bulkier branched isoleucine clashes with
larger R1 residues resulting in a drop of binding affinity for this
isoform.
R1 pocket
(S)-2Aa, which has decent physicochemical properties, an
excellent pharmacokinetic profile in several animal species21 and
disposes of a well-balanced ac ratio was selected for further profil-
ing.22 Figure 3 summarizes the results of an OGTT23 after adminis-
tering compound (S)-2Aa for 12 days to db/db mice at a dose of 0.3
or 3 mg/kg/d, respectively, in direct comparison with Rosiglitaz-
one, and the effect on non-fasted plasma glucose levels. Even at
ten times lower doses, (S)-2Aa compares with respect to both
parameters very favorably with the reference insulin sensitizer
Rosiglitazone (RSG).
Figure 2. X-ray structures of (S)-2Aa bound to hPPARa- (top) and c-ligand binding
domain (bottom).
Increasing length and size of R1 tends to boost potency at the
form, whereas the opposite is true with respect to PPAR . Branch-
c-iso-
a
ing is deleterious; unsaturation in the side chain, however, is well
tolerated ((S)-2Ga, (S)-2Gd). As could be anticipated, (R)-isomers
are almost inactive. The substitution pattern of the phenyl-oxazole
influences the affinity only marginally, with the only exception
(S)-2Aa was further profiled in a hyperinsulinemic euglycemic
clamp study in Zucker fa/fa rats.24 Figure 4 illustrates the robust ef-
oGTT
Post-prandial blood glucose
25
24
20
16
12
8
20
15
10
Glucose challenge
(2g/kg BW)
4
0
30
60
90
120
Time (min)
Placebo
RSG
3 mkd
(S)-2Aa
0.3 mkd
S)-2Aa
3.0 mkd
Figure 3. Oral Glucose Tolerance Test with (S)-2Aa and Rosiglitazone in db/db mice and effect on non-fasted plasma glucose. Post-prandial blood glucose levels (measured in
the morning, after switch from dark to light cycle) in db/db mice. Data are expressed as mean SEM, N = 8 per group. Oral Glucose challenge performed in overnight fasted db/
db mice treated chronically. Last drug was administered (per gavage) on day prior to challenge. Data are expressed as mean SEM. N = 6/group.

2472
A. Bénardeau et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2468–2473
22
20
18
16
14
12
10
8
Zucker fa/fa rat
7days treatment
Relative HOMA-Insulin Resistance
1
6
4
2
cont
RSG GW2570 (S)-2Aa
0
3
10
1 mkd
-20
0
20
40
60
80
100
120
Plasma
2677
776
773 ng/ml
Time (minutes)
Figure 4. Euglycemic clamp study and insulin resistance index in fa/fa rats. GIR measured in fa/fa rats submitted to Euglycemic-Hyperinsulinemic clamp; data are expressed
as mean values of N = 5 per group. SEM have not been included in the graph. HOMA-IR index was calculated from fasting plasma insulin and fasting blood glucose (FBG) levels
measured at time 0 of the oGTT (prior to glucose challenge) in fa/fa rats. HOMA-IR index = (Fasting insulin (mU/ml) ꢁ FBG (mM))/22.5.
fect, directly compared to 3 mg/kg/d of Rosiglitazone (RSG) and
10 mg/kg/d of Farglitazar (GW2570), respectively. The right panel
shows the insulin-resistance indices25,26 as determined in the same
model. Again, (S)-2Aa behaved very positively.
Last but not least, (S)-2Aa was also tested in different animal
models of dyslipidemia. Figure 5 illustrates the data generated in
human ApoAI-transgenic mice.27 A robust increase of the amount
of HDLc was observed, whereas the predominantly PPARc agonist
HDL
FF: 300 mkd
(S)-2Aa: 3mkd
(S)-2Aa: 0.3mkd
100
HDL
control
control
75
Farglitazar: 10 mkd
(S)-2Aa: 0.3 mkd
(S)-2Aa: 3.0 mkd
50
25
0
Fenofibrate: 300 mkd
VLDL
LDL
Peak shifted to larger HDL
30.0
40.0
50.0
60
Figure 5. Effects on HDLc in human ApoAI-transgenic mice. Effects on HDLc measured in human ApoAI-transgenic mice after 12 days of treatment. Plasma lipid levels were
measured by FPLC method from pooled plasma (N = 8 per group); data expressed as averaged values.
alpha-human
alpha-rat
alpha-mouse
EC50 ( M)
EC50 ( M)
EC50 ( M)
µ
µ
µ
O
O
O
O
N
0.050
2.26
2.34
O
S
( S) - 2Aa
Figure 6. Species-dependence of PPAR
a
affinity of (S)-2Aa.

A. Bénardeau et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2468–2473
2473
18. The coordinates of the structures of compound (S)-2Aa crystallized with the
human PPAR and ligand domains were deposited to the Protein Data Bank
(PDBid: PPAR -3G8I, PPAR -3G9E).
19. Co-crystals of PPAR -LBD with (S)–2Aa were obtained using an identical
Farglitazar had no effect in this model. The classical PPAR
a drug
a
a
c
Fenofibrate was found to be even more active, albeit at 100–1000
times higher doses. The right panel shows the drug-induced size
shift of the HDL-particles.
At first glance, this lipid-lowering effect does not seem to be too
impressive. However, one has to keep in mind that compounds 2–4
exhibit in general intriguing species selectivity with respect to
c
c
protocol as described in Burgermeister, E. et al. Mol. Endocrinol. 2006, 20, 809.
Data have been collected in-house on a rotating anode (k = 1.5418 Å) to a
maximum resolution of 2.3 Å. Crystals belong to the orthorhombic space group
P212121 with cell axes a = 53.9, b = 70.0, c = 88.5 Å. The structure was
determined by molecular replacement using the chain A of the pdb entry
1PRG and subsequently refined. Difference electron density was used to place
the ligand by real space refinement.
PPARa. Figure 6 exemplifies this phenomenon for (S)-2Aa. Func-
tional affinity drops almost two orders of magnitude with respect
to the human receptor. Taking this into account, the observed
in vivo effect is still remarkable; and the anticipation that the high-
er potency in primates will also translate into a much stronger
pharmacodynamic effect turned out to be true. In (pre)diabetic
rhesus monkeys, supposed to be one of the most predictive animal
models, (S)-2Aa proved to be extraordinarily efficacious.28 Based
on these highly promising data (S)-2Aa was eventually selected
as clinical candidate. Under the USAN name Aleglitazar it recently
completed successfully clinical phase II studies.
20. The corresponding phenyl compound, albeit in racemic form, exhibits, for
example, an EC50 value of 0.172 lM at the alpha receptor.
21. Selected PK data: total clearance: 6.2 ml/min/kg (rat), 1.6 ml/min/kg
(cynomolgus monkey); bioavailability: 70% (rat), 68% (cynomolgus monkey);
half-life: 4 h (rat), 12.9 h (cynomolgus monkey).
22. Preparation of (S)-2Aa: (a) (S)-4-Benzyl-3-(2-methoxy-acetyl)-oxazolidin-2-
one (16.97 g, 68.1 mmol) was dissolved under a stream of Argon in 120 mL of
CH2Cl2, treated with 11.4 mL of triethylamine (1.2 equiv), and cooled down to
ꢀ78 °C. 75 mL of 1 M nBu2BOTf in of CH2Cl2 was then slowly added, the
reaction mixture warmed to 0 °C, and recooled after 50 min to ꢀ78 °C. 4-[2-(5-
Methyl-2-phenyl-oxazol-4-yl)-ethoxy]-benzo[b]thiophene-7-carbaldehyde
(24.75 g, 68.1 mmol), dissolved in 300 mL of CH2Cl2, was then added within 2 h
while keeping the temperature carefully at ꢀ78 °C. After another 30 min, the
temperature was raised to 0 °C and the mixture stirred for 1 additional h.
Quenching with icewater, extracting with AcOEt, washing with water and
brine, drying over MgSO4, and evaporation of all solvents, followed by flash
chromatography (SiO2, AcOEt/hexane = 1/1), yielded 95.87 g of (S)-4-benzyl-
3-((2S,3R)-3-hydroxy-2-methoxy-3-{4-[2-(5-methyl-2-phenyl-oxazol-4-yl)-
ethoxy]-benzo[b]thiophen-7-yl}-propionyl)-oxazolidin-2-one as yellow foam,
slightly contaminated with another diastereomer; ISP MS: 613.1 (M+H+),
635.0 (M+Na+). (b) (S)-4-Benzyl-3-((2S,3R)-3-hydroxy-2-methoxy-3-{4-[2-(5-
methyl-2-phenyl-oxazol-4-yl)-ethoxy]benzo[b]thiophen-7-yl}-propionyl)-
oxazolidin-2-one (40.0 g, 65.3 mmol) was dissolved in 100 mL of
trifluoroacetic acid and treated at 0 °C with triethylsilane (50 mL, 4.8 equiv).
The reaction was then allowed to proceed for 18 h at ambient temperature.
Pouring onto crashed ice/NaHCO3, extracting with AcOEt, washing with water
and brine, drying over MgSO4, and evaporation of the solvents leaved the crude
product which was purified by flash chromatography (SiO2, AcOEt/hexane = 3/
7) to give 29.0 g of (S)-4-benzyl-3-((S)-2-methoxy-3-{4-[2-(5-methyl-2-
phenyl-oxazol-4-yl)-ethoxy]-benzo[b]thiophen-7-yl}-propionyl)-oxazolidin-
2-one as white foam; EI MS: 596.4 (M+), 564.3 (MꢀMeOH+); 1H NMR
Acknowledgments
The authors gratefully acknowledge Christian Bitsch, Stefan
Bürli, and Thomas Burger for chemical syntheses, Angele Flament
and Astride Schnoebelen for in vitro testing, Urs Sprecher and Phi-
lippe Verry for in vivo testing, Bernard Gsell and Martine Stihle for
protein preparation and crystallisation efforts, Dr. Manfred Kansy
for physicochemical and Dr. Beate Bittner for in vitro toxicological
and pharmacokinetical investigations and for in vivo screens.
References and notes
1. Zhang, Z.; Burch, P. E.; Cooney, A. J.; Lanz, R. B.; Pereira, F. A.; Wu, J.; Gibbs, R. A.;
Weinstock, G.; Wheeler, D. A. Genome Res. 2004, 14, 580.
2. Steinmetz, A. C.; Renaud, J. P.; Moras, D. Annu. Rev. Biophys. Biomol. Struct. 2001,
30, 329.
3. van Raalte, D. H.; Li, M.; Pritchard, P. H.; Wasan, K. M. Pharm. Res. 2004, 21,
1531.
(300 MHz, CDCl3):
d 2.40 (s, 3H), 2.78 (dd, J = 9.6, 13.5 Hz, 1H), 3.05 (t,
J = 6.6 Hz, 2H), 3.25–3.31 (m, 3H), 3.41 (s, 3H), 3.91 (t, J = 8.3 Hz, 1H), 4.07 (dd,
J = 2.4, 9.0 Hz, 1H), 4.37 (t, J = 6.6 Hz, 2H), 4.37–4.44 (m, 1H), 5.43 (t, J = 6.5 Hz,
1H), 6.75 (d, J = 8.1 Hz, 1H), 7.17–7.32 (m, 7H), 7.40–7.46 (m, 4H), 7.97 (br d,
J = 7 Hz, 2H). (c) (S)-4-Benzyl-3-((S)-2-methoxy-3-{4-[2-(5-methyl-2-phenyl-
oxazol-4-yl)-ethoxy]-benzo[b]thiophen-7-yl}-propionyl)-oxazolidin-2-one
(29.0 g, 48.6 mmol) was dissolved in 340 mL of THF and treated at 0 °C with
118 mL of 1 N NaOH (2.5 equiv) and the reaction allowed to proceed for 1 h at
ambient temperature, when TLC indicated the absence of starting material. The
reaction mixture was poured onto crashed ice and extracted twice with EtOEt
to remove the chiral auxiliary. The aqueous layer was then acidified with HCl
to pH1 and extracted again with AcOEt. The organic layer was washed with
water, dried over MgSO4, and the volume reduced i. V. to induce crystallisation.
The first crop, obtained at ꢀ10 °C, yielded after washing with hexane and
drying 17.39 g of (S)-2-methoxy-3-{4-[2-(5-methyl-2-phenyl-oxazol-4-yl)-
ethoxy]-benzo[b]thiophen-7-yl}-propionic acid ((S)-2Aa) as white crystals;
mp 146–47 °C; C24H23N1O5S1 theory%: C, 65.89; H, 5.30; N, 3.20; S, 7.33.
Found: C, 65.53; H, 5.52; N, 3.23; S 7.22; HR-MS: calcd 436.12242, found
436.12222; HPLC (Chiralpak-AD, 25 cm ꢁ 4.6 mm, 90% heptane/10% EtOH/0.2%
TFA): 99.4%; 1H NMR (400 MHz, CDCl3): d 2.40 (s, 3H), 3.06 (t, J = 6.4 Hz, 2H),
3.21(dd, J = 7.8, 14.5 Hz, 1H), 3.35 (dd, covered, 1H), 3.34 (s, 3H), 4.20 (dd,
J = 4.8, 7.6 Hz, 1H), 4.35 (t, J = 6.6 Hz, 2H), 6.73 (d, J = 8.0 Hz, 1H), 7.15 (d,
J = 8.0 Hz, 1H), 7.32 (d, J = 5.6 Hz, 1H), 7.39–7.47 (m, 3H), 7.48 (d, J = 5.6 Hz,
1H), 7.97 (br d, J = 8 Hz, 2H); ISN MS: 436.3 (MꢀHꢀ); 13C NMR (150 MHz,
CDCl3): d (ppm) 10.28, 26.16, 37.47, 58.75, 66.85, 79.99, 105.44, 121.32, 123.04,
124.21, 126.07, 126.10, 127.38, 128.73, 130.04, 130.50, 132.62, 141.56, 145.15,
153.26, 159.73, 174.78.
4. Feige, J. N.; Gelman, L.; Michalik, L.; Desvergne, B.; Wahli, W. Prog. Lipid Res.
2006, 45, 120.
5. Cheng, A. Y. Y.; Leiter, L. A. Diabetes Obes. Metab. 2008, 10, 691.
6. Leff, T.; Mathews, S. T.; Camp, H. S. Exp. Diabesity Res. 2004, 5, 99.
7. Fredenrich, A.; Grimaldi, P. A. Diabetes Metab. 2005, 31, 23.
8. Kuhn, B.; Hilpert, H.; Benz, J.; Binggeli, A.; Grether, U.; Humm, R.; Maerki, H. P.;
Meyer, M.; Mohr, P. Bioorg. Med. Chem. Lett. 2006, 16, 4016.
9. Goldstein, B. J.; Rosenstock, J.; Anzalone, D.; Tou, C.; Ohman, K. P. Curr. Med. Res.
Opin. 2006, 22, 2575.
10. Kendall, D. M.; Rubin, C. J.; Mohideen, P.; Ledeine, J.-M.; Belder, R.; Gross, J.;
Norwood, P.; O’Mahony, M.; Sall, K.; Sloan, G.; Roberts, A.; Fiedorek, F. T.;
DeFronzo, R. A. Diabetes Care 2006, 29, 1016.
11. Goto, Y.; Yamazaki, M.; Hamana, M. Chem. Pharm. Bull. 1971, 19, 2050.
12. Puentener, K.; Scalone, M. US 2005070714 2005070714, 2005.
13. Hamad, A.-S. S.; Schinzer, D. Phosphorus, Sulfur Silicon Relat. Elem. 2000, 158,
187.
14. Haigh, D.; Birrell, H. C.; Cantello, B. C. C.; Eggleston, D. S.; Haltiwanger, R. C.;
Hindley, R. M.; Ramaswamy, A.; Stevens, N. C. Tetrahedron: Asymmetry 1999,
10, 1353.
15. Gage, J. R.; Evans, D. A. Org. Synth. 1990, 68, 83.
16. Selected EC50 values (effects are reported in relation to (2-methyl-4-[4-methyl-
2(4-trifluoromethyl-phenyl)-thiazol-5-
ylmethylsulfanyl]-phenoxy)-acetic
M (22%),
M (29%),
M (44%).
a-LBD with (S)-2Aa were obtained using a similar protocol
acid (GW 501516) whose activity was set to 100%): (S)-2Aa: 0.053
l
l
23. Lestradet, H. Ann. Med. Int. 1980, 131, 103.
24. Koopmans, S. J.; Radder, J. K. Biol. Med. Rev. 1996, 5, 31.
25. Matthews, D. R.; Hosker, J. P.; Rudenski, A. S.; Naylor, B. A.; Treacher, D. F.;
Turner, R. C. Diabetologia 1985, 28, 412.
26. Duncan, M. H.; Singh, B. M.; Wise, P. H.; Carter, G.; Alaghband-Zadeh, J. Lancet
1995, 346, 120.
27. Fournier, N.; Atger, V.; Paul, J. L.; Moatti, N. Ann. Biol. Clin. (Paris) 1995, 53, 209.
28. Data will be published in due course.
29. Binggeli, A.; Boehringer, M.; Grether, U.; Hilpert, H.; Maerki, H.-P.; Meyer, M.;
Mohr, P.; Ricklin, F., Wo 2002092084, 2002.
(S)-2Ba: 0.117
(S)-2Ae: 0.678
l
M (100%), (S)-2Ca: 0.045
l
M (93%), (R)-2Aa: 2.71
lM (70%), (S)-3Ba: 0.169
l
M (64%), (S)-4Aa: 0.137 l
17. Co-crystals of PPAR
as described in Burgermeister, E. et al. Mol. Endocrinol. 2006, 20, 809. Data have
been collected in-house on a rotating anode (k = 1.5418 Å) to a maximum
resolution of 2.2 Å. Crystals belong to the orthorhombic space group P212121
with cell axes a = 42.4, b = 76.1, c = 98.5 Å. The structure was determined by
molecular replacement using the chain
A of the pdb entry 1K7l and
subsequently refined. Difference electron density was used to place the
ligand by real space refinement.