Design, synthesis, and structure-activity relationships of piperidine and dehydropiperidine carboxylic acids as novel, potent dual PPARα/γ agonists

Article abstract of DOI:10.1016/j.bmcl.2008.05.014

Several series of substituted dehydropiperidine and piperidine-4-carboxylic acid analogs have been designed and synthesized as novel, potent dual PPARα/γ agonists. The SAR of these series of analogs is discussed. A rare double bond migration occurred during the basic hydrolysis of the α,β-unsaturated dehydropiperidine esters 12, and the structures of the migration products were confirmed through a series of 2D NMR experiments.

Full text of DOI:10.1016/j.bmcl.2008.05.014

Bioorganic & Medicinal Chemistry Letters 18 (2008) 3545–3550
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis, and structure–activity relationships of piperidine and
dehydropiperidine carboxylic acids as novel, potent dual PPARa/c agonists
b
c
c
c
d
Xiang-Yang Ye a, , Yi-Xin Li , Dennis Farrelly , Neil Flynn , Liqun Gu , Kenneth T. Locke ,
*
Jonathan Lippy d, Kevin O’Malley d, Celeste Twamley d, Litao Zhang d, Denis E. Ryono a,
Robert Zahler a, Narayanan Hariharan c, Peter T. W. Cheng a,
*
a Metabolic Diseases Chemistry, Bristol-Myers Squibb R&D, PO Box 5400, Princeton, NJ 08543-5400, USA
b Discovery Analytical Sciences, Bristol-Myers Squibb R&D, PO Box 5400, Princeton, NJ 08543-5400, USA
c Metabolic Diseases Biology, Bristol-Myers Squibb R&D, PO Box 5400, Princeton, NJ 08543-5400, USA
d Lead Evaluation, Bristol-Myers Squibb R&D, PO Box 5400, Princeton, NJ 08543-5400, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Several series of substituted dehydropiperidine and piperidine-4-carboxylic acid analogs have been
designed and synthesized as novel, potent dual PPARa/c agonists. The SAR of these series of analogs is
discussed. A rare double bond migration occurred during the basic hydrolysis of the a,b-unsaturated
dehydropiperidine esters 12, and the structures of the migration products were confirmed through a ser-
ies of 2D NMR experiments.
Received 21 March 2008
Revised 1 May 2008
Accepted 2 May 2008
Available online 6 May 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Peroxisome proliferator-activated receptors
Piperidine and dehydropiperidine
carboxylic acids
Structure–activity relationships
Dual agonists
Peroxisome proliferator-activated receptors (PPARs) belong to a
nuclear hormone receptor superfamily which act as transcription
factors in the regulation of genes involved in glucose and lipid
homeostasis.1 Three PPAR subtypes have been identified: PPARa,
PPARc and PPARd. PPARa is highly expressed in liver, but is also
present in heart, kidney, and muscle, and regulates the transcrip-
tion of numerous genes encoding proteins involved in lipid metab-
olism.2 Currently marketed PPARa agonists are the fibrate class of
drugs (including fenofibrate3 and gemfibrozil4), which elevate HDL
levels and lower triglyceride and LDL levels. PPARc is predomi-
nantly expressed in adipose tissue and regulates insulin sensitivity,
glucose and fatty acid utilization as well as adipocyte differentia-
tion.5 Currently marketed drugs targeting PPARc are rosiglitazone6
and pioglitazone,7 both for the treatment of type 2 diabetes. We
have recently reported the design and synthesis of Muraglitazar
(1, Fig. 1), a dual PPARa/c agonist which has shown excellent effi-
cacy in animal models of type 2 diabetes and the associated dysl-
ipidemia as well as in human clinical trial.8 Starting from the
oxybenzyl-glycine template exemplified by 1, several series of po-
tent, conformationally constrained dual PPARa/c agonists have
been designed and synthesized.9 From this effort, a series of 3-ben-
zyl-4-carboxy-pyrrolidines, exemplified by compound 2, was iden-
tified which has good activity both in vitro and in vivo.10 To further
explore the scope of the SAR of this pyrrolidine acid series, we
decided to investigate the effect of expanding the pyrrolidine ring
of 2 into the corresponding piperidine acids (Fig. 2). This paper fo-
cuses on the design, synthesis, and SAR of various piperidine/dehy-
dropiperidine carboxylic acid analogs as dual PPARa/c agonists.
Chemistry. Based on results from the pyrrolidine acid series and
preliminary modeling studies, we focused our initial efforts on the
dehydropiperidine acid analogs 3 (with both 1,4- and 1,3-oxyben-
zyl substituents). These compounds should be readily derived from
dehydropiperidine intermediates 4 (Scheme 1). Our initial efforts
to synthesize 4 (especially where m = 1) were unsuccessful. For in-
stance, in one approach, the enolate of N-Boc-4-oxopiperidine was
successfully alkylated with 4-methoxybenzyl bromide; however,
subsequent attempts to convert the carbonyl group into the C-4
carboxylic acid (via carbonylation of the enol triflate of the ketone)
were unsuccessful. Recently, Bellina11 reported that vinyl nona-
flates (derived from b-keto esters) act as excellent electrophiles
in palladium-catalyzed cross-coupling reactions using a variety of
organozinc halides. This report encouraged us to apply this
approach to the successful synthesis of 4 outlined in Scheme 1.
Vinyl nonaflate 6 was easily obtained from N-benzyl-3-oxopiperi-
dine-4-carboxylate by treatment with base (NaH or KHMDS),
* Corresponding authors. Tel.: +1 609 818 4130; fax: +1 609 818 3460.
E-mail addresses: xiang-yang.ye@bms.com (X.-Y. Ye), peter.t.cheng@bms.com
(P.T.W. Cheng).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.05.014

3546
X.-Y. Ye et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3545–3550
CO2H
4
3
O
O
N
CO2H
N
O
N
N
O
O
O
N
N
Compound 2
(3R, 4S)
αEC50 = 165 nM
γEC50 = 7 nM
CF3
Muraglitazar (1)
αEC50 =1400 nM
γEC50 = 40 nM
OCH3
Figure 1. Structure of Muraglitazar (1) and the related pyrrolidine-acid analog 2.
CO2H
n
CO2H
Ar
ring expansion
m
Ar
O
O
N
R
N
R
double bond or
saturated
Pyrrolidine acid
m= 0, 1, 2 n = 0, 1
Dehydropiperidine/Piperidine acid
Figure 2. The design of piperidine/dehydropiperidine carboxylic acids as dual PPARa/c agonists.
nyl series (bearing a variety of carbon linkages where m = 0, 1 or 2)
linker
CO2Et
CO2H
by simply altering the organozinc halides 5. Organozinc reagents 5
which were not commercially available were prepared either: (1)
in situ via the exchange of the Grignard reagent or organolithium
with ZnCl2 or (2) from reaction of the alkyl or aryl halide with acti-
vated zinc.
O
Ph
m
m
N
PO
O
N
N
R
Bn
phenyl oxazole
PO
4
3
To illustrate the versatility of this synthetic route, the synthesis
of the 1,3-oxybenzylpiperidine acid analogs 13 is shown in Scheme
2. The Pd-catalyzed cross-coupling reaction of vinyl nonaflate 711
and 3-(tert-butyldimethylsilyloxy)benzylzinc chloride12 afforded
8 in 75% yield. Deprotection of the N-benzyl group with a-chloro-
ethyl chloroformate13 was achieved with concomitant desilylation
of the phenol. N-Boc reprotection of the piperidine afforded phenol
9, which was alkylated with mesylate 10 and then deprotected to
give the key intermediate secondary amine 11, which enabled us to
quickly explore SAR at the dehydropiperidine nitrogen. Intermedi-
ates 11 were treated with alkyl/aryl chloroformates to give carba-
mates, or with substituted 2-chloropyrimidines to form N-
pyrimidines. The desired carboxylic acid analogs 13 were obtained
via acid-mediated ester hydrolysis.
ONf
CO2Et
ZnCl
m
+
Bn
N
5
6
m = 0, 1, 2 P = protecting groups; Nf = CF3CF2CF2CF2SO2-
Scheme 1. Retrosynthetic analysis of dehydropiperidine-4-carboxylate 3.
followed by nonafluoro-1-butanesulfonyl fluoride (Nf-F). Without
further purification, nonaflate 6 was reacted with zinc halides 5
in the presence of catalytic Pd(dba)2/dppf to afford 4 in 70–80%
yield in two steps. This convergent approach allowed us to quickly
access the desired intermediates 4 in both the 1,3- and 1,4-oxyphe-
CO2Et
CO2Et
ONf
TBSO
HO
a
b
Bn
N
CO2Et
N
N
Bn
Boc
8
7
9
CO2Et
CO2Et
RO
RO
d or e
c
f
N
H
N
R1
11
12
CO2H
O
RO
Ph
R =
N
R1 = benzyl, alkoxy carbonyl,
phenoxy carbonyl, or substituted
2-pyrimidinyl
N
R1
13
Scheme 2. The synthesis of 1,3-oxybenzyl dehydropiperidine analogs 13. Reagents and conditions: (a) 3-(tert-butyldimethylsilyloxy)benzylzinc chloride, Pd(dba)2, dppf, THF,
65 °C, 12 h, 75%; (b) (i) a-chloroethyl chloroformate, 50 °C, 1 h; then MeOH; (ii) Boc2O, THF-NaHCO3 (aq), 2 h, 72% (two steps); (c) (i) 2-(5-methyl-2-phenyloxazol-4-yl)ethyl
methanesulfonate (10), K2CO3, CH3CN, 90 °C, 12 h, 90%; (ii) 4 N HCl in dioxane, MeOH, rt, 5 h, 95%; (d) THF–H2O, NaHCO3, alkyl or aryl chloroformate, 1 h, 90–95%;
(e) 2-chloropyrimidine, i-Pr2NEt, toluene, 100 °C, 3 h, 85–90%; (f) 15% HCl in HOAc, sealed tube, 100 °C, 8 h, 70–80%.

X.-Y. Ye et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3545–3550
3547
Analogs with different linkers to the dehydropiperidine ring
(e.g. 1,4-oxyphenethyl [14], 1,4-oxyphenyl [15], 1,3-oxyphenyl
[16] and 1,4-oxybenzyl [17]) were also readily synthesized using
this general route, as various organozinc chlorides could readily
be substituted for the exemplified 3-(tert-butyldimethylsiloxy)
benzylzinc chloride shown in Scheme 2. In all cases, the syntheses
worked well and multiple analogs from each of these series (N-
benzyl, N-alkoxycarbonyl, N-pyrimidinyl) were prepared (Fig. 3).
We then explored the synthesis of the corresponding piperidine
acid analogs. This was unexpectedly problematic, as hydrogenation
of C@C bond of the fully elaborated tetrasubstituted a,b-unsatu-
rated esters (i.e. 12) was accompanied by varying degrees of reduc-
tion of the phenyloxazole moiety (to the cyclohexyloxazole).
Therefore, we needed to address the reduction of the C@C bond
prior to the coupling with the phenyloxazole ‘left-hand-side’. The
successful alternative synthesis for these piperidine acid analogs
is shown in Scheme 3, and is exemplified for the 1,3-oxyphenyl
series; analogs 22–25 were also synthesized using this general
route.14
Based on previous experience in the pyrrolidine acid series,
increasing the distance between the piperidine ring and the car-
boxylic acid might improve both PPAR binding affinity and func-
tional activity. Therefore, the homologated dehydropiperidine
carboxylic acid analogs 31 and their corresponding piperidine ana-
logs 32 were synthesized as shown in Scheme 4.15 The key se-
quence is the conversion of 27 to 28, which involves the
palladium-catalyzed CO carbonylation of the allylic bromide de-
rived from 27.16 The rest of the synthetic route is identical to
Scheme 1. Piperidine analogs 32 were also synthesized using a
combination of the chemistry shown in Schemes 2 and 3.
RO
RO
CO2H
CO2H
CO2H
CO2H
RO
RO
N
R1
N
R1
N
R1
N
R1
17
14
16
15
R1 = benzyl, alkoxy carbonyl,
O
phenoxy carbonyl, or substituted
2-pyrimidinyl
Ph
R =
N
Figure 3. 1,4-Oxyphenethyl, 1,4-oxyphenyl, 1,3-oxyphenyl, and 1,4-oxybenzyl dehydropiperidine carboxylic acid analogs.
CO2H
CO2Et
CO2Et
CO2Et
f or g
d,e
a-c
RO
RO
HO
MeO
then h
N
R1
N
H
N
N
Boc
Bn
21
20
19
18
O
N
CO2H
CO2H
Ph
R =
RO
m
RO
N
R1
R1 = benzyl, alkoxy carbonyl,
phenoxy carbonyl, or substituted
2-pyrimidinyl
N
R1
24
m = 2
m = 0
m = 1
22
23
25
Scheme 3. The synthesis of piperidine carboxylic acids. Reagents and conditions: (a) 10% Pd-C, H2 (80 psi), HOAc, overnight; (b) BBr3 (3.5 equiv), CH2Cl2, À78 to 0 °C, 2 h; (c)
Boc2O, THF–H2O, NaHCO3, 2 h, 51% (three steps); (d) 2-(5-methyl-2-phenyloxazol-4-yl)ethanol (26), Bu3P@CH–CN; toluene, 50 °C, 2 h, 95%; (e) AcCl/MeOH, 0 °C to rt, 5 h,
99%; (f) THF–H2O, NaHCO3, alkyl or aryl chloroformate, 1 h, 90–95%; (g) 2-chloropyrimidine, i-Pr2NEt, toluene, 100 °C, 3 h, 85–90%; (h) 15% HCl in HOAc, sealed tube, 100 °C,
8 h, 70–80%.
CO2Et
CO2Et
CO2Et
CO2Et
e or f
a-c
d
RO
RO
RO
RO
N
N
N
H
N
R1
Boc
Boc
28
27
29
30
CO2H
CO2H
O
Ph
g
R =
R1 = benzyl, alkoxy carbonyl,
N
RO
RO
N
R1
N
R1
phenoxy carbonyl, or substituted
2-pyrimidinyl
31
32
Scheme 4. The synthesis of homologated piperidine carboxylic acids. Reagents and conditions: (a) LiAlH4, THF, À78 to 10 °C, 3 h, 95%; (b) PPh3, CBr4, CH2Cl2, 0 °C to rt, 2 h,
93%; (c) Pd(PPh3)4, KHCO3, MeOH, CO (100 psi), rt, overnight, 61%; (d) AcCl/MeOH, 0 °C to rt, 5 h, 95%; (e) THF–H2O, NaHCO3, alkyl or aryl chloroformate, 1 h, 90–95%;
(f) 2-chloropyrimidine, i-Pr2NEt, toluene, 100 °C, 3 h, 85–90%; (g) 15% HCl in HOAc, sealed tube, 100 °C, 8 h, 70–80%.

3548
X.-Y. Ye et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3545–3550
In previous schemes, the final step of analog synthesis was acid-
(1,3- vs 1,4-). The SAR at the piperidine ring was explored primarily
by modification of substituents at the piperidine nitrogen, e.g. N-
benzyl, N-alkoxycarbonyl, N-phenoxycarbonyl and N-pyrimidinyl.
The PPARa and c binding affinity as well as the transactivation
(functional) data of these novel piperidine analogs are shown in
Table 1.
mediated ester hydrolysis. Under these conditions, the C@C bond
of the a,b-unsaturated acid did not undergo migration (Scheme
2). However, due to the need for milder conditions required by
more sensitive functionalities, we explored base-mediated hydro-
lysis (LiOH, THF–H2O, rt) of the penultimate ethyl ester intermedi-
ate with analogs from the 1,3-oxybenzyl series 12. Due to the
sluggishness of the basic hydrolysis, three drops of methanol were
added to accelerate the reaction, which then proceeded smoothly
to completion. Surprisingly, the desired a,b-unsaturated acid 13
was isolated only as the minor product. The major product had
the same molecular weight as 13, but its 1H NMR spectrum con-
tained a new vinyl proton (d 6.73); the possible structures were
thus the C@C migration products 33 or 34 (both of which bear a vi-
nyl proton). The ratio of the C@C migration product versus 13 var-
ied from 4:1 to 6:1, depending on the structure of R1.
A set of 2D NMR experiments were carried out on the C@C
migration product from the hydrolysis of 12a (R = -CO2-iBu). These
studies enabled us to unequivocally assign each proton and carbon
in this compound (Fig. 4)17 and to unambiguously demonstrate
that its structure corresponded to 33a rather than 34 (R1 = i-Bu–
O–C(O)À). In particular, the 1H–13C Heteronuclear Multiple Bond
Connectivity (HMBC) analysis proved to be very useful in this
structure determination. The most important correlations (as
shown in Fig. 4) are: (1) the vinyl proton H-24 with the carbonyl
C-29 and, (2) the benzylic H-22 with C-20, C-18, C-24, and C-28.
This unexpected base-mediated C@C bond migration from the
a,b-unsaturated acid to the b,c-unsaturated acid is rare. However,
this result is in agreement with the previous work of Jacobsen,18
in which a similar C@C migration occurred upon treatment of a
dehydropiperidine 4-ester with benzylamine. Several analogs
bearing the general structure 33 have been synthesized according
to Scheme 5.
In the 1,4-oxyaryldehydropiperidine acid series, the linker be-
tween the oxyaryl and the dehydropiperidine moieties clearly
had a significant effect upon PPAR activity. Both PPARc and a po-
tency increased with increasing linker length m, e.g. the isobutyl
carbamates 15a (oxyphenyl linker; m = 0) versus 17a (oxybenzyl
linker; m = 1). With the oxyphenethyl analog 14a (m = 2; aEC50
56 nM; cEC50 74 nM), the PPARc potency was increased by an-
other 15-fold relative to 17a while PPARa potency was main-
tained, resulting in equivalent, potent agonist activity at PPARa
and c. The phenethyl linker may offer more flexibility (versus
the benzyl or phenyl linkers) allowing these analogs to bind in
the ligand-binding domains (LBDs) of both PPARa and c. How-
ever, in the piperidine acid series, both the 1,3-oxybenzyl analog
25a and 1,4-oxybenzyl analog 24a (m = 1) show significantly less
activity at PPAR.c than at PPARa. Changing the linker length as
in the dehydropiperidine series [e.g. for the 1,3-oxyphenyl ana-
log 21a (m = 0) and the 1,4-oxyphenethyl analog 22a (m = 2)]
did not result in compounds with equipotent activity at PPARa
and c. An alternative way to modulate PPARa/c functional activ-
ity is to introduce flexibility to the location of the carboxylic
acid. For instance, the 1,3-oxyphenyl dehydropiperidinyl acetic
acid analogs 31a and 31b (where the 1,3-oxyphenyl group is di-
rectly attached to the dehydropiperidine ring, [i.e. m = 0] but the
carboxymethyl group has flexibility of movement), are potent
dual PPARa/c agonists. The analogous piperidines 32a and 32b
are much less active at both PPARa and c. In the 1,3-oxybenzyl
series, the dehydropiperidine 13b (and to a lesser extent 13a)
shows good binding affinity but poor functional activity. Finally
(and perhaps most significantly) the serendipitously discovered
b, c-unsaturated acid analogs (33a–d) generally showed excel-
lent binding and functional activity at both PPARa and c. Com-
pounds 33a–d are significantly more potent at PPARc than the
corresponding a, b-unsaturated acid analogs 13a–d, and in the
cases of 33a and 33c, also show significantly more potent PPARa
activity.
Results and discussion. Following the versatile synthetic schemes
outlined above, we were able to quickly generate analogs which
explored the optimal linker length between the phenyloxazole
moiety and the dehydropiperidine/piperidine ring (m = 0–2), as
well as the optimal substitution of the central oxyphenyl ring
O
OH
31
28
21
5
32
37
Overall, the SAR from these series of piperidine and dehydropi-
peridine analogs shows that PPARa/c binding affinity and func-
tional activity can be modulated by: (a) varying the oxyphenyl
substitution (1,3 vs 1,4 series); (b) the linker between the oxyphe-
nyl and the piperidine ring (m = 0–2), (c) dehydropiperidine versus
piperidines, and (d) the linker between the piperidine and the car-
boxylic acid. From the SAR exploration of various piperidine and
dehydropiperidine series, analogs from the 1,3- oxyphenyl dehyd-
ropiperidines (31a, b), the 1,4- oxyphenethyl dehydropiperidines
12
20
17
22
O
4
13
15
16
19
18
23
24
27
26
14
O
1
2
7
N
3
N
25
29
6
11
10
34
36
35
38
O
O
30
33
8
9
33a
Figure 4. Some important HMBC correlations of 33a.
CO2H
CO2Et
CO2H
O
N
RO
RO
RO
RO
Ph
a
R =
+
R1 = benzyl, alkoxy carbonyl,
phenoxy carbonyl, or substituted
2-pyrimidinyl
N
R1
N
R1
N
R1
13
33
12
CO2H
N
R1
34
Scheme 5. Double bond migration in base-catalyzed ester hydrolysis. Reagents and conditions: (a) LiOH, THF–H2O (2:1), trace amount of MeOH, rt, 3 h.

X.-Y. Ye et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3545–3550
3549
Table 1
In vitro activities of selected analogs at human PPARc and a
CO2H
n
CO2H
n
CO2H
N
RO
O
m
m
RO
RO
Ph
R =
N
R1
N
N
R1
R1
33
21 (m = 0, n = 0, 3'-OR)
13 (m = 1, n = 0, 3'-OR)
22 (m = 2, n = 0, 4'-OR)
24 (m = 1, n = 0, 3'-OR)
25 (m = 1, n = 0, 4'-OR)
32 (m = 0, n = 1, 3'-OR)
14 (m = 2, n = 0, 4'-OR)
15 (m = 0, n = 0, 4'-OR)
16 (m = 0, n = 0, 3'-OR)
17 (m = 1, n = 0, 4'-OR)
31 (m = 0, n = 1, 3'-OR)
Compound
R1
PPARc binding
IC50(lM)a
HEK PPARc EC50(lM)
(% activity @ 1 lM)b
PPARa binding
IC50 (lM)a
HEK human PPARa EC50 (lM)
(% activity @ 1 lM)b
1
15a
17a
14a
16a
13a
25a ( )
22a ( )
21a ( )
24a ( )
31a
Muraglitazar
i-BuO-C(O)-
i-BuO-C(O)-
i-BuO-C(O)-
Benzyl
i-BuO-C(O)-
i-BuO-C(O)-
i-BuO-C(O)-
i-BuO-C(O)-
i-BuO-C(O)-
i-BuO-C(O)-
PhO-C(O)-
0.19
7.4
3.036
0.11
>27.5
4.51
1.25
0.45
>27.3
6.03
0.049
0.034
3.089
7.692
1.384
0.04 (139%)
4.65 (6%)
1.12 (52%)
0.074 (95%)
>7.5
3.86 (37%)
0.257 (82%)
0.263 (79%)
>7.5
1.07 (58%)
0.018 (116%)
0.046 (110%)
0.49 (75%)
1.29 (29%)
0.644 (65%)
0.25
>27.3
1.17
0.255
>27.5
2.56
0.215
0.711
7.62
1.37
0.13
0.016
>15
8.12
1.41 (107%)
5.57 (2%)
0.028 (104%)
0.056 (95%)
>7.5
0.235 (66%)
0.005 (115%)
0.044 (101%)
0.209 (67%)
0.018 (101%)
0.006 (82%)
0.004 (87%)
0.58 (48%)
0.17 (62%)
0.089 (37%)
31b
32a ( )
32b ( )
13b
i-BuO-C(O)-
PhO-C(O)-
4-Trifluoromethyl-2-
pyrimidinyl
PhCH2O-C(O)-
n-PrO-C(O)-
i-BuO-C(O)-
4-Trifluoromethyl-2-
pyrimidinyl
PhCH2O-C(O)-
n-PrO-C(O)-
1.328
13c
13d
33a ( )
33b ( )
4.73
1.32
0.13
0.287
>3.5
>15
0.97
0.861
1.078
4.92 (23%)
0.04 (85%)
0.01 (114%)
0.023 (85%)
0.18 (87%)
0.009 (132%)
0.014 (100%)
33c ( )
33d ( )
0.133
0.088
0.024 (118%)
0.008 (150%)
0.654
0.499
0.016 (125%)
0.008 (109%)
a
Ref. 19.
Ref. 20.
b
(14a), and the 1,3-oxybenzyl b, c-unsaturated acids (33a–d) show
potent functional activity at both PPARa and c.
References and notes
1. For general reviews on PPARs: Torra, I. P.; Chinetti, G.; Duval, C.; Fruchart, J.-C.;
Staels, B. Curr. Opin. Lipidol. 2001, 12, 245; Evans, A. J.; Krentz, A. J. Drugs R&D
1999, 2, 75; Sternbach, D. D. Annu. Rep. Med. Chem. 2003, 38, 71; Cheng, P. T. W.;
Mukherjee, R. Minirev. Med. Chem. 2005, 5, 741.
2. Isseman, I.; Green, S. Nature (London) 1990, 347, 771; Lee, S. S.; Pineau, T.;
Drago, J.; Lee, E. L.; Owens, J. W.; Kroetz, D. L.; Fernandez-Salguero, P. M.;
Westphal, H.; Gonzalez, F. J. Mol. Cell. Biol. 1995, 15, 3012.
3. Staels, B.; Dallongeville, J.; Auwerx, J.; Schoonjans, K.; Leitersdorf, E.; Fruchart,
J.-C. Circulation 1998, 98, 2088.
4. Todd, P. A.; Ward, A. Drugs 1988, 36, 314–339.
5. Tontonoz, P.; Hu, E.; Spiegelman, B. M. Cell 1994, 79, 1147.
6. Momose, Y.; Meguro, K.; Ikeda, H.; Hatanka, C.; Oi, S.; Sohda, T. Chem. Pharm.
Bull. 1991, 39, 1440.
7. (a) Willson, T. M.; Brown, P. J.; Sternbach, D. D.; Henke, B. R. J. Med. Chem 2000,
43, 527; (b) Cantello, B. C. C.; Cawthorne, M. A.; Haigh, D.; Hindley, R. M.; Smith,
S. A.; Thurlby, P. J. Med. Chem 1994, 37, 3977.
8. (a) Devasthale, P. V.; Chen, S.; Jeon, Y.; Qu, F.; Shao, C.; Wang, W.; Zhang, H.;
Farrelly, D.; Golla, R.; Grover, G.; Harrity, T.; Ma, Z.; Moore, L.; Ren, J.; Seethala,
R.; Cheng, L.; Sleph, P.; Sun, W.; Tieman, A.; Wetterau, J. R.; Doweyko, A.;
Chandrasena, G.; Chang, S. Y.; Humphreys, W. G.; Sasseville, V. G.; Biller, S. A.;
Ryono, D. E.; Selan, F.; Hariharan, N.; Cheng, P. T. W. J. Med. Chem. 2005, 48,
2248; (b) Harrity, T.; Farrelly, D.; Tieman, A.; Chu, C.; Kunselman, L.; Gu, L.;
Ponticiello, R.; Cap, M.; Qu, F.; Shao, C.; Wang, W.; Zhang, H.; Fenderson, W.;
Chen, S.; Devasthale, P.; Jeon, Y.; Seethala, R.; Yang, W.-P.; Ren, J.; Zhou, M.;
Ryono, D.; Biller, S.; Mookhtiar, K. A.; Wetterau, J.; Gregg, R.; Cheng, P. T.;
Hariharan, N. Diabetes 2006, 55, 240; (c) 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.
Conclusions. We have devised efficient and flexible synthetic
routes (using as the key step the palladium-catalyzed cross-cou-
pling of nonaflates with organozinc halides) to several series of
piperidine and dehydropiperidine carboxylic acids. Analogs from
several of these series were identified which show potent agonist
activity at both PPARa and c (e.g. 31b, 33a, b). One of the dehydro-
piperidine series (33) resulted from a C@C bond migration during
base-mediated hydrolysis of the penultimate a, b-unsaturated es-
ter intermediate. The structure of 33 was confirmed through a ser-
ies of 2D NMR experiments. In the course of these SAR studies, we
also identified PPARa-selective agonists (e.g. 17a and 25a). These
different varieties of analogs should serve as useful leads in our
continuing effort to explore the utility of dual PPARa/c agonists.
Acknowledgment
We thank the BMS Discovery Analytical Sciences Department
for analytical support.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2008.05.014.
9. (a) Devasthale, P. V.; Chen, S.; Jeon, Y.; Qu, F.; Ryono, D.; Wang, W.; Zhang, H.;
Cheng, L.; Farrelly, D.; Golla, R.; Grover, G.; Ma, Z.; Moore, L.; Seethala, R.; Sun,

3550
X.-Y. Ye et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3545–3550
W.; Doweyko, A. M.; Chandrasena, G.; Sleph, P.; Hariharan, N.; Cheng, P. T. W.
HMBC. For example, in the HMBC spectrum, H-24 (d 6.73) showed a correlation
with C-29 (carbonyl, d 153.2), while the benzylic protons H-22 (d 3.36, doublet)
showed connectivity with C-23 (d 115.3), C-24 (d 123.7), C-28 (d 40.7), C-16 (d
116.0), C-17 (d 141.5), and C-18 (d 121.9, as shown in Fig. 4). Complete analysis
from this set of NMR experiments showed that the structure was
unequivocally 33a rather than 34a.
Bioorg. Med. Chem. Lett. 2007, 17, 2312; (b) Wang, W.; Devasthale, P. V.; Farrelly,
D.; Gu, L.; Harrity, T.; Cap, M.; Chu, C.; Kunselman, L.; Morgan, N.; Ponticiello, R.;
Zebo, R.; Zhang, L.; Locke, K.; Lippy, J.; O’Malley, K.; Hosagrahara, V.; Zhang, L.;
Kadiyala, P.; Chang, C.; Muckelbauer, J.; Doweyko, A. M.; Zahler, R.; Ryono, D.;
Hariharan, N.; Cheng, P. T. W. Bioorg. Med. Chem. Lett. 2008, 18, 1939.
10. Cheng, P. T. W.; Chen, S.; Devasthale, P.; Ding, C. Z.; Herpin, T. F.; Wu, S.; Zhang,
H.; Wang, W.; Ye, X.-Y. WO 2004004665 A2.
18. (a) Jacobsen, P.; Schaumburg, K.; Krogsgaard-Larsen, P. Acta Chem. Scand. Ser. B
1980, 34, 319; (b) Allan, R. D.; Fong, J. Aust. J. Chem. 1983, 36, 601.
11. Bellina, F.; Ciucci, D.; Rossi, R.; Vergamini, P. Tetrahedron 1999, 55, 2103.
12. 3-(tert-Butyldimethylsilyloxy)benzylzinc chloride was prepared freshly from
the corresponding benzyl chloride and activated zinc.
13. (a) Olofson, R. A.; Martz, J. T. J. Org. Chem. 1984, 49, 2081; (b) Campbell, A. L.;
Pilipauskas, D. R.; Khanna, I. K.; Rhodes, R. A. Tetrahedron Lett. 1987, 28, 2331;
(c) Yang, B. V.; O’Rourke, D.; Li, J. Synlett 1993, 195.
19. PPARc and PPARa Binding Assays (Fluorescence-Polarization) were conducted
in human PPARa and PPARc ligand binding domains using a fluorescene-
containing oxybenzylglycine derivative as ligand. Binding IC50 values (lM) for
PPARa and PPARc were determined by calculating the amount of test ligand
required for the half-maximal inhibition of the specific binding of fluorescein-
labeled analog of a potent dual PPARa/c activator to the PPARa or c LBD,
respectively. See: Seethala, R.; Golla, R.; Ma, Z.; Zhang, H.; O’Malley, K.; Lippy,
J.; Cheng, L.; Mookhtiar, K.; Farrelly, D.; Zhang, L.; Hariharan, N.; Cheng, P. T. W.
Anal. Biochem. 2007, 363, 263.
14. Acid-mediated ester hydrolysis was chosen because our previous experience
with the analogous pyrrolidine acid series (e.g. 2) had shown that base-
mediated hydrolysis can result in epimerization at the carbon
a to the
carboxylic acid (e.g. analogs 22–25 in Scheme 3).
15. Compound 27 was synthesized following the procedure in Scheme 2.
16. Schoenberg, A.; Bartoletti, I.; Heck, R. F. J. Org. Chem. 1974, 39, 3318.
20. In vitro PPAR agonist functional assays were performed by transiently
transfecting GAL4-hPPARa-LBD or GAL4-hPPARc-LBD constructs, respectively,
into HEK293 cells stably expressing 5Â GAL4RE-Luciferase. Data were
normalized for efficacy at 1 lM to known agonists. Agonist binding results in
an increase in luciferase enzyme activity which can be monitored by
measuring luminescence upon cell lysing and the addition of luciferin
substrate. EC50 values (lM) for PPARa or c agonist activity were calculated as
the concentration of the test ligand (lM) required for the half-maximal fold
induction of HEK293 cells. The ‘intrinsic activity’ of a test ligand is defined as
its activity at 1 lM (expressed as a percentage) relative to the activity of the
primary standards at 1 lM. Also see Ref. 9a.
17. The major product from the reaction (a b,c unsaturated acid) exists as a
mixture of rotamers in d6-DMSO (25 °C), which results in broad signals in the
1H NMR spectrum. To achieve better long range coherence, a set of 2D NMR
experiments were performed at elevated temperature (ꢀ100 °C), including
1H–1H COSY, 1H–13C HMQC (Heteronuclear Multiple Quantum Coherence),
1H–13C HMBC (Heteronuclear Multiple Bond Connectivity). Each proton and
carbon of the product was initially assigned through the use of 1H–1H COSY
and HMQC. All proton and carbon assignments were also confirmed using