Phenylacetic acid derivatives as hPPAR agonists

Article abstract of DOI:10.1016/S0960-894X(03)00115-X

Beginning with the weakly active lead structure 1, a new series of hPPAR agonists was developed. In vivo glucose and triglyceride lowering activity was obtained by homologation and oxamination to 3, then conversion to substituted benzisoxazoles 4 and 5. F

Full text of DOI:10.1016/S0960-894X(03)00115-X

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1277–1280
Phenylacetic Acid Derivatives as hPPAR Agonists
Conrad Santini,a,* Gregory D. Berger,a,y Wei Han,a,{ Ralph Mosley,a Karen MacNaul,b
Joel Berger,b Thomas Doebber,b Margaret Wu,b David E. Moller,b Richard L. Tolmana,x
and Soumya P. Sahooa,*
aDepartment of Basic Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065 USA
bDepartment of Molecular Endocrinology, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065 USA
Received 21 July 2002; accepted 21 January 2003
Abstract—Beginning with the weakly active lead structure 1, a new series of hPPAR agonists was developed. In vivo glucose and
triglyceride lowering activity was obtained by homologation and oxamination to 3, then conversion to substituted benzisoxazoles 4
and 5. Further manipulation afforded benzofurans 6 and 7. Compound 7 was of comparable potency as a glucose and triglyceride
lowering agent in insulin resistant rodents to BRL 49653.
# 2003 Elsevier Science Ltd. All rights reserved.
The thiazolidinedione (TZD) class of insulin sensitizing
agents for the treatment of Type II diabetes is well
documented.1,2 A mechanism has been advanced for
their glucose lowering action invoking PPARg nuclear
receptor agonist activity, primarily in adipose tissue.3,4
This is supported by the correlation of in vitro and in
vivo structure–activity relationships.5,6 Interested in
discovering novel non-TZD PPARg agonists, we
screened in-house samples for hPPARg binding activity
using 2D similarities to various TZD’s as a searching
algorithm. Acetyl phenol 1 (Fig. 1) displayed modest in
vitro hPPARg agonist activity. A directed synthesis
program was initiated in a bid to both improve the
observed hPPARg activity and increase the probability
of eliciting in vivo antidiabetic efficacy.
reduction. Combination of 11 and 13 under basic con-
ditions afforded acyl phenols 14. Saponification of 14
(R=Et) gave 2, whereas oxamination and saponifica-
tion gave 3. Oxamination followed by acylation, cycli-
zation and hydrolysis gave benzisoxazoles 4 and 5. The
preparation of benzofurans is depicted in Scheme 2.
3-Methoxyphenol was alkylated with the appropriate
a-bromoketones, giving ketones 15. Acid catalyzed
cyclization cleanly gave benzofurans 16, which under-
went demethylation and propylation as above to give
17. Alkylation with 11 and hydrolysis afforded 6 and 7.
Recombinant hPPARg receptor affinity was determined
according to our new protocol.7 Agonist activity was
established using a cell-based transactivation assay.8
Similar assays using Gal-4-PPARd and Gal-4-PPARa
chimers characterized activity on related PPAR isoforms.
In vivo glucose and triglyceride lowering activities were
obtained using either male db/db mice or Zucker dia-
betic fatty (ZDF) rats as previously described.6,8 The in
vitro activity of compounds 1 through 7 along with that of
rosiglitazone (8) is shown in Table 1. Table 2 summarizes
the in vivo data for compounds 3–7 with rosiglitazone.
The preparation of propionyl phenol 2, oxime 3 and
benzisoxazoles 4 and 5 is shown in Scheme 1. The thio
halide 11 was prepared by thiocarbamoylation of 9,
thermal rearrangement to 10 and alkylation of the
derived sodium thiolate. Resorcinols 13 were prepared
by O-allylation of 12, Claisen rearrangement and
Unlike rosiglitazone and TZD’s in general, acids 1
through 7 were non-selective full hPPAR agonists,
selectivity being defined here as a difference in affinity or
activity equal to three orders of magnitude. By contrast,
none of these compounds were agonists of mPPARa. In
vitro activity against hPPARd was comparatively con-
*Corresponding author: E-mail: conrad_santini@merck.com
yCurrent address: Pfizer Central Research, Eastern Point Road, Gro-
ton, CT 06340, USA.
xCurrent address: Geron Corporation, 230 Constitution Drive, Menlo
Park, CA 94025, USA.
{Current address: Merck Research Laboratories, PO Box 4, Sumney-
town Pike, West Point, PA 19486, USA.
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00115-X

1278
C. Santini et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1277–1280
Figure 1.
Scheme 1. (a) 1.4 equiv (CH3)2NC(S)Cl, pyridine, reflux, 4 h, 80–85%; (b) sulfolane, reflux, 15 min, 45–60%; (c) (i) 1.2 equiv NaOCH3, CH3OH,
reflux, 2 h; (ii) 4 equiv Br(CH2)3Br, CH3OH, 0–20 ꢁC, inverse addition, 70–75%; (d) 1.1 equiv each allyl bromide, Cs2CO3, DMF, 20 ꢁC, 95–98%; (e)
ortho-dichlorobenzene, reflux, 20–24 h, 75–80%; (f) H2, 1 atm, 5% Pd/C, MTBE, 100%;ꢁ(g) 1.05 equiv 11, 1.1 equiv Cs2CO3, DMF, 65–70%; (h) 5
.
equiv each NH2OH HCl, NaOAc, CH3OH, reflux, 4 h, 95–99%; (j) Ac2O (as solvent), 20 C, 30 min; (k) pyridine, reflux, 3 h, 85–90% (two steps), (l)
1.3 equiv aq KOH, iPrOH, 50 ꢁC, 100%.
Scheme 2. (a) 0.9 equiv crude RC(O)CH2Br, 1.1 equiv Cs2CO3, DMF, 70–75%; (b) 15 vol% CH3SO2OH/CH2Cl2, ꢂ10–20 ꢁC, 65–70%; (c) 1.1 equiv
BBr3, CH2Cl2, ꢂ10 to 20 ꢁC, 1 h, 85–90%; (d) 1.1 equiv each allyl bromide, Cs2CO3, DMF, 95–98%; (e) ortho-dichlorobenzene, reflux, 75–80%; (f)
H2, 1 atm, 5% Pd/C, MTBE, 30 min, 90–95%; (g) 1.05 equiv 11, 1.1 equiv Cs2CO3, DMF, 65–70%; (h) 1.3 equiv aq KOH, iPrOH, 50 ꢁC, 100%.
stant across most of the compounds, but hPPARa and g
activity were more dependent on structural changes.
The weak in vitro activity of 1 was markedly improved
by simple homologation to 2, most notably in hPPARa
binding (12ꢀ) and, to a lesser extent (3–8ꢀ) in the
functional activation assays. Further potentiation was
obtained by converting 2 to oxime 3, which was much
more potent than 1 as an hPPARa and g agonist (12–
75ꢀ). Oxime 3 was also a more potent hPPARg agonist
than rosiglitazone, albeit clearly less active in vivo (ZDF
rat). The derived benzisoxazole 4 was similar to 3 as an
hPPARg agonist in vitro, but appreciably less potent
against hPPARa. Benzisoxazole 4 was also noticeably
less active in vivo (db/db mice) than rosiglitazone.
Pharmacokinetic profiling of 3 and 4 indicated low sys-
temic exposure following oral dosing. Benzisoxazole 4,

C. Santini et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1277–1280
Table 1. In vitro data for compounds 1–8
1279
Compd
hPPAR Binding IC50 (nM)
hPPAR Activation EC50 (nM)
ꢀ
ꢁ
ꢂ
ꢀ
ꢁ
ꢂ
1
2
3
4
5
6
7
8a
7400
600
100
570
20
20
10
4
10
40
6
580
300
48
110
140
560
94
1600
260
74
80
10
9
11
5
12
6
NAb
230
70
15
19
23
102
9
247
125
1300
18
310
1730
54
>5ꢀ105
>5ꢀ105
210
NAb
10
aBinding IC50’s are ꢃ25%. All activation EC50’s were reflective of full agonist activity (>80% activation).
bNA indicates <20% activation at all concentrations tested (max concentration=3 mM).
Table 2. In vivo data for compounds 3–7
Compd
Zucker diabetic fatty rat
TG
db/db mousea
Pharmacokinetic profilesb
iv (1.5 mpk) po (10 mpk)
Dose
(mpk/day)
Glu
(% Corr)c
FFA
(% Dec)d
Glu
(% Corr)c
TG
(% Dec)d
(% Dec)d
Clp
t1/2
nAUC
F (%)
3
3
4
4
5
5
5
6
7
7
8e
8e
100
30
100
30
30
10
3
30
30
10
30
10
89
44
66
29
95
83
18
80
65
70
42
ꢂ65
ꢂ53
ꢂ34
54
15
5.7
2.8
1.0
0.73
0.62
33
50
52
ꢂ14
33
67
17.6
74
52
96
71
30
74
94
77
81
46
46
60
86
72
45
2.6
4.7
2.52
100
56
27
109
99
66
13
94
78
88
67
86
54
3.3
2.6
2.1
3.8
3.14
4.08
84
64
65ꢃ5
75ꢃ5
aMale db/db mice (12–13 weeks of age, n=7) and non-diabetic mice (lean control, n=7) were provided ad libitum access to rodent chow and water
and received once-a-day oral dosing of the sodium-salts of tested compounds by gavage with vehicle (0.25% methylcellulose) for 11 days. Blood was
collected from the tail for measurement of plasma levels of glucose and triglyceride. For further experimental details and a complete description of
the related Zucker Diabetic Fatty Rat protocol, see refs 6 and 8.
bFasted male Sprague–Dawley rats (n=3), which have been surgically cannulated in the femoral artery and vein, received an intravenous dose by
bolus injection into the femoral vein, or an oral gavage dose. Blood samples were taken serially at selected time points from the femoral arterial
.
cannula. Clp=clearance (mL/min/kg); t1/2=half life (h); nAUC=dose-normalized AUC (mg h/mL); F, bioavailability. Variability of PK parameters
as follows: Clpꢃ20%, t1/2ꢃ25%, nAUCꢃ15%.
cDefined as the percentage of the difference between diabetic controls and lean controls (i.e., 100% correction=lean control).
dDefined as the percentage decrease in the absolute level of the indicated parameter (FFA, free fatty acids).
eUnless otherwise indicated, variability is ꢃ20%. Missing entries under the ZDF rat and db/db mouse headings indicate experiments not carried out.
in particular, was the most rapidly cleared compound
in this group. Replacement of the benzisoxazole ethyl
substituent with a trifluoromethyl group gave a com-
pound (5) that, although not much better than 4 in
vitro, was superior to it and comparable to rosiglita-
zone as a glucose lowering agent in vivo (ZDF rat and
db/db mouse). This was ascribed to the enhanced PK
profile relative to 4. Curiously, although 5 also effec-
tively lowered triglycerides in mice (although less so
than rosiglitazone), it raised them in rats. The origin of
this effect is under active investigation and its cause will
be disclosed in a separate report. Replacement of the
benzisoxazole nitrogen of 4 with methine gave benzo-
furan 6. Although appreciably less potent as an
hPPARa/g agonist than 4,6 was more effective in
reduction of serum glucose and triglycerides in vivo
(db/db mice).8 Focusing on the apparently more effi-
cacious benzofurans, a phenyl substituent was intro-
duced in place of ethyl, affording 7, the most potent
in vitro agonist in this series. Benzofuran 7 displayed
efficacy in db/db mice at least equal to rosiglitazone,
although it was less robust in the ZDF rat. Its pharma-
cokinetic profile was also among the more favorable in
this group and compared well with rosiglitazone (8).
Incorporation of a phenyl substituent at the benzisox-
azole 3-position led to a series that is being reported
elsewhere.9
In summary, beginning with 1, structural modifications
afforded a number of potent but non-selective hPPAR
agonists. The most notable of these, benzofuran 7, dis-
played serum glucose and triglyceride lowering efficacy
in db/db mice comparable to the clinical agent rosiglita-
zone. This compound can be used as a tool to further
understand the pharmacology of pan-hPPAR agonists
in vivo.

1280
C. Santini et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1277–1280
9455. (b) Jiang, G.; Dallas-Yang, Q.; Li, Z.; Szalkowski, D.;
Acknowledgements
Liu, F.; Shen, X.; Wu, M.; Zhou, G.; Doebber, T. W.; Berger,
J.; Moller, D. E.; Zhang, B. B. Diabetes 2002, 51, 2412.
7. Human PPARg2, PPARa and PPARd receptors were
expressed as GST-fusion proteins in Escherichia coli. The full
length human cDNAs for PPARd (provided by Dr. Azriel
Schmidt, MRL) and PPARa (provided by Dr. Tom Rushmore,
MRL) were subcloned into pGEX-KT expression vectors
(Pharmacia), whereas pGEX-hPPARg2 was constructed as
described by Elbrecht (JBC 1999, 274, 7913). Expression and
purification of GST-PPAR proteins, and establishment of a
Scintillation Proximity Assay (SPA)-based receptor binding
assays was similar that described by Elbrecht. For both
PPARa and PPARg 5 nM of [3H2]L-797773 (specific activity
of 34.3 Ci/mmol) was used, and for PPARd 2.5 nM of [3H2]L-
783483 (specific activity of 13.4 Ci/mmol) was used. Results
are expressed as percent inhibition and inflection points cal-
culated by a four-parameter logistic equation. Kis are calcu-
lated by the equation of Cheng and Prusofff. Cheng, Y.:
Prusoff, W. H. Biochem. Pharmacol. 1973, 22, 3099.
8. Berger, J.; Leibowitz, M. D.; Doebber, T. W.; Elbrecht, A.;
Zhang, B.; Zhou, G.; Biswas, C.; Cullinan, C. A.; Hayes, N. S.;
Li, L.; Tanen, M.; Ventre, J.; Wu, M. S.; Berger, G. D.; Mos-
ley, R.; Marquis, R.; Santini, C.; Sahoo, S.; Smith, R. G.;
Tolman, R.; Moller, D. E. J. Bio. Chem. 1999, 274, 6718.
9. Adams, A. D.; Yuen, W.; Hu, Z.; Santini, C.; Jones, A. B.;
MacNaul, K.; Berger, J.; Doebber, T.; Moller, D. E. Bioorg.
Med. Chem. Lett. 2003, 13, 931.
We acknowledge the technical support provided by
Cathy Cullinan, Neelam Sharma, Chhabi Biswas,
Nancy Hayes, Deborah Szalkowski, John Ventre, Sylvia
White-Carrington, Roger Meurer, and Drs. Victor
Ding, James V. Heck, Azriel Schmidt, Alex Elbrecht,
Gaochao Zhou, Yuli Chen and Ying Li.
References and Notes
1. Sohda, T.; Mizuno, K.; Tawada, H.; Sugiyama, Y.; Fujita,
T.; Kawamatzu, Y. Chem. Pharm. Bull. (Tokyo) 1982, 30, 3563.
2. Hulin, B.; McCarthy, P. A.; Gibbs, E. M. Curr. Pharm.
Design 1996, 2, 85.
3. Tontonoz, P.; Hu, E.; Spiegelman, B. M. Curr. Biol. 1995,
5, 571.
4. Schoonjans, K.; Staels, B.; Auwerx, J. Biochimica et Bio-
physica Acta 1996, 1302, 93.
5. Willson, T. M.; Cobb, J. E.; Cowan, D. J.; Wiethe, R. W.;
Correa, I. D.; Prakash, S. R.; Beck, K. D.; Moore, L. B.;
Kliewer, S. A.; Lehmann, J. M. J. Med. Chem. 1996, 39, 665.
6. (a) Zhang, B.; Graziano, M. P.; Doebber, T. W.; Liebowitz,
M. D.; White-Carrington, S.; Szalkowski, P. J. H.; Wu, M.;
Cullinan, C. A.; Bailey, P.; Lollman, B.; Frederich, R.; Flier,
J. S.; Strader, C. D.; Smith, R. G. J. Biol. Chem. 1996, 271,