The structure-activity relationship between peroxisome proliferator-activated receptor gamma agonism and the antihyperglycemic activity of thiazolidinediones.

Article abstract of DOI:10.1021/jm950395a

The structure-activity relationship between peroxisome proliferator-activated receptor gamma agonism and the antihyperglycemic activity of thiazolidinediones.

Full text of DOI:10.1021/jm950395a

J . Med. Chem. 1996, 39, 665-668
665
6 as first high-affinity ligand for PPARγ, a receptor
subtype selectively expressed in adipocytes and shown
to induce adipocyte differentiation.14 This report details
the structure-activity relationship between compounds
that activate PPARγ in vitro and their reported anti-
hyperglycemic activity in genetically obese and diabetic
mice in vivo.
Th e Str u ctu r e-Activity Rela tion sh ip
betw een P er oxisom e
P r olifer a tor -Activa ted Recep tor γ
Agon ism a n d th e An tih yp er glycem ic
Activity of Th ia zolid in ed ion es
Timothy M. Willson,*,† J effery E. Cobb,†
David J . Cowan,† Robert W. Wiethe,†
Itzela D. Correa,‡ Shimoga R. Prakash,‡
Kelli D. Beck,§ Linda B. Moore,§
Resu lts. A series of compounds (1b-11) was chosen
from the literature that had been previously tested in
rodent models of diabetes.4,7,9-11 Clofibric acid (1b), the
active form of clofibrate (1a ),15 was purchased from
Sigma. Compounds 2-11 were synthesized by the
literature methods.7,9-11 The compounds were tested
in a transient transfection assay in CV-1 cells for their
ability to activate PPARs. To allow comparison of the
relative transcriptional activity of the three receptors
on the same target gene and to prevent endogenous
receptors from complicating the interpretation of re-
sults, an established receptor chimera system was
used.14 The ligand binding domains of mouse PPARR,
PPARγ, and PPARδ were each fused to the DNA-
binding domain of the yeast transcription factor GAL4.
CV-1 cells were transiently transfected with expression
vectors for the respective PPAR chimera together with
a reporter construct containing five copies of the GAL4
upstream activator element driving expression of se-
creted placental alkaline phosphatase (SPAP). Weak
activity on PPARR was observed for only 1b and 2, and
no activity was seen on PPARδ for any of the compounds
(Table 1). By contrast, it was found that all the
compounds (1b-11) were efficacious activators of PPARγ,
showing 25-30-fold induction of reporter gene activity
at the highest dose tested (data not shown). The
potency of compounds 1b-11 as judged by the EC50 for
PPARγ activity, was structure dependent and spanned
a wide dose range from millimolar (for 1b) to nanomolar
(for 10).
In addition to screening for functional activity in cells,
compounds 1b-11 were tested for their ability to bind
directly to PPARγ in vitro. [3H]-6 was prepared by
tritiodebromination of 12 (Scheme 1). [3H]-6 was iso-
lated with a specific activity of 41 Ci/mmol. The ligand
binding domain of PPARγ was expressed in Escherichia
coli as a fusion protein with glutathione S-transferase.14
In competition binding assays, compounds 1b-11 showed
specific displacement of [3H]-6 from the PPARγ ligand
binding domain fusion protein. Thus, compounds 1b-
11 are PPARγ agonists since they bind directly to the
receptor and activate transcription in a comparable dose
range.
Steven A. Kliewer,§ and J u¨rgen M. Lehmann§
Glaxo Wellcome Research and Development, Five Moore
Drive, Research Triangle Park, North Carolina 27709
Received November 29, 1995
In tr od u ction . Resistance of peripheral tissues to the
action of insulin is a characteristic feature of human
obesity and non-insulin-dependent diabetes (NIDDM).1
A number of prospective studies have shown that the
development of insulin resistance is an early event in
the natural progression of the disease.2 Thus, drugs
that reverse the onset of insulin resistance fulfill a major
unmet medical need for the treatment of NIDDM.3
Thiazolidinediones are a class of oral insulin-sensitizing
agents that improve glucose utilization without stimu-
lating insulin release. They significantly reduce glu-
cose, lipid, and insulin levels in rodent models of
NIDDM and obesity,4 and recent clinical data supports
their efficacy in obese diabetic patients.5 The original
lead for the identification of this class of compounds was
clofibrate (1a , Figure 1), which possesses both antilipi-
demic and weak antihyperglycemic activity in humans.6
Scientists at Takeda, through screening analogs of 1a
for increased antihyperglycemic activity in diabetic
mice, identified the acid 2 and subsequently the thia-
zolidinedione 3 (Ciglitazone).7 Although the molecular
mechanism of action remained unknown, analog syn-
thesis and in vivo screening over a period of 15 years
led to the identification of thiazolidinediones with
increased potency.8 Examples of thiazolidinediones that
have progressed to clinical evaluation are the Pfizer
compound Englitazone (4),11 the Takeda compound
Pioglitazone (5),9 and the Beecham compound BRL
49653 (6).10
Peroxisome proliferator-activated receptors (PPARs)
are members of the nuclear receptor superfamily of
ligand-activated transcription factors. Three subtypes
of PPARs have been cloned from mouse and human:
PPARR, PPARγ, and PPARδ (also known as NUCI).12
The PPARs are believed to play a physiological role in
the regulation of lipid metabolism. They can be acti-
vated by high concentrations of fatty acids (FA) and
have been shown to regulate the expression levels of
FA binding proteins or enzymes involved in FA oxida-
tion.13 Since the natural ligand for these receptors has
yet to be identified, they are classified as “orphan”
receptors. We recently identified the thiazolidinedione
Discu ssion . The results in Table 1 show that thia-
zolidinediones 3-11 were selective PPARγ agonists,
whereas acids 1b and 2, although less potent, showed
comparable activity on PPARR and PPARγ. None of
the compounds demonstrated measurable activity on
PPARδ.16 Compounds 3 and 5-10 have been reported
to show antihyperglycemic activity in genetically obese
C57 Bl/6 ob/ob mice,10 an animal model of NIDDM that
is insulin resistant, hyperinsulinemic, and glucose
intolerant. Table 1 shows the reported minimum ef-
fective dose (MED) for antihyperglycemic activity using
an oral glucose tolerance test following dosing of 3
and 5-10 in the diet for 8 d.10,17 Comparison of the
EC50 for activation of PPARγ with the MED for anti-
* Address correspondence to: Timothy M. Willson, Glaxo Wellcome
Research and Development, NTH-M2134, Five Moore Drive, Research
Triangle Park, NC 27709. Tel: (919) 483-9875. FAX: (919) 315-5668.
E-mail: willson∼tm@glaxo.com.
† Department of Medicinal Chemistry.
‡ Department of Drug Metabolism.
§ Department of Cell Biology.
0022-2623/96/1839-0665$12.00/0 © 1996 American Chemical Society

666 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 3
Communications to the Editor
F igu r e 1.
Ta ble 1. In Vitro and in Vivo Activity of Compounds (1b-11)
transactivationa
binding:b
in vivoc
PPARR
EC50 (µM)
PPARγ
EC50 (µM)
PPARδ
EC50 (µM)
PPARγ
(% displacement)
MED
no.
(µmol kg-1
)
1b
2
3
4
5
6
7
8
9
51 ( 2
600 ( 200
110 ( 20
ia
ia
ia
ia
ia
ia
ia
ia
ia
ia
ia
56 ( 2
86 ( 3
81 ( 4
89 ( 1
90 ( 1
89 ( 2
79 ( 6
85 ( 6
80 ( 6
93 ( 1
83 ( 5
80 ( 20
ia
ia
ia
ia
ia
ia
ia
ia
ia
3.0 ( 0.7
3000
13 ( 2
0.69 ( 0.08
0.060 ( 0.004
1.0 ( 0.7
300d
3
300
300
300
3
ia at 1000
0.19 ( 0.03
0.14 ( 0.04
0.013 ( 0.003
10 ( 2
10
11
a
GAL4-PPAR chimeric expression constructs were prepared as described previously (ref 14). Reporter plasmid UAS-tk-SPAP was
generated by insertion of five copies of a GAL4 DNA-binding element into pG12-tk-SPAP (ref 22). CV-1 cells were transfected with the
relevant receptor expression plasmid, reporter construct and â-galactosidase expression plasmid (pCH110, Pharmacia) as described
previously (ref 14). After 16 h, the medium was exchanged to DME medium supplemented with 10% delipidated fetal calf serum and the
test compound at the appropriate concentration. After an additional 24 h, cell extracts were prepared and assayed for alkaline phosphatase
activity and â-galactosidase activity. Alkaline phosphatase activity was corrected for transfection efficiency using the â-galactosidase
activity as internal standard (ref 23). EC50 equals the concentration of compound required to induce 50% of the maximum alkaline
b
phosphatase activity ( standard error, n ) 3. ia ) inactive at 10-3 M (1 and 2) or 10-4 M (3-11). Competition binding was performed
with bacterial extracts containing the glutathione S-transferase-PPARγ ligand binding domain as described previously (ref 14). Percent
displacement equals the percent of [3H]-6 (10-7 M) competitively displaced by the test compound at a concentration of 10-4 M (except 1b
and 2 were at 10-3 M) in triplicate ( standard error. c MED ) minimum effective dose for significant reduction in blood glucose in ob/ob
d
mice (from ref 10a). See comment in ref 17.
Sch em e 1a
a
Reagents: (i) neat, 150 °C, 18 h; (ii) NaH, 4-fluorobenzaldehyde, DMF, 80 °C, 18 h; (iii) 2,4-thiazolidinedione, piperidine, EtOH,
reflux, 5 h; (iv) 3H2(g), 10% Pd-C, DMF, 5 h.
hyperglycemic activity revealed a significant correla-
tion (r2 ) 0.92, P < 0.0005, Figure 2). Thus, for
compounds 3 and 5-10, both the relative and absolute
rank order potency for activation of PPARγ in vitro
mirrored the in vivo antihyperglycemic activity in
diabetic ob/ob mice.

Communications to the Editor
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 3 667
discussion during the course of this work and J ohn
Katzenellenbogen (University of Illinois at Urbanas
Champaign) for critical reading of the manuscript.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails and spectral and analytical data for the preparation of
the radioligand ([3H]-6) (2 pages). Ordering information is
given on any current masthead page.
Refer en ces
(1) Ferrannini, E. The Insulin Resistance Syndrome. Curr. Opin.
Nephrol. Hypertens. 1992, 1, 291-298.
F igu r e 2. Correlation of in vitro PPARγ activity with in vivo
antihyperglycemic activity in ob/ob mice for compounds 3 and
5-10. In vitro PPARγ EC50 from Table 1. In vivo minimum
effective dose (MED) for antihyperglycemic activity from ref
10a; r2 ) 0.92, P < 0.0005.
(2) (a) Warram, J . H.; Martin, B. C.; Krolewski, A. S.; Soeldner, J .
S.; Kahn, C. R. Slow Glucose Removal Rate and Hyperinsuline-
mia Precede the Development of Type II Diabetes in the
Offspring of Diabetic Parents. Ann. Intern. Med. 1990, 113, 909-
915. (b) Lillioja, S.; Mott, D. M.; Spraul, M.; Ferraro, R.; Foley,
J . E.; Ravussin, E.; Knowler, W. C.; Bennett, P. H.; Bogardus
C. Insulin Resistance and Insulin Secretory Dysfunction as
Precursors of Non-Insulin-Dependent Diabetes Mellitus: Pro-
spective Studies of Pima Indians. N. Engl. J . Med. 1993, 329,
1988-1992.
(3) Colca, J . R.; Tanis, S. P. Recent Advances in the Discovery and
Development of Potential Antidiabetic Agents. In Annual Re-
ports in Medicinal Chemistry; Bristol, J . A., Ed.; Academic Press
Inc.: San Diego, 1992; Vol. 27, pp 219-226.
(4) (a) Fujita, T.; Sugiyama, Y.; Taketomi, S.; Sohda, T.; Kawamatsu,
Y.; Iwatsuka, H.; Suzjoka, Z. Reduction of Insulin Resistance in
Obese and/or Diabetic Animals by 5-[4-(1-Methylcyclohexyl-
methoxy)benzyl]-thiazolidine-2,4-dione (ADD-3878, U-63,287,
Ciglitazone), a New Antidiabetic Agent. Diabetes 1983, 32, 804-
810. (b) Chang, A. Y.; Wyse, B. M.; Gilchrist, B. J .; Peterson, T.;
Diani, A. R. Ciglitazone, a New Hypoglycemic Agent. Diabetes
1983, 32, 830-838. (c) Sugiyama, Y.; Taketomi, S.; Shimara, Y.;
Ikeda, H.; Fujita, T. Effects of Pioglitazone on Glucose and Lipid
Metabolism in Wistar Fatty Rats. Arzneim. Forsch./ Drug Res.
1990, 40, 263-267. (d) Stevenson, R. W., Hutson, N. J .; Krapp,
M. N.; Volkmann, R. A.; Holland, G. F.; Eggler, J . F.; Clark, D.
A.; McPherson, R. K.; Hall, K. L.; Danbury, B. H.; Gibbs, E. M.;
Krentler, D. K.; Actions of Novel Antidiabetic Agent Englitazone
in Hyperglycemic Hyperinsulinemic ob/ ob Mice. Diabetes 1990,
39, 1218-1227.
(5) Nolan, J . J .; Ludvik, B.; Beerdsen, P.; J oyce, M.; Olefsky, J .
Improvement in Glucose Tolerance and Insulin Resistance in
Obese Subjects Treated with Troglitazone. N. Engl. J . Med.
1994, 331, 1188-1193.
(6) (a) Ferrari, C.; Testori, G. P.; Betazzoni, A.; Romussi, M.;
Caldara, R.; Frezzati, S. Increased Glucose Disappearance Rate
after Short-term Clofibrate Administration in Normal Subjects
and in Patients with Chemical Diabetes. Horm. Metab. Res.
1978, 10, 4-6. (b) Barnett, D.; Craig, J . G.; Robinson, D. S.;
Perenna Rogers, M. Effect of Clofibrate on Glucose Tolerance
in Maturity Onset Diabetes. Br. J . Clin. Pharmacol. 1977, 4,
455-458.
(7) (a) Kawamatsu, Y.; Saraie, T.; Imamiya, E.; Nishikawa, K.;
Hamuro, Y. Studies on Antihyperlipidemic Agents. Arzneim.-
Forsch./ Drug Res. 1980, 30, 454-459. (b) Sohda, T.; Mizuno,
K.; Imamiya, E.; Sugiyama, Y.; Fujita, T.; Kawamatsu, Y.
Studies on Antidiabetic Agents. II. Synthesis of 5-[4-(1-Methyl-
cyclohexylmethoxy)benzyl]thiazolidine-2,4-dione (ADD-3878) and
its Derivatives. Chem. Pharm. Bull. 1982, 30, 3580-3600.
(8) Hulin, B. New Hypoglycemic Agents. Prog. Med. Chem. 1994,
31, 1-58.
Table 1 shows that compounds 1b, 2, 4, and 11 also
demonstrated PPARγ agonist activity. Compound 11
was reported to be inactive at 1000 µmol kg-1 in ob/ob
mice,10a and Figure 2 predicts that a dose of >10 000
µmol kg-1 would be required to show in vivo activity.
Thus, the in vitro PPARγ activity of 11 is consistent
with the reported in vivo data. Englitazone (4) was
reported to be active in ob/ob mice at doses comparable
to ciglitazone (3);11 however, differences in dosing
regimes (diet vs gavage) and duration of study (8 vs 4
d) did not allow inclusion of this compound in the
correlation.10,11 Finally, the activities of 1b and 2 have
not been reported in ob/ob mice, although 2 was reported
to lower blood glucose in diabetic KK-Ay mice at high
doses (0.2% of diet for 4 d).7a
Within a single compound class that is known to have
acceptable pharmacokinetic properties, it is likely that
differences in in vivo potency reflect changes in the
pharmacodynamic properties of the molecules. Thus,
the correlation between in vivo potency of compounds
3 and 5-10 and the measured in vitro activity impli-
cates PPARγ as the molecular target for the antidiabetic
effects of thiazolidinediones.18 PPARγ is predominantly
expressed in adipose tissue,19 and it was recently shown
that PPARγ, through ectopic expression, can act as a
master regulator of adipocyte differentiation in pre-
adipocyte and multipotential cell lines.14,20 Since skel-
etal muscle is the predominant insulin-sensitive tissue
in the body, accounting for >80% of the insulin-
regulated glucose disposal, an increase in glucose
disposal in adipocytes alone is unlikely to account for
the dramatic antihyperglycemic activity of the thiazoli-
dinediones. Thus, the mechanism by which activation
of PPARγ-regulated genes in adipocytes reverses insulin
resistance remains to be discovered.21
(9) Shoda, T.; Momose, Y.; Meguro, K.; Kawamatsu, Y.; Sugiyama,
Y.; Ikeda, H. Studies on Antidiabetic Agents: Synthesis and
Hypoglycemic Actvity of 5-[4-(Pyridylalkoxy)benzyl]-2,4-thiazoli-
dinediones. Arzneim.-Forsch./ Drug Res. 1990, 40, 37-42.
(10) (a) Cantello, B. C. C.; Cawthorne, M. A.; Cottam, G. P.; Duff, P.
T.; Haigh, D.; Hindley, R. M.; Lister, C. A.; Smith, S. A.; Thurlby,
P. L. [[ω-(Heterocyclylamino)alkoxy]benzyl]-2,4-thiazolidinedi-
ones as Potent Antihyperglycemic Agents, J . Med. Chem. 1994,
37, 3977-3985. (b) Cantello, B. C. C.; Cawthorne, M. A.; Haigh,
D.; Hindley, R. M.; Smith, S. A.; Thurlby, P. L. The Synthesis
of BRL 49653 - a Novel and Potent Antihyperglycemic Agent.
Bioorg. Med. Chem. Lett. 1994, 10, 1181-1184.
Con clu sion . The structure-activity relationship for
PPARγ agonist activity in vitro accurately predicts the
in vivo antihyperglycemic activity of thiazolidinediones
in diabetic mice. The ability to rapidly screen com-
pounds against this target in radioligand competition-
binding and functional transactivation assays should
allow the rapid development of more potent and selec-
tive antidiabetic agents.
(11) Clark, D. A.; Goldstein, S. W.; Volkmann, R. A.; Eggler, J . F.;
Holland, G. F.; Hulin, B.; Stevenson, R. W.; Kreutter, D. K.;
Michael Gibbs, E.; Krupp, M. N.; Merrigan, P.; Kelbaugh, P. L.;
Andrews, E. G.; Tickner, D. L.; Suleske, R. T.; Lamphere, C. H.;
Ack n ow led gm en t. We thank Ken Batchelor, Stew-
art Noble, Steve Blanchard, and Bill Wilkison (Glaxo
Wellcome Research and Development) for support and

668 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 3
Communications to the Editor
Rajeckas, F. J .; Kappeler, W. H.; McDermott, R. E.; Hutson, N.
J .; J ohnson, M. R. Substituted Dihydrobenzopyran and Dihy-
drobenzofuran Thiazolidine-2,4-diones as Hypoglycemic Agents.
J . Med. Chem. 1991, 34, 319-325.
Common Mechanism of Action for Fatty Acids and Thiazoli-
dinedione Antidiabetic Agents on Gene Expression in Preadipose
Cells. Mol. Pharmacol. 1994, 46, 1070-1076.
(17) The MED for compound 5 is reported as 300 µmol kg-1 in ref
10a and 200 µmol kg-1 in ref 10b. The higher value has been
used in Table 1 and in Figure 2. If the lower value is used, the
correlation in Figure 2 shows r2 ) 0.91, P < 0.0005.
(18) A reviewer has noted that a number of structurally diverse
thiazolidinediones of varying potencies exist which were not
included in our analysis. Therefore caution should be exercised
in extending our results to cover these other classes of com-
pounds.
(12) (a) Isseman, I.; Green, S. Activation of a Member of the Steroid
Hormone Receptor Superfamily by Peroxisome Proliferatos.
Nature 1990, 347, 645-650. (b) Schmidt, A.; Endo, N.; Rutledge,
S. J .; Vogel, R.; Shinar, D.; Rodan, G. A. Identification of a New
Member of the Steroid Hormone Receptor Superfamily that is
Activated by a Peroxisome Proliferator and Fatty Acids. Mol.
Endocrinol. 1992, 6, 1634-41. (c) Kliewer, S. A.; Forman, B. M.;
Blumberg, B.; Ong, E. S.; Borgmeyer, U.; Manglesdorf, D. J .;
Umesono, K.; Evans, R. M. Differential Expression and Activa-
tion of a Family of Murine Peroxisome Proliferator-activated
Receptors. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 7355-7359.
(19) Chawla, A.; Schwarz, E. J .; Dimaculangan, D. D.; Lazar, M. A.
Peroxisome Proliferator-activated Receptor (PPAR) γ: Adipose-
predominant Expression and Induction Early in Adipose Dif-
ferentiation. Endocrinology 1994, 135, 798-800.
(20) Tontonoz, P.; Hu, E.; Spiegelman, B. M. Stimulation of Adipo-
genesis in Fibroblasts by PPARγ2, a Lipid-activated Transcrip-
tion Factor. Cell 1994, 79, 1-20.
(21) Flier, J . S. The Adipocyte: Storage Depot or Node on the Energy
Information Superhighway? Cell 1995, 80, 15-18.
(22) Berger, J .; Hauber, R.; Geiger, R.; Cullen, B. R. Secreted Alkaline
Phosphatase: a Powerful New Quantitative Indicator of Gene
Expression in Eukaryotic Cells. Gene 1988, 66, 1-10.
(23) Pfahl, M; Tzuckerman, M; Zhang, X.-K.; Lehmann, J . M.;
Hermann, T.; Wills, K. N.; Graupner, G. Nuclear Retinoic Acid
Receptors: Cloning, Analysis, and Function. Methods Enzymol.
1990, 26, 256-270.
(13) Desvergne, B.; Wahli, W. PPAR:
a Key Nuclear Factor in
Nutrient/Gene Interactions? In Inducible Gene Expression; Bae-
uerle, P., Ed.; Birkha¨use: Boston, 1995; pp 142-176.
(14) Lehmann, J . M.; Moore, L. B.; Smith-Oliver, T. A.; Wilkison, W.
O.; Willson, T. M.; Kliewer, S. A. An Antidiabetic Thiazolidinedi-
one is
a High Affinity Ligand for Peroxisome Proliferator-
Activated Receptor γ (PPARγ). J . Biol. Chem. 1995, 270, 12953-
12956.
(15) Thorpe, J . M. Experimental Evaluation of an Orally Active
Combination of Andosterone and Ethyl Chlorophenoxy-iso-
butryrate. Lancet 1962, 1323-1326.
(16) Compound 6 has also been reported to activate FAAR, a homolog
of mouse PPARδ, in 3T3-C3 fibroblasts with an EC50 ∼10 µM:
Ibrahimi, A.; Teboul, L; Gaillard, D; Amri, E.-Z.; Ailhaud, G.;
Young, P.; Cawthorne, M. A.; Grimaldi, P. A. Evidence for a
J M950395A