Pyridine-3-propanoic acids: Discovery of dual PPARα/γ agonists as antidiabetic agents

Article abstract of DOI:10.1016/j.bmcl.2006.08.110

A series of novel pyridine-3-propanoic acids was synthesized. A structure-activity relationship study of these compounds led to the identification of potent dual PPARα/γ agonists with varied isoform selectivity. Based on the results of efficacy studies in

Full text of DOI:10.1016/j.bmcl.2006.08.110

Bioorganic & Medicinal Chemistry Letters 16 (2006) 6120–6123
Pyridine-3-propanoic acids: Discovery of dual PPARa/c agonists
as antidiabetic agents
a,
b
c
Paul S. Humphries,a,* Simon Bailey, Jonathon V. Almaden, Sandra J. Barnum,
Thomas J. Carlson,b Lance C. Christie,a Quyen-Quyen T. Do,a James D. Fraser,c
Mary Hess,c Jack Kellum,a Young H. Kim,c Guy A. McClellan,a
Kathleen M. Ogilvie,c Brett H. Simmons,c Donald Skalitzky,a
*
Shaoxian Sun,b David Wilhitea and Luke R. Zehndera
aDepartment of Diabetes Chemistry, Pfizer Global R&D, 10770 Science Center Drive, San Diego, CA 92121, USA
bDepartment of Biochemistry, Pfizer Global R&D, 10628 Science Center Drive, San Diego, CA 92121, USA
cDepartment of Diabetes Biology, Pfizer Global R&D, 10724 Science Center Drive, San Diego, CA 92121, USA
Received 9 May 2006; revised 23 August 2006; accepted 28 August 2006
Available online 14 September 2006
Abstract—A series of novel pyridine-3-propanoic acids was synthesized. A structure–activity relationship study of these compounds
led to the identification of potent dual PPARa/c agonists with varied isoform selectivity. Based on the results of efficacy studies in
diabetic (db/db) mice, and the desired pharmacokinetic parameters, compounds (S)-14 and (S)-19 were selected for further profiling.
Ó 2006 Elsevier Ltd. All rights reserved.
Type 2 diabetes is a metabolic disorder that accounts
for 120 million patients worldwide and the number is
likely to grow to greater than 200 million by the year
2010.1 This is a complex disease and invariably type
2 diabetic patients also display cardiovascular risk fac-
tors including hypertension and dyslipidemia.2,3 Both
Avandia (rosiglitazone) and Actos (pioglitazone) are
PPARc agonists and elicit their insulin sensitizing effect
through activating the PPARc nuclear receptor. The
clinical use of PPARc agonists in type 2 diabetes has
been plagued by mechanism based side effects including
weight gain, fluid retention, and edema. PPARa is the
molecular target for the fibrate class of lipid-modulat-
ing drugs.4 Designing compounds with PPARa activity
in addition to PPARc agonist activity may offer im-
proved alternatives toward control of hyperglycemia
and hypertriglyceridemia in type 2 diabetic patients.5
Scientists at Kyorin disclosed novel antidiabetic
KRP-297,6 the first published example of a dual
PPARa and PPARc agonist.7
In 1991, a series of PPAR analogues were disclosed,
which for the first time did not contain a thiazolidine-
2,4-dione pharmacophore.8 These were propanoic acid
derivatives with a substituent placed in the a-position
such that the whole group could mimic the thiazoli-
dine-2,4-dione ring. Based on the above and a knowl-
edge of PPAR ligands publicly disclosed, we wished to
synthesize compounds represented by the general struc-
ture 1 (Fig. 1). These efforts led to the identification of a
new class of potent dual PPARa/c agonists with excel-
lent in vivo efficacy. Herein, we report the synthesis,
structure–activity relationships (SAR), and in vivo activ-
ity of this new class of compounds.9
The synthetic route for the preparation of compounds
5–22 is shown in Scheme 1. Coupling of commercially
Keywords: PPAR; Diabetes; Carboxylic acid; SAR.
*
Corresponding authors. Tel.: +1 858 622 7956; fax: +1 858 526 4438
(P.S.H.); tel.: +1 858 526 4692; fax: +1 858 526 4315 (S.B.); e-mail
addresses: paul.humphries@pfizer.com; simon.bailey@pfizer.com
Figure 1. Thiazolidine-2,4-dione mimic and chosen lead scaffold 1.
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.08.110

P. S. Humphries et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6120–6123
6121
(R1 and R2), Table 1. Compounds 5–10 highlight mono-
alkyl derivatives, which demonstrated that binding affin-
ity and isoform selectivity are sensitive to the size of the
substituent. Propyl and iso-butyl both proved optimal
(compounds 7 and 9). Compound 11 showed that acy-
clic dialkyl substitution offered no benefits over mono-
alkyl (cf. compound 5). Compounds 12–15 highlight
acyclic alkoxy derivatives, with compounds 12–14 prov-
ing especially interesting. Compounds 16–22 all con-
tained cyclic a-substituents, with 16–18 being
cycloalkyl and 19–22 being cycloalkoxy. Of particular
note are compounds 17 and 19, which displayed excel-
lent potency in both the binding and transactivation as-
says. Differences between compounds 19 versus 20 and
21 versus 22 demonstrated that there was an optimal
positioning of the heteroatom in these cycloalkoxy
moieties.
Scheme 1. Reagents and conditions: (a) 2,5-dibromopyridine, KOtBu,
THF, reflux, 16 h, 83%; (b) nBuLi, Et2O, THF, ꢀ78 °C then DMF,
0 °C, 1 h, 72%; (c) NaBH4, MeOH, rt, 1 h, 100%; (d) (COCl)2, DMF,
CH2Cl2, rt, 3 h, 100%; (e) NaI, Me2CO, rt, 4 h, 75%; (f)
R1R2CHCO2Et, NaHMDS, THF, ꢀ50 °C, 3 h; (g) LiOH, THF,
MeOH, H2O, rt, 16 h.
available 2,5-dibromopyridine and alcohol 2, followed
by conversion of the 5-bromo for an aldehyde moiety,
afforded 3. Aldehyde 3 was reduced to the alcohol, con-
verted to the benzyl chloride and further transformed to
iodide 4 via a Finkelstein reaction. Reaction of 4 with a
variety of ester enolates allowed for an efficient variation
of a-substituents. The final step was ester hydrolysis,
giving final products 5–22.
The observed potency enhancement associated with the
2-tetrahydrofuran moiety (e.g., compound 19) prompted
us to hold this moiety constant and make changes to the
substituents on the phenyloxazole group (Table 2).
Extension of the linker between the pyridyl core and
the phenyloxazole moiety affected PPARc more than
PPARa, resulting in a more balanced dual agonist (com-
pound 26). Compounds 27–35 demonstrated that bind-
ing affinity and isoform selectivity were sensitive to the
positioning of substituents on the phenyl ring. 4-substit-
uents tended to afford compounds with a more balanced
isoform profile, whereas 3-substituents tended to impart
isoform selectivity favoring PPARc (e.g., compound 27
versus 28).
In an effort to look at the effects of substitution on the
phenyloxazole moiety, a new synthetic route was re-
quired (Scheme 2). Commercially available alcohol 23
was converted to the benzyl bromide, which was then
reacted with the enolate of ethyl tetrahydrofuran-2-car-
boxylate to afford 24. Alcohols 25, prepared in two
steps, were then coupled to chloropyridine 24 via the
method described by Buchwald.10 Finally, hydrolysis
of the esters gave products 26–35.
Finally, we shifted our attention to finding alternatives
to the phenyloxazole group. Figure 2 depicts compound
36 as an example of this effort (Table 2). The synthesis of
this class of compounds has been described previously
and further information will be reported elsewhere.9
The newly synthesized compounds were evaluated in the
PPAR SPA binding assay to ascertain c and a binding
affinity.11 The active analogs were also tested for func-
tional activity in a PPAR-GAL4 transactivation (TA)
assay, where EC50 values as well as percent maximal
activation were measured.12 Initially, we decided to
investigate the effects of varying the a-substituents of 1
The synthetic route for the preparation of compound
(S)-19 is shown in Scheme 3. Reaction of 4 with the
sodium enolate of oxazolidinone 37 (prepared from
tetrahydrofuran-2-carbonyl chloride and (4S)-4-benzyl-
1,3-oxazolidin-2-one) furnished a single diastereomer
13
(as determined by HPLC, 1H and C NMR).13 The
final step was hydrolysis of the chiral auxiliary, affording
compound (S)-19 in >95% ee (as determined by chiral
SFC). Compound (S)-19 was subsequently determined
to be the eutomer and thus predicted to have the S
configuration based on literature precedent.14
The synthetic route for the preparation of compound
(S)-14 is shown in Scheme 4. Following work by
Andersson,14 racemic 14 was successfully reacted with
(R)-phenylglycinol to afford a pair of diasteromeric
amides. The two diastereomeric amides were easily sep-
arated by flash column chromatography. The first elut-
ing product 38 (a single diastereomer as determined by
HPLC, 1H and 13C NMR) was a white solid and the sec-
ond eluting product (not shown) an oil. The final step
was hydrolysis of the chiral auxiliary, affording com-
pound (S)-14 in >95% ee (as determined by chiral
SFC). Compound (S)-14 was subsequently determined
Scheme 2. Reagents and conditions: (a) PBr3, 160 °C, 3 h, 44%; (b)
ester, NaHMDS, THF, ꢀ50 °C, 3 h, 78%; (c) Pd(OAc)2, racemic-2-(di-
tert-butylphosphino)-1,10-binaphthyl, Cs2CO3, PhMe, 115 °C, 16 h;
(d) LiOH, THF, MeOH, H2O, rt, 16 h.

6122
P. S. Humphries et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6120–6123
Table 1. In vitro activities of compounds 5–22
Compound
R1
R2
hPPARc
hPPARa
hPPARc TA EC50 (lM)
hPPARa TA EC50 (lM)
SPA Kia (lM)
SPA Kia (lM)
(% max activation)b
(% max activation)b
Rosiglitazone
5
0.44
2.1
45% at 100 lM
3.0
0.158 (100%)
0.268 (113%)
0.065 (99%)
0.454 (69%)
1.28 (87%)
1.65 (104%)
NT
>10 (6%)c
0.287 (93%)
0.15 (106%)
0.77 (92%)
0.426 (93%)
1.72 (98%)
NT
Me
Et
H
6
H
1.1
0.24
1.3
0.88
7
nPr
H
8
iPr
H
8.8
0.36
4.1
0.43
9
iBu
H
10
11
12
13
14
(S)-14
15
16
17
18
19
(S)-19
20
21
22
tBu
Me
H
12.0
5.1
9.4
6.8
Me
H
0.984 (72%)
0.013 (98%)
0.006 (105%)
0.232 (116%)
0.110 (97%)
2.81 (64%)
1.01 (80%)
0.125 (93%)
NT
0.534 (112%)
0.083 (95%)
0.033 (96%)
0.144 (106%)
0.111 (99%)
4.67 (64%)
6.15 (39%)
0.238 (94%)
NT
OMe
OEt
0.063
0.15
0.56
0.18
0.043
H
OMe
OMe
OEt
Me
Me
Et
0.33
0.57
0.27
14.4
15.1
Cyclopropyl
Cyclobutyl
Cyclohexyl
2-Tetrahydrofuran
2-Tetrahydrofuran
3-Tetrahydrofuran
2-Tetrahydropyran
4-Tetrahydropyran
7.98
0.394
3.5
44% at 100 lM
0.633
8.3
0.038
0.006
17.0
0.078
0.032
0.012 (89%)
0.014 (93%)
NT
0.048 (106%)
0.165 (89%)
NT
33% at 100 lM
0.078
0.081
32% at 100 lM
0.209 (103%)
NT
0.228 (96%)
NT
2% at 11 lM
a Binding affinities were measured using radioligands (darglitazone for PPARc and GW2331 for PPARa) following published procedures.12
b Agonist activities were measured in human PPAR-GAL4 chimeric HepG2 cells analogous to published procedures.13 The EC50 refers to the
concentration at which 50% of a given compounds’ intrinsic maximal response has been reached. % max activation refers to the level of maximal
activation achieved by a given compound when compared with the standard reference full agonists (darglitazone for PPARc and GW2331 for
PPARa).
c No EC50 was obtained; no plateau reached in titration; maximal activity only reported. NT, not tested
Table 2. In vitro activities of compounds 26–36
Compound
R1
n
hPPARc
SPA Ki (lM)
hPPARa
SPA Ki (lM)
hPPARc TA EC50 (lM)
(% max activation)
hPPARa TA EC50 (lM)
(% max activation)
26
27
28
29
30
31
32
33
34
35
36
H
2
1
1
1
1
1
1
1
1
1
0.81
1.84
NT
NT
4-Me
3-Me
4-Cl
0.023
0.016
0.052
0.011
0.094
0.015
0.017
0.007
0.334
0.015
0.031
0.277
0.014
0.099
0.021
0.100
0.065
0.215
0.314
0.013
0.022 (96%)
0.005 (130%)
0.042 (89%)
0.001 (103%)
0.006 (95%)
NT
0.023 (84%)
0.134 (86%)
0.010 (87%)
0.046 (117%)
0.011 (143%)
0.008 (81%)
0.486 (96%)
0.090 (104%)
0.031 (167%)
0.133 (100%)
3-Cl
4-CF3
3-CF3
4-OMe
3-OMe
4-CN
0.005 (115%)
0.002 (93%)
0.028 (96%)
0.05 (95%)
NT, not tested.
(S)-14 also displayed acceptable PK parameters—87%
oral bioavailability, dose normalized oral AUC of
13.4 h lg/mL, iv clearance of 2.2 mL/min/kg, and oral
half-life of 9.7 h.
Figure 2. Biaryl ether 36.
In light of the above PK results, both compounds were
evaluated in a db/db mouse model, using rosiglitazone as
the comparator (Table 3).15 Compound (S)-19 was
shown to effectively lower glucose by 99% at 1 mg/kg
in an 8-day study. Compound (S)-14 was shown to effec-
tively lower glucose by 106% at 10 mg/kg. Rosiglitazone
exhibited a lowering of 78% glucose at 30 mg/kg in this
study.
to be the eutomer, and again predicted to have the S
configuration.14
Both compounds were then selected for rat pharmacoki-
netic (PK) studies. Administration to male Sprague–
Dawley (SD) rats resulted in satisfactory PK parameters
for compound (S)-19—59% oral bioavailability, dose
normalized oral AUC of 8.0 h lg/mL, iv clearance of
2.5 mL/min/kg, and oral half-life of 4.7 h. Compound
Compounds (S)-19 and (S)-14 were also shown to effec-
tively lower triglycerides (Table 3). Unfortunately, both

P. S. Humphries et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6120–6123
6123
script. We also thank Bill Farrell and Christie
Aurigemma for chiral SFC analysis.
References and notes
Scheme 3. Reagents and conditions: (a) 5, NaHMDS, THF, ꢀ50 °C,
2 h, 85%; (b) LiOH, THF, MeOH, H2O, 50 °C, 5 h, 17%.
1. Amos, A. F.; McCarty, D. J.; Zimmet, P. Diabetic Med.
1997, 14, S1.
2. Moller, D. E. Nature 2001, 414, 821.
3. Smith, A.; Fogelfield, L.; Bakris, G. Emerg. Drugs 2000, 5,
441.
4. Willson, T. M.; Brown, P. J.; Sternbach, D. D.; Henke, B.
R. J. Med. Chem. 2000, 43, 527.
5. Murakami, K.; Tobe, K.; Ide, T.; Mochizuki, T.; Ohashi,
M.; Akanuma, Y.; Yazaki, Y.; Kadowaki, T. Diabetes
1998, 47, 1841.
Scheme 4. Reagents and conditions: (a) H2SO4, 1,4-dioxane, H2O,
90 °C, 16 h, 96%.
6. Nomura, M.; Kinoshita, S.; Satoh, H.; Maeda, T.;
Murakami, K.; Tsunoda, M.; Miyacvhi, H.; Awano, K.
Bioorg. Med. Chem. Lett. 1999, 9, 553.
Table 3. Effect of rosiglitazone (Rosi), (S)-20, and (S)-15 on plasma
glucose, triglycerides (TG), hematocrit (Hct), and body weight (BW)
7. For recent publications on dual PPARa/c agonists, see: (a)
Liu, K.; Xu, L.; Berger, J. P.; MacNaul, K. L.; Zhou, G.;
Doebber, T. W.; Forrest, M. J.; Moller, D. E.; Jones, A. B.
J. Med. Chem. 2005, 48, 2262; (b) 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.
8. Hulin, B. WO 91/19702 A1, 1991; Chem. Abstr. 1991, 117,
26552.
9. Bailey, S.; Humphries, P. S.; Skalitzky, D. J.; Su, W.-G.;
Zehnder, L. R. WO 2004/092145, 2004; Chem. Abstr. 2004,
141, 379637.
Compound Dose
Glucose TG
Hct
change change change
BW
(mg/kg) change
(%)a
(%)a
(%)a
(%)a
Rosi
Rosi
3
30
ꢀ46
ꢀ78
ꢀ37
ꢀ99
ꢀ88
ꢀ106
ꢀ67
ꢀ138
ꢀ67
ꢀ5.3
ꢀ8.5
ꢀ7.2
ꢀ8.7
ꢀ3.8
ꢀ7.2
+13.1
+17.3
+11.7
+15.7
+13.2
+16.8
(S)-19
(S)-19
(S)-14
(S)-14
0.03
1
1
ꢀ147
ꢀ177
ꢀ205
10
Male db/db mice (7 weeks old) and lean mice were dosed daily for 8
days by oral gavage with vehicle or the indicated doses of test com-
pound. Plasma glucose, triglycerides, hematocrit, and body weight
were measured before dosing on day ꢀ1 and 2 h post-dose on day 8.
a p < 0.05 versus vehicle control.
10. Humphries, P. S.; Bailey, S.; Do, Q.-Q. T.; Kellum, J. H.;
McClellan, G. A.; Wilhite, D. M. Tetrahedron Lett. 2006,
47, 5333.
11. Sun, S.; Almaden, J.; Carlson, T. J.; Barker, J.; Gehring,
M. R. Metab. Eng. 2005, 7, 38.
compounds also caused an increase in body weight and
a reduction in hematocrit (suggesting the presence of
hemodilution) to a similar extent as rosiglitazone.
12. Berger, J.; Leibowitz, M. D.; Doebber, T. W.; Elbrecht,
A.; Zhang, B.; Zhou, G.; Biswas, C.; Cullinan, C. A.;
Hayes, N. S.; Li, Y.; Tanem, M.; Venture, J.; Wu, S. M.;
Berger, G. D.; Mosley, R.; Marquis, R.; Santini, C.;
Sahoo, S. P.; Tolman, R. L.; Smith, R. G.; Moller, D. E. J.
Biol. Chem. 1999, 274, 6718.
13. (a) Nakamura, Y.; Hirata, M.; Kuwano, E.; Taniguchi, E.
Biosci. Biotechnol. Biochem. 1998, 62, 1550; (b) Miyata,
O.; Shinada, T.; Ninomiya, I.; Naito, T. Tetrahedron 1997,
53, 2421; (c) Miyata, O.; Shinada, T.; Ninomiya, I.; Naito,
T. Tetrahedron Lett. 1991, 32, 3519.
In summary, we have identified a pyridine-3-propa-
noic acid class of potent dual PPARa/c agonists. Sys-
tematic SAR studies generated a multitude of potent
compounds with varied isoform selectivity. Com-
pounds (S)-14 and (S)-19 displayed oral efficacy with
greater apparent potency, at a given dose, than the
benchmark rosiglitazone in the db/db mouse model
of diabetes.
14. Andersson, K. WO 99/62872, 1999; Chem. Abstr. 1999,
132, 22757.
Acknowledgments
15. All in vivo procedures were approved by the Pfizer La
Jolla Institutional Animal Care and Use Committee, and
principles of laboratory animal care (NIH publication no.
85-23, revised 1985) were followed.
The authors gratefully acknowledge all our colleagues in
the PPAR program for their technical support in the
evaluation of the compounds presented in this manu-