N-(2-benzoylphenyl)-L-tyrosine PPARγ agonists. 1. Discovery of a novel series of potent antihyperglycemic and antihyperlipidemic agents

Article abstract of DOI:10.1021/jm9804127

We have identified a novel series of antidiabetic N-(2-benzoylphenyl)- L-tyrosine derivatives which are potent, selective PPARγ agonists. Through the use of in vitro PPARγ binding and functional assays (2S)-3-(4- (benzyloxy)phenyl)-2-((1-methyl-3-oxo-3-ph

Full text of DOI:10.1021/jm9804127

5020
J . Med. Chem. 1998, 41, 5020-5036
N-(2-Ben zoylp h en yl)-L-tyr osin e P P ARγ Agon ists. 1. Discover y of a Novel Ser ies
of P oten t An tih yp er glycem ic a n d An tih yp er lip id em ic Agen ts
Brad R. Henke,*,† Steven G. Blanchard,| Marcus F. Brackeen,†,∇ Kathleen K. Brown,# J eff E. Cobb,†
J on L. Collins,† W. Wallace Harrington, J r.,# Mir A. Hashim, Emily A. Hull-Ryde,# Istvan Kaldor,†
Steven A. Kliewer,‡ Debra H. Lake,† Lisa M. Leesnitzer,| J u¨rgen M. Lehmann,‡ J ames M. Lenhard,#
Lisa A. Orband-Miller,| J ohn F. Miller,† Robert A. Mook, J r.,† Stewart A. Noble,†,[ William Oliver, J r.,#
Derek J . Parks,| Kelli D. Plunket,‡ J erzy R. Szewczyk,† and Timothy M. Willson†
Glaxo Wellcome Research and Development, Five Moore Drive, Research Triangle Park, North Carolina 27709
Received J uly 13, 1998
We have identified a novel series of antidiabetic N-(2-benzoylphenyl)-L-tyrosine derivatives
which are potent, selective PPARγ agonists. Through the use of in vitro PPARγ binding and
functional assays (2S)-3-(4-(benzyloxy)phenyl)-2-((1-methyl-3-oxo-3-phenylpropenyl)amino)-
propionic acid (2) was identified as a structurally novel PPARγ agonist. Structure-activity
relationships identified the 2-aminobenzophenone moiety as a suitable isostere for the
chemically labile enaminone moiety in compound 2, affording 2-((2-benzoylphenyl)amino)-3-
(4-(benzyloxy)phenyl)propionic acid (9). Replacement of the benzyl group in 9 with substituents
known to confer in vivo potency in the thiazolidinedione (TZD) class of antidiabetic agents
provided a dramatic increase in the in vitro functional potency and affinity at PPARγ, affording
a series of potent and selective PPARγ agonists exemplified by (2S)-((2-benzoylphenyl)amino)-
3-{4-[2-(methylpyridin-2-ylamino)ethoxy]phenyl}propionic acid (18), 3-{4-[2-(benzoxazol-2-
ylmethylamino)ethoxy]phenyl}-(2S)-((2-benzoylphenyl)amino)propanoic acid (19), and (2S)-((2-
benzoylphenyl)amino)-3-{4-[2-(5-methyl-2-phenyloxazol-4-yl)ethoxy]phenyl}propanoic acid (20).
Compounds 18 and 20 show potent antihyperglycemic and antihyperlipidemic activity when
given orally in two rodent models of type 2 diabetes. In addition, these analogues are readily
prepared in chiral nonracemic fashion from L-tyrosine and do not show a propensity to undergo
racemization in vitro. The increased potency of these PPARγ agonists relative to troglitazone
may translate into superior clinical efficacy for the treatment of type 2 diabetes.
Peroxisome proliferator-activated receptors (PPARs)
are orphan receptors belonging to the steroid/thyroid/
retinoid receptor superfamily of ligand-activated tran-
scription factors. Three mammalian PPARs have been
identified thus far and are termed PPARR, PPARγ, and
PPARδ.1-5 PPARs regulate expression of target genes
by binding to DNA sequences termed PPAR response
elements (PPREs). PPREs have been identified in the
regulatory regions of a number of genes encoding
proteins that are involved in lipid metabolism and
energy balance, such as aP2,6 phosphoenolpyruvate
carboxykinase (PEPCK),7 acyl-CoA synthetase,8 and
lipoprotein lipase (LPL).9 The physiological role of the
PPAR family in the regulation of lipid metabolism and
storage has been the subject of several recent re-
views.10-15
adipocyte differentiation in vitro, suggesting that the
PPARγ is an important component in the adipogenic
signaling cascade and in lipid storage and utiliza-
tion.6,16,17 The human PPARγ gene structure has been
characterized, and two isoforms with a common ligand
binding domain, PPARγ1 and PPARγ2, have been
identified. PPARγ1 is the predominantly expressed
isoform both in man18 and in rodents.19 A number of
naturally occurring fatty acids, eicosanoids, prostaglan-
dins, and their metabolites have been shown to modu-
late PPARγ activity,20-24 consistent with the hypothesis
that fatty acids or their metabolites may be naturally
occurring PPARγ ligands.
We recently identified a class of antidiabetic drugs,
the thiazolidinediones (TZDs), as high-affinity PPARγ
agonists,25 which was subsequently corroborated by
others.22,26,27 In addition, we have demonstrated a
correlation between the in vitro potency at the PPARγ
receptor and the in vivo antihyperglycemic potency in
C57 Bl/6 ob/ob mice for a series of TZDs.28 These results
suggest that PPARγ is the molecular target responsible
for the antidiabetic activity of the thiazolidinedione class
of agents. TZDs, or “glitazones”, enhance the sensitivity
of target tissues to insulin and also reduce lipid and
insulin levels in animal models of type 2 diabetes and
clinically in type 2 diabetics.29-35 TZDs do not cause
an increase in insulin secretion or in the number or
affinity of insulin receptor binding sites, suggesting that
The PPARγ receptor subtype is predominantly ex-
pressed in adipose tissue and plays a pivotal role in
* Address correspondence to: Brad R. Henke, Glaxo Wellcome
Research and Development, NORTH M2111, Five Moore Drive,
Research Triangle Park, NC 27709. Tel: (919) 483-1363. Fax: (919)
315-5668. E-mail: brh14990@glaxowellcome.com.
† Department of Medicinal Chemistry.
| Department of Molecular Biochemistry.
‡ Department of Molecular Endocrinology.
Department of Molecular Pharmacology.
#
Department of Metabolic Diseases.
∇ Current address: MonomerChem Inc., Research Triangle Park,
NC 27709.
[ Current address: Synaptic Pharmaceutical Corp., Paramus, NJ
07652.
10.1021/jm9804127 CCC: $15.00 © 1998 American Chemical Society
Published on Web 11/11/1998

N-(2-Benzoylphenyl)-L-tyrosine PPARγ Agonists. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25 5021
Ch a r t 1
TZDs amplify postreceptor events in the insulin signal-
ing cascade.36
known thiazolidinediones and thus offer an advantage
as an antidiabetic agent. This manuscript provides an
account of the discovery of a novel set of tyrosine-based
PPARγ agonists and describes the synthesis, initial
structure-activity relationship (SAR) studies, and in
vitro activity of this series of compounds. In addition,
the in vivo antihyperglycemic activity of a subset of
these compounds is described. The detailed SAR of
these N-(2-benzoylphenyl)-L-tyrosine PPARγ agonists is
the subject of the following two papers.
A number of TZDs have been advanced into the clinic
for the treatment of type 2 diabetes without benefit of
the knowledge of their molecular target, using in vivo
assays to discriminate among the candidate molecules.
The TZDs currently undergoing clinical evaluation
include troglitazone,37 pioglitazone,38-40 and rosiglita-
zone (BRL 49653)41,42 (Chart 1). The TZD class of
compounds is not without drawbacks, however. Al-
though troglitazone has been approved for use in the
United States and J apan, it displays modest in vivo
potency both in animal models of type 2 diabetes43,44
and clinically,29 suggesting that improvements in an-
tidiabetic potency are feasible. A number of TZDs have
been dropped from development due to their unaccept-
able side-effect profile.45 It remains unclear whether
the side effects are caused by the mechanism of action
of these compounds or originate within the 2,4-thiazo-
lidinedione chemical structure common to this class.
Notably, all TZDs have been developed as racemates
since the stereogenic center present in the 2,4-thiazo-
lidinedione moiety is known to undergo facile racem-
ization under physiological conditions.46
Ch em istr y
Synthesis of the vinylogous amide analogues 2, 5, 7,
and 8 (Table 1) was carried out by condensation of
O-benzyltyrosine or its corresponding methyl ester with
the requisite â-diketone in the presence of either dicy-
clohexylamine or molecular sieves as a catalyst (Scheme
1).47,48 The remainder of O-benzyltyrosine analogues
from Table 1 were prepared via a rhodium-catalyzed
insertion reaction49 of a suitable aniline derivative with
3-(4-(benzyloxy)phenyl)-2-diazopropionic acid methyl
ester,50 followed by saponification to the desired acid
(Scheme 2).
Analogues found in Table 2 were prepared via the
general route depicted in Scheme 3. L-Tyrosine methyl
ester and 2-benzoylcyclohexanone51 were refluxed in
anisole in the presence of 10% Pd/C with removal of
We were interested in developing a series of potent,
selective, non-thiazolidinedione PPARγ agonists which
might surmount the problems associated with the
Sch em e 1a
a
Reagents: (i) dicyclohexylamine, MeOH, reflux, 24 h or 4-Å molecular sieves, MeOH, reflux, 72 h.
Sch em e 2a
a
Reagents: (i) Rh2(OAc)4, toluene, rt to 80 °C, 1-12 h; (ii) LiOH, THF/MeOH (3:1), rt, 4-24 h.

5022 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25
Henke et al.
Sch em e 3a
a
Reagents: (i) 2-benzoylcyclohexanone, 10% Pd/C, anisole, reflux, 2 h; (ii) Ph3P, DEAD, HET-OH, THF, rt, 6-18 h; (iii) LiOH, THF/
MeOH (3:1), rt, 4-24 h.
Sch em e 4a
a
Reagents: (i) Ph3P, DEAD, 2-(N-Boc-N-methylamino)ethanol, THF, rt, 6 h; (ii) trifluoroacetic acid, DCM, rt, 30 min; (iii) 2-fluoropyridine,
reflux, 16 h; (iv) LiOH, THF/MeOH (3:1), rt, 7 h.
Sch em e 5a
a
Reagents: (i) NaH, 2-(2-benzoxazolylmethylamino)ethanol methylsulfonyl ester, DMF, 100 ˚C, 2 h; (ii) isoamyl nitrite, AcOH, CHCl3,
rt to 60 ˚C, 15 min; (iii) Rh2(OAc)4, Ar-NH2, toluene, rt to 80 ˚C, 1-12 h; (iv) DBU, toluene, reflux, 1 h; (v) LiOH, THF/MeOH (3:1), rt,
4-24 h.
water via a Dean-Stark trap to effect condensation to
the vinylogous amide and subsequent dehydrogenation
to the benzophenone derivative 31 without loss of
stereochemical integrity. The resulting (S)-2-((2-ben-
zoylphenyl)amino)-3-(4-hydroxyphenyl)propionic acid
methyl ester was then reacted with the desired hetero-
cyclic alcohol via a Mitsunobu reaction to afford the
intermediate methyl ester. Saponification with LiOH
in THF afforded compounds 15-17, 19, and 20 in
enantiomerically pure fashion. Synthesis of the requi-
site heterocyclic alcohol coupling partners was carried
out via literature methods.41,52,53
Analogues 29 and 30 (Table 4) were prepared from 18
via standard functional group transformations.
Most of the analogues in Table 4 were prepared as
outlined in Scheme 5. Alkylation of tyrosine methyl
ester with 2-(2-benzoxazolylmethylamino)ethanol meth-
ylsulfonyl ester41 followed by treatment with isoamyl
nitrite provided the key diazo intermediate 33. Rhodium-
catalyzed diazo insertion followed by treatment with
LiOH or NaOH then afforded 24-28. In most cases
when a carbonyl was ortho to the amino moiety of the
insertion partner, the desired benzophenone product
was accompanied by a byproduct determined to be cyclic
derivative 34, which may arise from rearrangement of
an epoxide intermediate. Intermediate 34 could be
converted to the desired benzophenone species simply
by refluxing with DBU in toluene to effect a formal
retroaldol reaction. The cyclic derivative 34 appeared
to be a single diastereomer by 1H NMR, but assignment
of relative stereochemistry was not attempted. At-
tempts at asymmetric insertion reactions using chiral
rhodium catalysts54-56 led to poor (<10% ee) asymmetric
induction under all conditions tried.
We suspected compound 18 could not be prepared via
the Mitsunobu coupling of 31 with 2-(N-methyl-N-
pyridin-2-ylamino)ethanol as described in Scheme 3 due
to competing intramolecular cyclization of the activated
alcohol onto the pyridine nitrogen.41 Therefore 18 was
synthesized by the procedure outlined in Scheme 4 via
Mitsunobu reaction with 2-(N-tert-butoxycarbonyl-N-
methylamino)ethanol to afford 32, followed by removal
of the Boc group, addition to 2-fluoropyridine, and
saponification. Synthesis of ent-18 was carried out in
identical fashion starting with D-tyrosine methyl ester.
Compounds 22 and 23 were prepared by the route

N-(2-Benzoylphenyl)-L-tyrosine PPARγ Agonists. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25 5023
Sch em e 6a
a
Reagents: (i) t-BuO-K+, 18-C-6, Ar-XH, THF, rt, 3 h; (ii) LiOH, THF/MeOH (3:1), rt, 4-7 h.
Sch em e 7a
a
Reagents: (i) NaN(TMS)2, 2-iodobenzyl bromide, THF, -78 ˚C, 90 min; (ii) phenylboronic acid, PdCl2(PPh3)2, CO(g), K2CO3, 1,4-
dioxane, 95 °C, 22 h; (iii) LiOH, THF/MeOH (3:1), rt, 6 h.
outlined in Scheme 6 via alkylation of the known
R-bromo ester 3657 with either 2-mercapto- or 2-hy-
droxybenzophenone58 followed by saponification. Com-
pound 21 was prepared by alkylation of the known ester
3757 with 2-iodobenzyl bromide, followed by palladium-
catalyzed carbonylative cross-coupling59 with phenyl-
boronic acid to form the desired benzophenone deriva-
tive (Scheme 7). Saponification then afforded compound
21.
tor system was utilized to allow comparison of the
relative transcriptional activity of the three receptor
subtypes on the same target gene and to prevent
endogenous receptor activation from complicating the
interpretation of results.25 The ligand binding domains
for murine and human PPARR, PPARγ, and PPARδ
were each fused to the yeast transcription factor GAL4
DNA binding domain. CV-1 cells were transiently
transfected with expression vectors for the respective
PPAR chimera along with a reporter construct contain-
ing five copies of the GAL4 DNA binding site driving
expression of secreted placental alkaline phosphatase
(SPAP) and â-galactosidase as described previously.25
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 and â-galactosi-
dase activity. Alkaline phosphatase activity was cor-
rected for transfection efficiency using the â-galactosi-
dase activity as an internal standard. Compounds
which elicited on average at least 70% activation of
PPARγ versus rosiglitazone (positive control) were
considered full agonists.
Biology
In Vitr o. Compounds were evaluated in three sepa-
rate in vitro assays to determine their binding and
functional potency. Compounds were tested for their
ability to bind to PPARγ using a scintillation proximity
assay (SPA). The PPAR ligand binding domain (LBD)
was expressed in Escherichia coli as polyHis-tagged
fusion proteins and purified. The LBD was then labeled
with biotin and immobilized on streptavidin-modified
scintillation proximity beads. The beads were then
incubated with a constant amount of the appropriate
radioligand ([3H]BRL 49653 for PPARγ) and variable
concentrations of test compound, and after equilibration
the radioactivity bound to the beads was measured by
a scintillation counter. The amount of nonspecific
binding, as assessed by control wells containing 50 µM
of the corresponding unlabeled ligand, was subtracted
from each data point. For each compound tested, plots
of ligand concentration versus CPM of radioligand
bound were constructed and apparent Ki values were
estimated from nonlinear least-squares fit of the data
assuming simple competitive binding. The details of
this assay have been reported elsewhere.60
Certain compounds were also tested for their ability
to promote differentiation of C3H10T1/2 stem cells to
adipocytes via measurement of glucose incorporation
into total lipid of the cells (lipogenesis assay). The
amount of glucose uptake into lipid in these cells
provides a measure of the cell differentiation and
indirectly a measure of insulin sensitization mediated
through activation of PPARγ. Several lines of evidence
indicate the importance of the role PPARγ and insulin
play in adipocyte differentiation. First, induction of
PPARγ has been shown to be an early event in the
course of differentiation of several preadipocyte cell lines
and precedes the induction of other adipocyte markers
Compounds were screened for functional potency in
a transient transfection assay in CV-1 cells for their
ability to activate the three PPAR subtypes (transacti-
vation assay). A previously established chimeric recep-

5024 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25
Ta ble 1. In Vitro Profile of PPARγ Agonists
Henke et al.
a
b
See figure. pKi, -log of the concentration of test compound required to achieve an apparent Ki value according to the equation Ki )
IC50/(1 + [L]/Kd), where IC50 is the concentration of test compound required to inhibit 50% of the specific binding of the radioligand, [L]
is the concentration of the radioligand used, and Kd is the dissociation constant for the radioligand at the receptor. c pEC50, -log of the
concentration of test compound required to induce 50% of the maximum alkaline phosphatase activity ( standard error (number of
determinations); IA, inactive at 10-4 M.

N-(2-Benzoylphenyl)-L-tyrosine PPARγ Agonists. 1
Ta ble 2. In Vitro Profile of PPARγ Agonists
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25 5025
a
b
See figure. pKi, -log of the concentration of test compound required to achieve an apparent Ki value according to the equation Ki )
IC50/(1 + [L]/Kd), where IC50 is the concentration of test compound required to inhibit 50% of the specific binding of the radioligand, [L]
is the concentration of the radioligand used, and Kd is the dissociation constant for the radioligand at the receptor. c pEC50, -log of the
concentration of test compound required to induce 50% of the maximum alkaline phosphatase activity ( standard error (number of
d
determinations). pEC50, -log of the concentration of test compound required to induce 50% of the maximum lipogenic activity ( standard
error (number of determinations).
such as aP2 and C/EBPR.6,16 Importantly, this dif-
ferentiation was dependent on the presence of insulin
in the media. Furthermore, ectopic expression of PPARγ
in fibroblasts has been shown to induce adipocyte
differentiation; this differentiation was significantly
enhanced in the presence of weak PPARγ activators.17
In addition, thiazolidinediones have been shown to
induce the insulin-stimulated differentiation of preadi-
pocyte cell lines,61 which suggests that the insulin-
enhancing action of TZDs in these cells is mediated
through the PPARγ receptor.
C3H10T1/2 cells were grown in a DME medium
supplemented with 10% fetal calf serum. One day
before reaching confluence, differentiation was induced
by addition of insulin (200 nM) with or without test
compound. Medium and compound were exchanged
every 3 days. [3H]Glucose was added to the cells after
3 days in culture. Cells treated with compound were
evaluated for lipogenic activity after 7 days by deter-
mination of radioactive glucose uptake into total lipid
and confirmed by accumulation of lipid droplets in the
cytoplasm by staining with oil red O. The details of this
assay have been recently reported.62
In Vivo. Compounds which showed potent PPARγ
activation in vitro were evaluated for in vivo antidiabetic
activity using two established rodent models of type 2
diabetes: the db/db mouse model and/or the Zucker
diabetic fatty (ZDF) rat model.63,64 Both models are
characterized by a single-gene mutation (autosomal
recessive trait) that results in a severe diabetic pheno-
type resembling human type 2 diabetes. The animals
are hyperglycemic, hyperinsulinemic, hyperlipidemic,
and insulin-resistant.
A. Db/d b Mice. Test compound was administered
to 60-day-old male db/db mice (n ) 7/group) at a dose
of 5 mg/kg bid for 14 days by oral gavage. Blood

5026 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25
Ta ble 3. In Vitro Comparison of TZDs and Tyrosine PPARγ Agonists
Henke et al.
a
b
See figure. pKi, -log of the concentration of test compound required to achieve an apparent Ki value according to the equation Ki )
IC50/(1 + [L]/Kd), where IC50 is the concentration of test compound required to inhibit 50% of the specific binding of the radioligand, [L]
is the concentration of the radioligand used, and Kd is the dissociation constant for the radioligand at the receptor. c pEC50, -log of the
concentration of test compound required to induce 50% of the maximum alkaline phosphatase activity ( standard error (number of
determinations).
samples were obtained by cardiac puncture at days 0,
4, 7, and 14 and analyzed for plasma glucose, triglyc-
erides (TGs), nonesterified free fatty acids (NEFAs), and
insulin. A subset of mice was housed in metabolic cages
to obtain baseline food and water consumption, body
weight changes, and urine glucose and output. Values
in Table 6 represent percent reduction from vehicle
control animals at day 14, with the exception of body
weight which is represented as a percent increase over
vehicle controls.
ZDF Ra ts. Male Zucker diabetic fatty (ZDF) rats at
81/2 weeks of age (n ) 6/group) were baseline-matched
for plasma glucose, and test compound was adminis-
tered bid for 14 days by oral gavage. Blood samples
were obtained by cardiac puncture at days 0, 4, 7, and
14 and analyzed for plasma glucose, glycosylated he-
moglobin (HbA1c), which is a surrogate marker for
overall glycemic control, triglycerides (TGs), and non-
esterified free fatty acids (NEFAs). Values in Table 7
represent percent reduction from vehicle control animals
at day 14.
tional assays (transactivation and lipogenesis). Inter-
pretation of the latter assays can be complicated by
differences in the ability of a given compound to perme-
ate cell walls. In addition, a number of factors are
known to influence PPARγ activity besides ligand
binding. These include phosphorylation of PPARγ65 and
the presence of coactivator or corepressor proteins.66 It
is not known how well the in vitro cell-based assays
mimic the in vivo expression of these modifying condi-
tions. Thus there may be a number of reasons why a
molecule with good binding affinity may not have
analogous potency in the functional assays; however,
the converse is rare. To date we have identified no
molecules that are substantially more potent in the
functional assays than in the binding assay. The
functional assays are, nevertheless, extremely impor-
tant in choosing molecules for further progression to in
vivo assays. In particular, in our hands, compounds
which have good potency in the lipogenesis assay also
have good antihyperglycemic and antihyperlipidemic
activity in vivo in rodent models of type 2 diabetes. Since
adipose tissue is the major site for PPARγ expression,
it is not surprising that an assay based on preadipocyte
differentiation to adipocytes, a process in which PPARγ
plays a major role,67 would have good predictive value
for in vivo activity by PPARγ agonists.
Resu lts a n d Discu ssion
In general, structure-activity relationships were
developed and pursued based primarily on binding
affinity to PPARγ rather than on the cell-based func-

N-(2-Benzoylphenyl)-L-tyrosine PPARγ Agonists. 1
Ta ble 4. In Vitro Profile of PPARγ Agonists
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25 5027
a
b
See figure. pKi, -log of the concentration of test compound required to achieve an apparent Ki value according to the equation Ki )
IC50/(1 + [L]/Kd), where IC50 is the concentration of test compound required to inhibit 50% of the specific binding of the radioligand, [L]
is the concentration of the radioligand used, and Kd is the dissociation constant for the radioligand at the receptor; NT, not tested. c pEC50
,
-log of the concentration of test compound required to induce 50% of the maximum alkaline phosphatase activity ( standard error
d
(number of determinations). pEC50, -log of the concentration of test compound required to induce 50% of the maximum lipogenic activity
( standard error (number of determinations); NT, not tested.
Our goal was to develop a potent (<10 nM), subtype-
selective (>1000-fold) PPARγ agonist which (a) did not
contain the 2,4-thiazolidinedione moiety, (b) either was
achiral or could be prepared easily in chiral nonracemic
fashion, and (c) displayed both antihyperglycemic and
antihyperlipidemic activity in an established animal
model of type 2 diabetes when dosed orally. A number
of approaches toward finding an appropriate bioisostere

5028 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25
Henke et al.
Ta ble 5. In Vitro Profile of PPARγ Agonists 18-20
Bindinga
Transactivationb
hPPARγ
pEC50
hPPARR
pEC50
hPPARδ
pEC50
mPPARγ
pEC50
mPPARR mPPARδ
no.
PPARγ pKi
PPARR pKi
PPARδ pKi
pEC50
pEC50
18 8.84 ( 0.09 (2) <5.5 (2)
6.15 ( 0.05 (2) 8.04 ( 0.33 (7) IA
IA
IA
IA
7.88 ( 0.10 (2)
8.33 ( 0.07 (2)
9.45 ( 0.01 (2)
IA
IA
IA
IA
IA
IA
19 8.83 ( 0.14 (2) 5.92 ( 0.40 (2) 6.80 ( 0.00 (2) 8.58 ( 0.39 (8) 5.19 ( 0.60 (6)
20 8.94 ( 0.13 (8) 6.31 ( 0.02 (4) 6.07 ( 0.12 (3) 9.47 ( 0.44 (11) 6.34 ( 0.10 (7)
a
pKi, -log of the concentration of test compound required to achieve an apparent Ki value according to the equation Ki ) IC50/(1 +
[L]/Kd), where IC50 is the concentration of test compound required to inhibit 50% of the specific binding of the radioligand, [L] is the
b
concentration of the radioligand used, and Kd is the dissociation constant for the radioligand at the receptor. pEC50, -log of the
concentration of test compound required to induce 50% of the maximum alkaline phosphatase activity ( standard error (number of
determinations); IA, inactive at 10-4 M.
Ta ble 6. In Vivo Activity of PPARγ Agonists 18-20 in db/db Mice
Antihyperglycemic activitya
Antihyperlipidemic activityb
Plasma Glucose
(% reduction)
Triglycerides
(% reduction)
NEFAs
(% reduction)
Body weightc
(% increase)
no.
18
19
20
58 ( 5
41 ( 11
65 ( 3
61 ( 4
36 ( 9
54 ( 8
46 ( 13
38 ( 7
56 ( 16
6 ( 3
9 ( 1
17 ( 2
a
Antihyperglycemic activity of test compound dosed at 5 mg/kg bid via oral gavage for 14 days; values represent percent reduction as
b
compared to vehicle control group on day 14, ( standard error. Antihyperlipidemic activity of test compound dosed at 5 mg/kg bid via
oral gavage for 14 days; values represent percent reduction as compared to vehicle control group on day 14, ( standard error. c Values
represent percent body weight increase of drug-treated animals as compared to vehicle control group on day 14.
for the 2,4-thiazolidinedione ring have been described
which have met with varying degrees of success,68-72
and several suitable replacements have been report-
ed.73-75 We opted to “mine” the Glaxo Wellcome data-
base in an attempt to identify a structurally novel
PPARγ agonist which did not contain the 2,4-thiazo-
lidinedione ring. Since PPARγ belongs to the steroid/
thyroid family of nuclear receptors, we selected a subset
of compounds from the database which contained the
general size, shape, hydrophobicity, and hydrogen-
bonding characteristics found in known steroid/thyroid
ligands and screened them for in vitro activity. Com-
pound 1 (Chart 1) emerged as the most chemically
tractable from a set of several structurally novel ago-
nists having micromolar potency at human PPARγ.
Replacement of the L-phenylalanine in 1 with O-benzyl-
L-tyrosine afforded compound 2 (Table 1) which dis-
played submicromolar (pKi ) 7.93, pEC50 ) 6.64)
potency at PPARγ in both binding and functional
assays. Continuous iv infusion of 2 in glucose-matched
9-week old ZDF rats for 5 days (Css ) 5 µM) produced a
46% decrease in postprandial plasma glucose levels
compared to vehicle controls. In addition, the serum
levels of triglycerides (TGs) and free fatty acids (FFAs)
were significantly decreased in the treated animals as
compared to vehicle controls. Having established that
a compound selected solely on the basis of interaction
with the PPARγ receptor in vitro displays antidiabetic
activity in vivo, we began to explore the SAR around
compound 2.
fold loss in binding affinity and a compound which had
no measurable agonist activity at concentrations up to
10 µM. Removal of the phenyl enaminone altogether
to afford O-benzyltyrosine 6 also led to a complete
abolishment of receptor affinity and activity. Finally,
ent-2 was only marginally active, suggesting that com-
pounds with the (S)-configuration, derived from natu-
rally occurring L-tyrosine, are more active at the PPARγ
receptor. This result is in agreement with data pub-
lished on a series of antidiabetic R-alkoxy- and R-thio-
phenylpropanoic acids in which both in vitro and in vivo
activity was also shown to reside in the (S)-enanti-
omer.74,76
Despite the promising PPARγ activity of compound
2 and the ease in which analogues could be synthesized,
this compound unfortunately suffered from poor phar-
macokinetic properties. Oral administration of com-
pound 2 in CD rats (5 mg/kg, n ) 4) resulted in both
low bioavailability (%F ) 10%) and a short half-life (t1/2
) 2.3 h). In addition, 2 lacked in vivo efficacy in
lowering glucose and lipids in the db/db mouse assay
when dosed orally, in contrast to the results seen with
iv dosing (vide supra). We felt that the most plausible
reason for the poor pharmacokinetics surrounding 2 was
the potential lability of the enaminone moiety. Sys-
tematic structure-stability studies of enaminones sug-
gest that these compounds are readily cleaved in acid.77
Indeed, the enaminone moiety has been used as an acid-
removable protecting group in peptide synthesis. The
lability of this group might also explain the 20-fold loss
in potency of 2 in the transactivation assay versus the
binding assay (Table 1), as the transactivation assay
requires extended incubation times in cell culture. We
thus concentrated on analogues which would either
stabilize or provide a suitable replacement for the
enaminone moiety of 2. Ring-constrained analogue 7
(Table 1) was completely inactive at the PPARγ recep-
tor. The 2-benzoylcyclohexenyl derivative 8, although
approximately 70-fold less potent than 2, still displayed
weak agonist activity and offered hope that the steric
environment around the enaminone was not too restric-
We briefly explored the region around the phenyl
enaminone and carboxylic acid moieties. Conversion of
the carboxylic acid of 2 to either the primary amide 3
or the methyl ester 4 led to a considerable drop in
binding affinity for the PPARγ receptor (Table 1). Only
a small decrease in potency of activation in the trans-
activation assay was observed for both 3 and 4; however,
we believe the apparent discrepancy is due to conversion
of both the ester and amide to 2 via cellular enzymes
during the incubation time of the assay. Replacement
of phenyl with methyl (compound 5, Table 1) led to 100-

N-(2-Benzoylphenyl)-L-tyrosine PPARγ Agonists. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25 5029
tive. Replacement of the cyclohexenyl ring of 8 with
phenyl afforded benzophenone 9, which was roughly
equipotent to 8. However, benzophenone 9 no longer
contained the acid-labile vinylogous amide moiety and
represented a chemically stable alternative to the
enaminone 2.
tyrosine analogues are 5-50-fold more potent than their
corresponding TZDs in the transactivation assay. These
data suggest that the 2-benzoylphenylamino side chain
is responsible for additional receptor interactions which
the TZDs are unable to achieve. In addition, the rank
order of potencies in both binding and transactivation
assays conferred by a given “left-hand” is the same for
both the TZDs and the N-(2-benzoylphenyl)tyrosine
analogues. Ciglitazone, the least potent TZD, is also
the least potent in our series, while the 2-(5-methyl-2-
phenyloxazol-4-yl)ethyl side chain is the most potent in
both series. This trend supports the hypothesis that the
TZDs and our N-(2-benzoylphenyl)tyrosine series bind
to the PPARγ receptor in a similar fashion.79
We selected the 2-aminobenzoxazole compound 19 as
a template for exploring the SAR around the 2-ami-
nobenzophenone portion of this series in greater detail.
Replacement of the nitrogen atom with carbon, sulfur,
or oxygen (Table 4, entries 1-4) led to a 10-fold decrease
in affinity for PPARγ. An even greater decrease (>100
fold) in functional potency at PPARγ was observed in
both the transactivation and lipogenesis assays, with
the rank order of potency being N . C > O > S. The
origin of the difference between receptor affinity and
functional potency of compounds 21-23 is unclear but
may be due to poorer cellular uptake and/or increased
metabolic lability of these analogues. Interestingly,
conformational constraint of the benzophenone via the
fluorenone derivative 24 led to roughly 100-fold decrease
in both affinity and functional potency at PPARγ (Table
4, entry 1 versus 5), while constraint via the an-
thraquinone-type compound 25 afforded no loss of
affinity or potency at PPARγ (Table 4, entry 1 versus
6). Insertion of a methylene unit between the distal
phenyl ring and the carbonyl group afforded compound
26 which was 10-fold less potent than 19 in both
functional assays, despite showing similar affinity for
the PPARγ receptor. Finally, a short series of replace-
ments for the carbonyl were explored (Table 4, entries
8-11). Replacement of the carbonyl with a sulfone,
ether linkage, hydrazone, or carbinol all led to at least
a 50-fold decrease in both functional potency and
binding affinity versus the corresponding carbonyl
analogues 18 and 19.
A brief investigation into the structure-activity re-
lationships of the benzophenone revealed that this
moiety is fairly sensitive to structural changes. Re-
moval of the ketone to afford compound 10 abolished
all activity at PPARγ. Replacement of the phenyl
ketone portion with either a phenyl amide (11) or a
benzyl ester (12) also resulted in a loss of activity. In
addition, the ortho orientation between the amine and
ketone is important, as the corresponding m- and
p-aminobenzophenones (13 and 14, Table 1) show
reduced activity at PPARγ.
Having established compound 9 as a chemically stable
alternative to 2, we began to explore the “left-hand”
portion of the molecule in an attempt to regain both the
affinity and functional potency at PPARγ. We postu-
lated that if the carboxylic acid moiety of 9 mimicked
the 2,4-thiazolidinedione moiety of the glitazones, then
the most expedient method for improving potency would
be to replace the benzyl group of 9 with “left-hand”
groups known to confer potency in the TZDs. Using this
strategy a rapid improvement in potency at PPARγ was
realized (Table 2). Replacement of the benzyl moiety
with the “left-hand” portion of ciglitazone (compound 15)
resulted in a 50-fold improvement in functional potency
at PPARγ and a modest improvement in binding affinity
over 9. Further increases in PPARγ potency and
affinity were realized upon replacement of the benzyl
moiety with the “left-hand” portions of pioglitazone (16),
troglitazone (17), rosiglitazone (18), and BRL 48482
(19), affording compounds with binding pKi and func-
tional pEC50 values > 7.3. The “left-hand” group which
proved to be the most potent was the 2-(5-methyl-2-
phenyloxazol-4-yl)ethyl side chain first employed by
Takeda.78 Incorporation of this side chain onto the N-(2-
benzoylphenyl)tyrosine framework afforded compound
20, which displayed low-nanomolar binding to PPARγ
and subnanomolar in vitro functional potency. As in
the O-benzyl enaminone series, the enantiomer derived
from L-tyrosine was considerably more potent than its
antipode (compare entries 18 and ent-18). All the
analogues in Table 2 were at least 500-fold selective for
PPARγ over PPARR in both receptor affinity and in vitro
functional agonism (Table 5, data not shown), and none
of the compounds in Table 2 showed activity at PPARδ
at concentrations up to 10 µM (data not shown). It was
also reassuring to note that the rank order of potency
in this series of compounds was maintained across the
binding and transactivation assays, suggesting that this
class of compounds has good cell permeability.
The data in Tables 1 and 4 suggest that the pocket of
the PPARγ protein which binds the 2-aminobenzophe-
none is fairly intolerant of stereoelectronic perturbations
away from the parent 2-aminobenzophenone. We be-
lieve the amine and ketone moieties of the aminoben-
zophenone engage in both a critical internal hydrogen
bond and additional hydrogen bond(s) with residues in
the protein which help place the benzophenone phenyl
rings in an optimal position within the binding pocket.
To evaluate receptor subtype selectivity of the ty-
rosine series, the most potent PPARγ agonists (ana-
logues 18-20) were tested for functional activity at the
three human and murine PPAR subtypes as well as for
binding affinity to PPARR and PPARδ. The data are
shown in Table 5. All three compounds displayed at
least 100-fold binding selectivity for PPARγ versus
PPARR and PPARδ. Compound 19 showed the highest
binding affinity for PPARδ of the three, while compound
20 displayed the most affinity for PPARR. All three
compounds displayed almost identical functional po-
A comparison of the in vitro binding and functional
potency of the N-(2-benzoylphenyl)tyrosine analogues
found in Table 2 along with their corresponding thia-
zolidinediones is found in Table 3. Several interesting
trends emerge from examination of these data. First,
our N-(2-benzoylphenyl)tyrosine analogues show 5-600-
fold greater affinity for hPPARγ than the corresponding
thiazolidinediones. This trend also holds true for
PPARγ functional potency, as the N-(2-benzoylphenyl)-

5030 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25
Henke et al.
Ta ble 7. In Vivo Activity of PPARγ Agonists 18 and 20 in Zucker Diabetic Fatty Rats
Antihyperglycemic activitya
Antihyperlipidemic activityb
Triglycerides NEFAs
Plasma Glucose
(% reduction)
HbA1C
(% reduction)
no.
(% reduction)
(% reduction)
troglitazone
18
20
61 ( 9
68 ( 5
65 ( 5
29 ( 4
37 ( 2
33 ( 4
87 ( 2
87 ( 6
77 ( 5
81 ( 2
73 ( 2
75 ( 4
a
Antihyperglycemic activity of the maximum dose of test compound (5 mg/kg for 18 and 20; 500 mg/kg for troglitazone) given bid via
oral gavage for 14 days; values represent percent reduction as compared to vehicle control group on day 14, ( standard error.
b
Antihyperlipidemic activity of the maximum dose of test compound (5 mg/kg for 18 and 20; 500 mg/kg for troglitazone) given bid via
oral gavage for 14 days; values represent percent reduction as compared to vehicle control group on day 14, ( standard error.
tency at the murine PPARγ receptor versus the human
PPARγ receptor, which is not surprising in view of the
high degree of sequence homology between the two
receptors.18 In addition, all three compounds were
functionally inactive against the murine PPARR and
PPARδ receptors as well as human PPARδ at concen-
trations up to 10 µM. Compound 18 also was inactive
at human PPARR; however, analogues 19 and 20
displayed weak affinity and activity at human PPARR.
The origin of the observed potency difference of our
analogues at murine versus human PPARR is unclear.
The human and murine PPARR receptors show a lower
degree of sequence homology within the ligand binding
domain as compared to PPARγ, and key difference(s)
F igu r e 1. Plasma concentration-response curves for glucose
lowering of compounds 18, 20, and troglitazone in Zucker
in the amino acid sequence could be responsible for
changing the promiscuity in the binding pocket of the
diabetic fatty rats.
receptor. However, differences in phosphorylation states
and/or corepressor and coactivator proteins cannot be
ruled out (vide supra). Nevertheless, it is noteworthy
that 19 and 20 show greater than 1000-fold functional
selectivity for PPARγ over PPARR and PPARδ.
These compounds displayed efficacy similar to trogli-
tazone dosed at 500 mg/kg. Figure 1 shows the plasma
concentration-response curves for troglitazone and
compounds 18 and 20. Both 18 and 20 elicit glucose-
lowering activity similar to troglitazone at plasma
concentrations approximately 100-fold lower, which
parallels their differences in binding and activation
potencies at PPARγ. This provides further evidence
that activity at PPARγ is a predictor of in vivo antihy-
perglycemic activity.28
Compounds 18 and 20 meet the criteria we initially
outlined (vide supra) of being potent and selective
PPARγ agonists which do not contain the 2,4-thiazo-
lidinedione ring. These compounds are easily prepared
in enantiomerically pure form via a convergent ap-
proach, using L-tyrosine as a commerically available
source of chirality. Furthermore, the chemical structure
of these compounds suggests that the facile racemiza-
tion seen with the TZD class of agents is not likely to
occur in this series. Data from our cell-based functional
assays support this hypothesis. The marked difference
in the functional potency of 18 and ent-18 (Table 2) in
both the transactivation and lipogenesis assays, which
mirrors the difference in binding affinity, suggests these
enantiomers do not racemize under the cell culture
conditions of these assays. The in vivo chiral stability
of these analogues will be reported elsewhere.
On the basis of their in vitro PPARγ potency and
selectivity, compounds 18-20 were evaluated for in vivo
antihyperglycemic and antihyperlipidemic efficacy in
the db/db mouse. The results are displayed in Table 6.
Both 18 and 20 showed a marked ability to lower
plasma glucose, serum triglycerides, and nonesterified
free fatty acids (NEFAs) when dosed orally at 5 mg/kg
bid for 14 days. Compound 19 was less efficacious in
all respects than 18 and 20 at the same dose. Since 19
displayed comparable in vitro potency to 18 and 20, the
relatively poor in vivo efficacy is most likely due to
suboptimal pharmacokinetics of this compound. All
three compounds caused statistically significant weight
gain despite showing no increase in food consumption
(data not shown). This increase in weight mirrors that
seen with insulin in rodent models of diabetes and
correlates with the amelioration of glycosuria and the
diabetic state of the animals.80 This effect may be
model-specific, as no significant weight gain has been
seen in human clinical trials with the TZD troglita-
zone.81
Compounds 18 and 20 were also evaluated for in vivo
antihyperglycemic and antihyperlipidemic efficacy in
the ZDF rat. Results are listed in Table 7. Troglitazone
at doses of 50, 150, and 500 mg/kg was used for
comparison in this assay. When dosed orally at 5 mg/
kg bid, both 18 and 20 reduced postprandial plasma
glucose to 30-40% of the vehicle animals after 14 days
of administration. Glycosylated hemoglobin (HbA1c) of
the drug-treated animals was also significantly reduced
compared to vehicle controls. Both compounds also had
marked effects on serum triglycerides and NEFAs.
These N-(2-benzoylphenyl)tyrosine analogues repre-
sent a novel series of antidiabetic agents which were
identified and optimized solely on the basis of in vitro
potency at PPARγ. Concurrent with our work on this
series, researchers at both Pfizer74 and SmithKline
Beecham75 independently reported that R-heteroatom-
substituted carboxylic acids function as effective re-
placements for the 2,4-thiazolidinedione. Although our
series bears some structural resemblance to the afore-

N-(2-Benzoylphenyl)-L-tyrosine PPARγ Agonists. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25 5031
instrument equipped with a Delta-Pak radial compression
cartridge (C-18, 300 A, 15 µm, 47 mm × 300 mm) as the
stationary phase. The mobile phase employed 0.1% aqueous
TFA with acetonitrile as the organic modifier. Linear gradi-
ents were used in all cases, and the flow rate was 100 mL/
min (t0 ) 5 min). Analytical purity was assessed by RP-HPLC
using a Hewlett-Packard series 1050 system equipped with a
diode array spectrometer (λ range 200-400 nm). The station-
ary phase was a Keystone Scientific BDS Hypersil C-18 column
(5 µm, 4.6 mm × 200 mm). The mobile phase employed 0.1%
aqueous TFA with acetonitrile as the organic modifier and a
flow rate of 1.0 mL/min (t0 ) 3 min). Analytical data is
reported as retention time, tR, in minutes (% acetonitrile, time,
flow rate).
1H NMR spectra were recorded on either a Varian VXR-
300, a Varian Unity-400, or a Varian Unity-300 instrument.
Chemical shifts are reported in parts per million (ppm, δ
units). Coupling constants are reported in units of hertz (Hz).
Splitting patterns are designated as s, singlet; d, doublet; t,
triplet; q, quartet; m, multiplet; b, broad. Low-resolution mass
spectra (MS) were recorded on a J EOL J MS-AX505HA, J EOL
SX-102, or SCIEX-APIiii spectrometer. High-resolution mass
spectra were recorded on an AMD-604 (AMD Electra GmbH)
high-resolution double focusing mass spectrometer (Analytical
Instrument Group, Raleigh, NC). Mass spectra were acquired
in the positive ion mode under electrospray ionization (ESI)
or fast atom bombardment (FAB) methods. Combustion
analyses were performed by Atlantic Microlabs, Inc. (Norcross,
GA).
mentioned series, it is worthwhile to note the contrast
in the approaches taken to develop the respective series.
Both the Pfizer and SmithKline groups sought specific
replacements of the 2,4-thiazolidinedione moiety having
already started with TZDs which displayed potent
antidiabetic activity in vivo. Our approach utilized in
vitro screening for a functional PPARγ agonist response
to identify a novel isostere of the 2,4-thiazolidinedione
ring first, and then potency was increased using the
existing thiazolidinedione SAR. The in vivo activity of
analogues 18 and 20 further supports the hypothesis
that activation of PPARγ leads to potent antihypergly-
cemic and antihyperlipidemic effects in the diabetic
state. Although the molecular target of these com-
pounds has been identified, the downstream cellular
mechanisms by which activation of PPARγ-regulated
genes exert their antidiabetic effects remain to be
discovered.82
Con clu sion
We have identified a novel series of antidiabetic N-(2-
benzoylphenyl)-L-tyrosine derivatives which are potent,
selective PPARγ agonists. Starting from screening lead
1, optimization of receptor potency and compound
stability identified the 2-benzoylphenyl moiety as an
optimal substituent on the tyrosine nitrogen. Com-
pounds from this series show increased in vitro potency
at PPARγ when compared to the corresponding thiazo-
lidinedione antidiabetic agents, suggesting that the
2-benzoylphenylamino moiety occupies a previously
undiscovered binding pocket in the PPARγ receptor. In
addition, the â-phenylpropionic acid moiety functions
as a suitable isosteric replacement for the benzyl 2,4-
thiazolidinedione ring. Compounds 18 and 20 from this
series are potent, efficacious antihyperglycemic and
antihyperlipidemic agents when dosed orally in rodent
models of type 2 diabetes. In addition, these analogues
can be prepared in chiral nonracemic fashion from
readily available L-tyrosine and do not show a propen-
sity to undergo racemization in vitro. Additional struc-
ture-activity relationships of these N-(2-benzoylphenyl)-
L-tyrosine antidiabetic agents are detailed in the following
two papers. The increased potency of these PPARγ
agonists versus the currently available TZDs may
translate into superior clinical efficacy of these com-
pounds for the treatment of type 2 diabetes, diabetic
dyslipidemia, and cardiovascular risk management in
diabetic patients.
Gen er a l P r oced u r e for F or m a t ion of Vin ylogou s
Am id es 2, 3, 5, 7, a n d 8. A mixture of 1.0 equiv of O-benzyl-
L-tyrosine (or the corresponding methyl ester or primary
amide), 1.0 equiv of the requisite â-diketone, and 1.1 equiv of
dicyclohexylamine were refluxed in methanol (solution con-
centration approximately 0.1 M) until the reaction was com-
plete as judged by TLC (24-72 h). The reaction mixture was
cooled to room temperature and the solvent removed in vacuo.
The crude reaction was purified either by trituration or by
silica gel flash column chromatography.
3-(4-(Ben zyloxy)p h en yl)-(2S)-((1-m eth yl-3-oxo-3-p h en -
ylp r op en yl)a m in o)p r op ion ic Acid (2). Trituration with
cold (-20 °C) absolute ethanol three times yielded 7.60 g of a
white solid: 1H NMR (DMSO-d6, 400 MHz) δ 11.38 (d, 1 H, J
) 8.9), 7.76 (d, 2 H, J ) 5.8), 7.3 (m, 8 H), 7.09 (d, 2 H, J )
8.5), 6.84 (d, 2 H, J ) 8.5), 5.53 (s, 1 H), 5.00 (s, 2 H), 4.02 (m,
1 H), 3.05 (dd, 1 H, J ) 4.4, 12.9), 2.93 (m, 2 H), 2.76 (dd, 1 H,
J ) 8.4, 13.7), 1.92 (m, 4 H), 1.71 (s, 3 H), 1.66 (m, 4 H), 1.53
(d, 2 H, J ) 12.7), 1.2 (m, 8 H), 1.03 m (2 H); low-resolution
MS (FAB+) m/e 597 (MH+), 182 (DCAH+). Anal. (C26H23
NO4‚(C6H11)2NH) C, H, N.
-
2-Am in o-3-{4-[2-(ben zoxa zol-2-ylm eth yla m in o)eth oxy]-
p h en yl}p r op ion ic Acid Meth yl Ester . To a stirred solution
of 3.61 g (18.5 mmol) of (S)-tyrosine methyl ester and 0.81 g
(20.4 mmol) of sodium hydride (60% suspension in mineral oil)
in 50 mL of DMF at room temperature was added 5.0 g (18.5
mmol) of N-2-(N-methylaminoethyl 1-methylsulfonate)-1,3-
benzoxazole.41 The resulting solution was heated to 100 °C for
2 h. After cooling to room temperature, the solution was
quenched with water and extracted with EtOAc. The com-
bined organics were dried (MgSO4) and solvent removed in
vacuo. The residue was purified by silica gel chromatography
using hexane/EtOAc (gradient of 3:7 to 0:1) as eluent to give
1.45 g (21% yield) of the title compound: 1H NMR (CDCl3, 300
MHz) δ 7.29 (d, 1 H, J ) 7.8), 7.18 (s, 1 H), 7.04 (m, 4 H), 6.76
(d, 2 H, J ) 8.6), 4.18 (t, 2 H, J ) 5.2), 3.88 (t, 2 H, J ) 5.2),
3.64 (s, 3 H), 3.61 (m, 1 H), 3.29 (s, 3 H), 2.95 (m, 1 H), 2.74
(m, 1 H), 1.52 (bs, 2 H); low-resolution MS (ES+) m/e 370
(MH+); TLC (MeOH/EtOAc (1:9)) Rf ) 0.35.
Exp er im en ta l Section
Gen er a l. All commercial chemicals and solvents are re-
agent grade and were used without further purification unless
otherwise specified. The following solvents and reagents have
been abbreviated: tetrahydrofuran (THF), ethyl ether (Et2O),
dimethyl sulfoxide (DMSO), ethyl acetate (EtOAc), dichlo-
romethane (DCM), trifluoroacetic acid (TFA), dimethylforma-
mide (DMF), methanol (MeOH). All reactions except those in
aqueous media were carried out with the use of standard
techniques for the exclusion of moisture. Reactions were
monitored by thin-layer chromatography on 0.25-mm silica gel
plates (60F-254, E. Merck) and visualized with UV light, iodine
vapors, or 5% phosphomolybdic acid in 95% ethanol.
2-Dia zo-3-{4-[2-(ben zoxa zol-2-ylm eth yla m in o)eth oxy]-
p h en yl}p r op ion ic Acid Meth yl Ester . Safety note: Reaction
was performed behind a blast shield. To a stirred solution of
1.45 g of 2-amino-3-{4-[2-(benzoxazol-2-ylmethylamino)ethoxy]-
phenyl}propionic acid methyl ester and 0.7 mL (11.8 mmol)
of glacial acetic acid in 40 mL of chloroform was added 0.5
Final compounds were typically purified either by flash
chromatography on silica gel (E. Merck, 40-63 mm) or by
preparative reverse-phase high-pressure liquid chromatogra-
phy (RP-HPLC) using a Waters model 3000 Delta Prep

5032 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25
Henke et al.
mL (3.93 mmol) of isoamyl nitrite. The resulting solution was
heated to 60 °C for 0.25 h. The solution under went a color
change to orange/brown after heating. The solution was cooled
to room temperature, extracted with water, and then washed
with a saturated solution of sodium bicarbonate. The organics
were then dried (MgSO4) and the solvent removed in vacuo to
quantitatively yield the title compound which was used
directly without further purification: 1H NMR (CDCl3, 400
MHz) δ 7.29 (d, 1 H, J ) 7.7), 7.18 (s, 1 H), 7.08 (m, 3 H), 6.95
(m, 1 H), 6.77 (d, 2 H, J ) 8.5), 4.18 (t, 2 H, J ) 5.2), 3.88 (t,
2 H, J ) 5.3), 3.71 (s, 3 H), 3.50 (s, 2 H), 3.28 (s, 3 H); TLC
(EtOAc/hexane (3:7)) Rf ) 0.23; low-resolution MS (ES+) m/e
381 (MH+), 353.
Gen er a l P r oced u r e for Dia zo In ser tion Rea ction s.
Safety note: This reaction should always be performed behind
a blast shield. To a stirred solution of 1.0 equiv (0.5-5 mmol)
of the diazo ester and 1.1 equiv of the desired aniline in toluene
at room temperature was added 5 mol % of rhodium(II) acetate
dimer. The resulting greenish solution was stirred at room
temperature for 15 min and then warmed to 80 °C until
starting material was consumed as judged by TLC (1-12 h).
The reaction was cooled to room temperature, and the solvent
was removed in vacuo. The crude reaction was then purified
by silica gel flash column chromatography.
3-(4-(Ben zyloxy)p h en yl)-2-((p h en ylca r b a m oyl)p h en -
yla m in o)p r op ion ic Acid Meth yl Ester (11). Purification
by flash chromatography using hexane/EtOAc (5:1) as eluent
afforded 354 mg of the title compound: 1H NMR (CDCl3, 400
MHz) δ 7.83 (s, 1 H), 7.66 (d, 1 H, J ) 7.9), 7.54 (d, 2 H, J )
8.0), 7.47 (d, 1 H, J ) 7.9), 7.35 (m, 8 H), 7.13 (m, 3 H), 6.85
(d, 2 H, J ) 8.6), 6.68 (t, 1 H, J ) 7.5), 6.61 (d, 1 H, J ) 8.4),
4.96 (s, 2 H), 4.30 (q, 1 H, J ) 7.2), 3.64 (s, 3 H), 3.10 (m, 2 H);
low-resolution MS (ES+) m/e 481 (MH+). Anal. (C30H27N2O4‚
0.7H2O) C, H, N.
(2S)-2-((2-Ben zoylp h en yl)a m in o)-3-(4-h yd r oxyp h en yl)-
p r op ion ic Acid Meth yl Ester (31). A stirred mixture of 92
g (0.45 mol) of 2-benzoylcyclohexanone,51 78 g (0.40 mol) of
l-tyrosine methyl ester, and 17.0 g of palladium on activated
carbon (10%) was refluxed for 2 h in 1 L of anisole while the
resulting water was removed by a Dean-Stark apparatus. The
mixture was cooled to 80 °C, and the Pd/C was filtered and
washed with anisole (3 × 50 mL). The mixture was cooled to
40 °C; 1 L of hexane was added and the mixture kept at -20
°C for 48 h. The solid was filtered and washed with hexane
(5 × 200 mL) to yield 89.0 g of crude (S)-2-((2-benzoylphenyl)-
amino)-3-(4-hydroxyphenyl)propionic acid methyl ester. This
solid was mixed with 220 mL of methanol, and the slurry was
refluxed for 30 min. The mixture was cooled to 0 °C; the
product was filtered, washed with 50 mL of cold (-20 °C)
methanol twice, and then dried under reduced pressure to yield
67.4 g the of the title compound: mp 185-6 °C; 1H NMR
(DMSO-d6, 200 MHz) δ 9.27 (s, 1 H), 8.68 (d, 1 H, J ) 7.8),
7.69-7.32 (m, 7 H), 7.00 (d, 2 H, J ) 8.4), 6.82 (d, 1 H, J )
8.4), 6.65 (m, 3 H), 4.64 (m, 1 H), 3.67 (s, 3 H), 3.05 (m, 2 H);
low-resolution MS (ES+) m/e 376 (MH+). Anal. (C23H21NO4)
C, H, N.
1 H, J ) 8.4), 6.58 (t, 1 H, J ) 7.8), 4.37 (m, 1 H),
4.02 (m, 2 H), 3.70 (s, 3 H), 3.56 (m, 2 H), 3.16 (m, 2
H), 2.95 (s, 3 H), 1.44 (s, 9 H); low-resolution MS (ES+)
m/e 555 (MNa+), 533 (MH+); TLC (hexane/EtOAc (2:1)):
Rf ) 0.49. Anal. (C31H36N2O6) C, H, N.
(2S)-((2-Ben zoylph en yl)am in o)-3-{4-[2-(5-m eth yl-2-ph en -
yloxa zol-4-yl)eth oxy]p h en yl}p r op ion ic Acid Meth yl Es-
ter . Purification by silica gel flash column chromatography
using hexane/EtOAc (7:3): as eluent afforded 0.26 g (70%) of
the title compound as a yellow foam: mp 55-60 °C; 1H NMR
(DMSO-d6, 200 MHz) δ 8.66 (d, 1 H, J ) 7.8), 7.91 (m, 2 H),
7.65-7.32 (m, 10 H), 7.12 (d, 2 H, J ) 8.4), 6.84 (d, 3 H, J )
8.6), 6.64 (t, 1 H, J ) 7.5), 4.68 (m, 1 H), 4.16 (t, 2 H, J ) 6.7),
3.66 (s, 3 H), 3.11 (m, 2 H), 2.90 (t, 2 H, J ) 6.6), 2.34 (s, 3 H);
low-resolution MS (ES+) m/e 561 (MH+); TLC (hexane/EtOAc
(7:3)) Rf ) 0.50. Anal. (C35H32N2O5‚0.2H2O) C, H, N.
2(S)-((2-Ben zoylp h en yl)a m in o)-3-{4-[2-(m eth ylp yr id in -
2-yla m in o)eth oxy]p h en yl}p r op ion ic Acid Meth yl Ester .
A solution of 2.56 g (4.81 mmol) of 2(S)-((2-benzoylphenyl)-
amino)-3-{4-[2-((tert-butoxycarbonyl)methylamino)ethoxy]-
phenyl}propionic acid methyl ester in 56 mL of DCM at 25 °C
was treated with 56 mL (0.73 mol, 152 equiv) of TFA. After
stirring for 30 min, the solution was neutralized with saturated
NaHCO3 and extracted with DCM (2 × 50 mL). The combined
organics were dried (Na2SO4), filtered, and concentrated in
vacuo. The crude amine was used immediately in the next
reaction. A solution of 2.08 g (4.81 mmol) of the crude amine
from above in 480 mL of 2-fluoropyridine was allowed to reflux
for 16 h and then concentrated in vacuo. Purification by silica
gel flash column chromatography using hexane/EtOAc (2:1)
as eluent provided 1.85 g (76%) of the title compound as a
yellow oil: 1H NMR (CDCl3, 400 MHz) δ 8.89 (d, 1 H, J ) 7.3),
8.14 (dd, 1 H, J ) 1.6, 4.8), 7.59 (d, 2 H, J ) 7.0), 7.45 (m, 5
H), 7.31 (dd, 1 H, J ) 7.2, 7.2), 7.16 (d, 2 H, J ) 8.6), 6.81 (d,
2 H, J ) 8.6), 6.57 (m, 4 H), 4.37 (dd, 1 H, J ) 6.0, 6.0), 4.14
(t, 2 H, J ) 5.6), 3.95 (t, 2 H, J ) 5.6), 3.68 (s, 3 H), 3.12 (m,
5 H); low-resolution MS (CI) m/e 511 (MH+), 510 (M+); TLC
(hexane/EtOAc (2:1)) Rf ) 0.40. Anal. (C31H31N3O4) C, H, N.
3-{4-[2-(Ben zoxa zol-2-ylm eth yla m in o)eth oxy]p h en yl}-
2(S)-((2-ben zoylp h en yl)a m in o)pr op ion ic Acid Meth yl Es-
ter . A solution of 319 mg (0.60 mmol) of 2(S)-((2-benzoylphen-
yl)amino)-3-{4-[2-((tert-butoxycarbonyl)methylamino)ethoxy]-
phenyl}propionic acid methyl ester in 7 mL of DCM at 25 °C
was treated with 7 mL (90.9 mmol, 152 equiv) of TFA. After
stirring for 30 min, the solution was neutralized with saturated
NaHCO3 and extracted with DCM (2 × 50 mL). The combined
organics were dried (Na2SO4), filtered, and concentrated in
vacuo. The crude amine was used immediately in the next
reaction. To a solution of 259 mg (0.60 mmol) of the above
amine in 6 mL of THF at 25 °C was added 0.250 mL (1.80
mmol, 3 equiv) of Et3N followed by 0.103 mL (0.90 mmol, 1.5
equiv) of 2-chlorobenzoxazole. After stirring for 24 h, the
reaction was diluted with EtOAc, poured into saturated
NaHCO3, and extracted with EtOAc. The combined organics
were dried (Na2SO4), filtered, and concentrated in vacuo.
Purification by silica gel flash column chromatography eluting
with hexane/EtOAc (2:1 to 1:1) provided 244 mg (74%) of the
title compound as a yellow solid: 1H NMR (CDCl3, 300 MHz)
δ 8.88 (d, 1 H, J ) 6.9), 7.62-7.55 (m, 2 H), 7.54-7.10 (m, 10
H), 7.00 (t, 1 H, J ) 7.7), 6.81 (d, 2 H, J ) 8.4), 6.62 (d, 1 H,
J ) 8.7), 6.57 (t, 1 H, J ) 7.5), 4.37 (m, 1 H), 4.22 (t, 2 H, J )
5.2), 3.93 (t, 1 H, J ) 5.2), 3.68 (s, 3 H), 3.33 (s, 3 H), 3.15 (m,
2 H); low-resolution MS (ES+) m/e 572 (MNa+), 550 (MH+);
TLC (hexane/EtOAc (2:1)) Rf ) 0.24. Anal. (C33H31N3O5) C,
H, N.
Gen er a l P r oced u r e for Mitsu n obu Cou p lin g w ith
P h en ol (31). A stirring solution of 1.0 equiv of (2S)-((2-
benzoylphenyl)amino)-3-(4-hydroxyphenyl)propionic acid meth-
yl ester, 1.1 equiv of the requisite alcohol, and 1.5 equiv of
triphenylphosphine in THF (0.1 M total solution concentration)
at room temperature was treated with 1.5 equiv of diethyl
azodicarboxylate in 5-50 mL of THF via dropwise addition.
The resulting solution was stirred for 6-18 h at room tem-
perature, and then the solvent was removed in vacuo. The
crude reaction mixture was purified as described below.
(2S )-((2-B e n z o y lp h e n y l)a m i n o )-3-{4-[2-(t er t -b u -
t oxyca r b on yl)m et h yla m in o)et h oxy]p h en yl}p r op ion ic
Acid Meth yl Ester . Purification by silica gel flash column
chromatography using hexane/EtOAc (2:1) as eluent
afforded 1.37 g (65%) of title compound as a viscous yellow
oil: 1H NMR (CDCl3, 300 MHz) δ 8.91 (bs, 1 H), 7.59
(m, 2 H), 7.56-7.40 (m, 4 H), 7.33 (t, 1 H, J ) 7.8),
7.19 (d, 2 H, J ) 8.6), 6.81 (d, 2 H, J ) 8.6), 6.63 (d,
3-{4-[2-(Ben zoxa zol-2-ylm eth yla m in o)eth oxy]p h en yl}-
2-[(2-(ben zoylp h en yl)th io]p r op ion ic Acid Meth yl Ester .
To a solution of 360 mg (1.68 mmol) of 2-mercaptobenzophe-
none and 18-crown-6 (50 mg) in 5 mL of THF was added
dropwise 1.55 mL (1.55 mmol) of 1.0 M potassium tert-butoxide
in THF. After 5 min, the resulting red solution was concen-
trated by rotary evaporation to give a red foam which was
redissolved in THF (5 mL). To this solution was added a
solution of 560 mg (1.29 mmol) of 2-bromo-3-{4-[2-(benzoxazol-

N-(2-Benzoylphenyl)-L-tyrosine PPARγ Agonists. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25 5033
2-ylmethylamino)ethoxy]phenyl}propionic acid methyl ester57
in 2 mL of THF. The resulting orange solution was stirred
for 3 h at room temperature and then diluted with EtOAc (75
mL). This mixture was washed with 2.0 M NaOH (4 × 25
mL) and brine (25 mL), dried over MgSO4, and concentrated
to a brown oil. This material was flash chromatographed on
silica gel (75 g) and the product eluted with hexane/EtOAc (2:
1) to afford 600 mg of the title compound as a white foam: 1H
NMR (CDCl3, 400 MHz) δ 7.74 (d, 2 H, J ) 7.2), 7.56 (t, 2 H,
J ) 6.8), 7.45-7.35 (m, 6 H), 7.16 (t, 1 H, J ) 7.6), 7.00 (m, 3
H), 6.74 (d, 2 H, J ) 8.6), 4.21 (t, 2 H, J ) 5.3), 3.93 (t, 2 H,
J ) 5.1), 3.83 (dd, 1 H, J ) 9.7, 6.0), 3.52 (s, 3 H), 3.34 (s, 3
H), 3.01 (dd, 1 H, J ) 9.9, 14.1), 2.86 (dd, 1 H, J ) 6.0, 13.8);
low-resolution MS (ES+) m/e 567 (MH+).
(m, 1 H), 3.30-3.10 (m, 2 H); low-resolution MS (FAB+) m/e
452 (MH+). Anal. (C29H25NO4) C, H, N.
(2S)-((2-Ben zoylp h en yl)a m in o)-3-{4-[2-(m eth ylp yr id in -
2-yla m in o)eth oxy]p h en yl}p r op ion ic Acid (18). Tritura-
tion of the crude reaction mixture with EtOAc followed by
filtration and drying gave 1.24 g of the title compound as a
yellow solid: 1H NMR (DMSO-d6, 300 MHz) δ 13.04 (s, 1 H),
8.71 (d, 1 H, J ) 7.5), 8.11 (dd, 1 H, J ) 1.2, 5.1), 7.69-7.52
(m, 6 H), 7.48 (t, 1 H, J ) 7.8), 7.41 (dd, 1 H, J ) 1.8, 8.0),
7.15 (d, 2 H, J ) 8.7), 7.73-6.81 (m, 3 H), 6.75 (d, 1 H, J )
8.7), 6.71-6.60 (m, 2 H), 4.59 (m, 1 H), 4.12 (t, 2 H, J ) 5.9),
3.93 (t, 2 H, J ) 5.9), 3.15 (m, 2 H), 3.11 (s, 3 H); low-resolution
MS (CI) m/e 518 (MNa+), 496 (MH+); TLC (EtOAc/AcOH (99:
1)) Rf ) 0.13; SFC (Chiralpak AD, 4.6 mm × 25 cm; 75% CO2/
25% MeOH with 0.1% DEA; 35 min; 1 mL/min; 3000 psi, 40
°C) tR ) 24.3 min (t0) 3 min); >99% ee. Anal. (C30H29N3O4‚
0.8H2O) C, H, N.
3-{4-[2-(Ben zoxa zol-2-ylm eth yla m in o)eth oxy]p h en yl}-
(2S)-((2-ben zoylp h en yl)a m in o)p r op ion ic Acid (19). Tri-
turation of the crude reaction with Et2O/hexane followed by
filtration and drying provided 209 mg of the title compound
as a yellow solid: 1H NMR (DMSO-d6, 300 MHz) δ 13.1 (bs, 1
H), 8.71 (d, 1 H, J ) 7.5), 7.68-7.36 (m, 8 H), 7.32 (d, 1 H, J
) 7.2), 7.23-7.12 (m, 3 H), 7.03 (dt, 1 H, J ) 2.1, 7.7), 6.92-
6.82 (m, 3 H), 6.66 (t, 1 H, J ) 7.5), 4.58 (m, 1 H), 4.23 (t, 2 H,
J ) 5.4), 3.91 (t, 2 H, J ) 5.4), 3.25 (s, 3 H), 3.15 (m, 2 H);
low-resolution MS (CI) m/e 558 (MNa+), 536 (MH+); TLC
(EtOAc/MeOH (9:1)) Rf ) 0.23; HPLC (Chiralcel OD-H, 4.6 mm
3-{4-[2-(Ben zoxa zol-2-ylm eth yla m in o)eth oxy]p h en yl}-
2-[(2-ben zoylp h en yl)m eth yl]p r op ion ic Acid Meth yl Es-
ter . To a solution of 1.56 g (4.42 mmol) of 3-{4-[2-(benzoxazol-
2-ylmethylamino)ethoxy}phenyl}propionic acid methyl ester
in 75 mL of THF maintained at -78 °C was added dropwise
6.62 mL of 1.0 M sodium hexamethyldisilazide in THF. The
resulting pale-yellow solution was stirred at -78 °C for 90 min,
and then 1.97 g (6.62 mmol) of 2-iodobenzyl bromide was
added. The solution was stirred for an additional 90 min at
-78 °C and was then quenched by the addition of saturated
NH4Cl (50 mL). This mixture was warmed to room temper-
ature and extracted with EtOAc (200 mL). This extract was
washed with saturated NH4Cl and brine (50 mL each), dried
over MgSO4, and concentrated to a pale-yellow oil which was
flash chromatographed on silica gel. The product 2-(2-iodo-
benzyl)propionic acid methyl ester was eluted with hexane/
EtOAc (3:1) and isolated as a colorless oil (0.48 g, 19%): 1H
NMR (CDCl3, 400 MHz) δ 7.79 (d, 1 H, J ) 7.9), 7.36 (d, 1 H,
J ) 7.7), 7.24-7.14 (m, 4 H), 7.08 (d, 2 H, J ) 8.5), 7.00 (t, 1
H, J ) 7.7), 6.88 (t, 1 H, J ) 7.7), 6.80 (d, 2 H, J ) 7.7), 4.23
(t, 2 H, J ) 5.3), 3.94 (t, 2 H, J ) 5.2), 3.46 (s, 3 H), 3.34 (s, 3
H), 3.04-2.76 (m, 5 H); low-resolution MS (ES+) m/e 571
(MH+).
× 25 cm; 23%EtOH/hexane/0.1%TFA; 45 min; 1 mL/min) tR
)
9.3 min (t0) 3 min); >98% ee. Anal. (C32H29N3O5‚1.1H2O) C,
H, N.
(2S)-((2-Ben zoylph en yl)am in o)-3-{4-[2-(5-m eth yl-2-ph en -
yloxa zol-4-yl)eth oxy]p h en yl}p r op ion ic Acid (20). Tritu-
ration of the crude reaction with Et2O/hexane followed by
filtration and drying provided 202 mg of the title compound:
mp 148-50 °C; 1H NMR (DMSO-d6, 200 MHz) δ 8.66 (d, 1 H,
J ) 7.8), 7.91 (m, 2 H), 7.64-7.33 (m, 10 H), 7.13 (d, 2 H, J )
8.3), 6.82 (m, 3 H), 6.62 (t, 1 H, J ) 7.5), 4.54 (m, 1 H), 4.15 (t,
2 H, J ) 6.5), 3.10 (m, 2 H), 2.90 (t, 2 H, J ) 6.5), 2.34 (s, 3
H); low-resolution MS (ES+) m/e 547 (MH+); TLC (DCM/MeOH
(95:5)) Rf ) 0.49. Anal. (C34H30N2O5‚0.3H2O) C, H, N.
A solution of the above iodide (213 mg, 0.37 mmol), PdCl2-
(PPh3)2 (8 mg, 0.011 mmol), phenylboronic acid (50 mg, 0.41
mmol), and K2CO3 (153 mg, 1.12 mmol) in 8 mL of 1,4-dioxane
was warmed to 95 °C and stirred under 1 atm of CO(g) for 22
h. The resulting tan suspension was diluted with EtOAc (75
mL), washed with water (10 mL) and brine (20 mL, 2×), dried
over MgSO4, and concentrated to a yellow oil which was flash
chromatographed on silica gel eluted with hexane/EtOAc (7:
3) to afford 144 mg of the title compound as a pale-yellow oil:
1H NMR (CDCl3, 400 MHz) δ 7.76 (d, 2 H, J ) 7.20), 7.56 (t,
1 H, J ) 7.8), 7.45-7.24 (m, 8 H), 7.16 (t, 1 H, J ) 7.6), 7.00
(t, 1 H, J ) 7.7), 6.94 (d, 2 H, J ) 8.6), 6.72 (d, 2 H, J ) 8.6),
4.21 (t, 2 H, J ) 5.3), 3.93 (t, 2 H, J ) 5.2), 3.39 (s, 3 H), 3.34
(s, 3H), 2.95 (m, 3 H), 2.82 (dd, 1 H, J ) 8.0,13.9), 2.66 (dd, 1
H, J ) 5.4,13.8); 13C NMR (CDCl3, 100 MHz) δ 198.22, 175.23,
162.45, 156.95, 149.00, 143.42, 138.52, 138.46, 137.75, 133.19,
131.36, 130.78, 130.35, 130.25, 129.84, 129.16, 128.40, 125.89,
123.96, 120.36, 116.08, 114.28, 108.71, 66.11, 51.3, 50.06,
49.73, 37.82, 37.38, 35.70; low-resolution MS (ES+) m/e 549
(MH+).
(2S)-3-{4-[2-(1,3-Ben zoxa zol-2-ylm eth yla m in o)eth oxy]-
ph en yl}-2-(2-ph en oxyan ilin o)pr opion ic Acid Sodiu m Salt
(28). A stirred solution of 210 mg (0.40 mmol) of (2S)-3-{4-
[2-(1,3-benzoxazol-2-ylmethylamino)ethoxy]phenyl}-2-(2-phe-
noxyanilino)propionic acid methyl ester in 5 mL of THF was
treated with 0.60 mL (0.60 mmol) of a 1.0 M solution of LiOH
in H2O. The resulting solution was stirred at room temper-
ature for 17 h. The reaction mixture was then poured into 15
mL of 1 N HCl and extracted with EtOAc (2 × 15 mL). The
organics were separated and dried (MgSO4) and the solvents
removed in vacuo. The resulting tan foam was dissolved in 2
mL of dry methanol, and 1 equiv of NaH was added. The
resulting solution was stirred at room temperature for 1 h;
then the solvent was removed in vacuo. Purification of the
residue by trituration with ether/DCM (1:1) followed by
filtration and drying (house vacuum, 50 °C) afforded 170 mg
of the sodium salt of the title compound as a white solid: mp
1
Gen er a l P r oced u r e for Ester Hyd r olysis To F or m
Com p ou n d s 9, 10, a n d 13-27. A stirred solution of 1.0 equiv
of the appropriate methyl ester in THF/MeOH (3:1) was
treated with 1.5 equiv of a 1.0 M solution of LiOH in H2O.
The resulting solution was stirred at room temperature until
reaction was complete as judged by TLC (4-24 h). The
reaction mixture was then poured into 1 N HCl and extracted
with EtOAc. The organic layer was separated and dried
(MgSO4), and the solvents were removed in vacuo. The crude
reaction was then purified as outlined below.
210-4 °C; H NMR (DMSO-d6, 400 MHz) δ 7.37 (d, 1 H, J )
7.9), 7.12 (dd, 1 H, J ) 7.5, 7.5), 7.00-6.89 (m, 5 H), 6.79 (d,
2 H, J ) 8.0), 6.69 (d, 1 H, J ) 7.1), 6.61 (d, 2 H, J ) 8.6), 6.50
(d, 1 H, J ) 8.0), 6.42 (dd, 1 H, J ) 7.7, 7.7), 5.42 (d, 1 H, J )
5.7), 4.12 (t, 2 H, J ) 5.4), 3.83 (t, 2 H, J ) 5.4), 3.64 (m, 1 H),
3.19 (s, 3 H), 2.98 (dd, 1H, J ) 5.1, 14.5), 2.86 (dd, 1 H, J )
4.6, 14.5); low-resolution MS (ES+) m/e 569 (MNa+), 546 (M+).
Anal. (C31H28N3O5Na‚2.0H2O) C, H, N.
(2S )-2-[2-(H y d r o x y p h e n y lm e t h y l)a n ilin o ]-3-{4-[2-
(m eth ylpyr idin -2-ylam in o)eth oxy]ph en yl}pr opion ic Acid
(29). To a stirred solution of 5 mg (0.13 mmol) of NaBH4 in 1
mL of absolute ethanol was added 50 mg (0.10 mmol) of 18 in
1 mL of absolute ethanol. The resulting solution was stirred
30 min at room temperature, then an additional 5 mg NaBH4
was added, and the solution was stirred an additional 30 min.
The reaction was quenched by addition of 0.1 N HCl and then
2-((2-Ben zoylp h en yl)a m in o)-3-(4-(ben zyloxy)p h en yl)-
p r op ion ic a cid (9). Purification by silica gel flash column
chromatography using DCM/MeOH (9:1) afforded 133 mg of
the title compound: 1H NMR (CDCl3, 300 MHz) δ 8.87 (bs, 1
H), 7.60 (d, 2 H, J ) 6.0), 7.94-7.32 m (10 H), 7.23 (d, 2 H, J
) 3.9), 6.90 (m, 2 H), 6.67-6.56 (m, 2 H), 4.98 (s, 2 H), 4.39

5034 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25
Henke et al.
(8) Schoonjans, K.; Watanabe, M.; Suzuki, H.; Mahfoudi, A.; Krey,
G.; Wahli, W.; Grimaldi, P.; Staels, B.; Yamamoto, T.; Auwerx,
J . Induction of the acyl-coenzyme A synthetase gene by fibrates
and fatty acids is mediated by a peroxisome proliferator response
element in the C promoter. J . Biol. Chem. 1995, 270, 19269-
76.
(9) Schoonjans, K.; Peinado-Onsurbe, J .; Lefebvre, A.-M.; Heyman,
R. A.; Briggs, M.; Deeb, S.; Staels, B.; Auwerx, J . PPARR and
PPARγ activators direct a distinct tissue-specific transcriptional
response via a PPRE in the lipoprotein lipase gene. EMBO J .
1996, 15, 5336-48.
poured into brine (15 mL) and extracted with EtOAc (2 × 15
mL). The organic layers were combined and dried (MgSO4),
and solvent was removed in vacuo. Purification of the cream-
colored foam via silica gel flash column chromatography using
CHCl3/MeOH (9:1) as eluent afforded 25 mg of the title
compound as a pale-yellow solid: 1H NMR (DMSO-d6, 400
MHz) δ 8.06 (d, 1 H, J ) 3.3), 7.46 (m, 1 H), 7.28 (d, 2 H, J )
7.3), 7.21 (dd, 2 H, J ) 7.1, 7.1), 7.12 (m, 1 H), 7.02 (d, 1 H, J
) 7.1), 6.92 (m, 3 H), 6.64 (m, 3 H), 6.54 (dd, 1 H, J ) 5.0,
6.8), 6.46 (dd, 1 H, J ) 7.3, 7.3), 6.40 (d, 1 H, J ) 8.0), 5.58 (s,
1 H), 4.04 (t, 2 H, J ) 6.0), 3.86 (t, 2 H, J ) 6.0), 3.72 (m, 1 H),
3.05 (s, 3 H), 2.90 (dd, 1 H, J ) 4.8, 13.5), 2.69 (m, 1 H); low-
resolution MS (ES) m/e 499 (MH+), 498 (M+), 481, 480, 436.
Anal. (C30H31N3O4‚HCl) C, H, N.
(10) Schoonjans, K.; Martin, G.; Staels, B.; Auwerx, J . Peroxisome
proliferator-activated receptors, orphans with ligands and func-
tions. Curr. Opin. Lipidol. 1997, 8, 159-66.
(11) Willson, T. M.; Wahli, W. Peroxisome proliferator-activated
receptor agonists. Curr. Opin. Chem. Biol. 1997, 1, 235-41.
(12) Staels, B.; Schoonjans, K.; Fruchart, J . C.; Auwerx, J . The effects
of fibrates and thiazolidinediones on plasma triglyceride me-
tabolism are mediated by distinct peroxisome proliferator acti-
vated receptors (PPARs) Biochimie 1997, 79, 95-9.
(13) Mandrup, S.; Lane, M. D. Regulating adipogenesis. J . Biol.
Chem. 1997, 272, 5367-70.
(14) Latruffe, N.; Vamecq, J . Peroxisome proliferators and peroxisome
proliferator activated receptors (PPARs) as regulators of lipid
metabolism. Biochimie 1997, 79, 81-94.
(15) Devchand, P. R.; Wahli, W. PPARR: tempting fate with fat.
Horm. Signaling 1998, 1, 235-56.
(16) Chawla, A.; Schwarz, E. J .; Dimaculangan, D. D.; Lazar, M. A.
Peroxisome proliferator-activated receptor (PPAR)γ: adipose-
predominant expression and induction early in adipocyte dif-
ferentiation. Endocrinology 1994, 135, 798-800.
(17) Tontonoz, P.; Hu, E.; Spiegelman, B. M. Stimulation of adipo-
genesis in fibroblasts by PPARγ2, a lipid-activated transcription
factor. Cell 1994, 79, 1147-56.
(18) Fajas, L.; Auboeuf, D.; Raspe, E.; Schoonjans, K.; Lefebvre, A.-
M.; Saladin, R.; Najib, J .; Laville, M.; Fruchart, J .-C.; Deeb, S.;
Vidal-Puig, A.; Flier, J .; Briggs, M. R.; Staels, B.; Vidal, H.;
Auwerx, J . The organization, promoter analysis, and expression
of the human PPARγ gene. J . Biol. Chem. 1997, 272, 18779-
789.
(2S)-2-[2-(H yd r a zon op h en ylm et h yl)a n ilin o]-3-{4-[2-
(m eth ylpyr idin -2-ylam in o)eth oxy]ph en yl}pr opion ic Acid
(30). A solution of 150 mg (0.30 mmol) of 18 in 5 mL of
anhydrous hydrazine was heated to 80 °C for 9 h. The reaction
was then cooled to room temperature, poured into EtOAc/
water, and acidified to pH 2 by addition of 1 N HCl. The layers
were separated; then the aqueous layer was adjusted to pH 6
by addition of aqueous NaHCO3. The aqueous layer was
extracted with DCM (2 × 15 mL). The organics were dried
(MgSO4), and the DCM was removed in vacuo to afford a tan
solid. Purification via trituration with ether/hexane (1:1)
1
afforded 48 mg of a cream-colored solid: H NMR (DMSO-d6,
400 MHz, 1:1 mixture of hydrazone isomers) δ 8.94 (d, 0.5 H,
J ) 7.4), 8.06 (m, 1 H), 7.55-7.45 (m, 2 H), 7.32-7.10 (m, 5
H), 7.03-6.23 (m, 12.5 H), 5.87 (m, br, 0.5 H), 4.28 (m, 1 H),
4.17 (m, 0.5 H), 4.05 (m, 1.5 H), 3.95 (m, 0.5 H), 3.86 (m, 2 H),
3.10-2.82 (m, 5 H); low-resolution MS (ES+) m/e 511 (MH+),
510 (M+). Anal. (C30H31N5O3) C, H, N.
(19) Xue, J .-C.; Schwarz, E. J .; Chawla, A.; Lazar, M. A. Distinct
stages in adipogenesis revealed by retinoid inhibition of dif-
ferentiation after induction of PPARγ. Mol. Cell. Biol. 1996, 16,
1567-75.
(20) Yu, K.; Bayona, W.; Kallne, C. B.; Harding, H. P.; Ravera, C.
P.; McMahon, G.; Brown, M.; Lazar, M. A. Differential activation
of peroxisome proliferator-activated receptors by eicosanoids. J .
Biol. Chem. 1995, 270, 23975-83.
Ack n ow led gm en t. We gratefully acknowledge Ro-
derick Davis, Doug Minick, Peter Kitrinos, Larry Shamp-
ine, and Manon Villeneuve for analytical support. We
also thank Thomas Consler and Bruce Wisely for PPAR
protein expression and purification.
(21) Kliewer, S. A.; Lenhard, J . M.; Willson, T. M.; Patel, I.; Morris,
D. C.; Lehmann, J . A prostaglandin J 2 metabolite binds per-
oxisome proliferator-activated receptor γ and promotes adipocyte
differentiation. Cell 1995, 83, 813-9.
(22) Forman, B. M.; Tontonoz, P.; Chen, J .; Brun, R. P.; Spiegelman,
B. M.; Evans, R. M. 15-Deoxy-∆12,14-prostaglandin J 2 is a
ligand for the adipocyte determination factor PPARγ. Cell 1995,
83, 803-12.
Su p p or tin g In for m a tion Ava ila ble: Table of physical
data for compounds 2-30 and experimental data for all
compounds which were not included in the Experimental
Section (13 pages). Ordering information is given on any
current masthead page.
(23) Krey, G.; Braissant, O.; L′Horset, F.; Kalkhoven, E.; Perroud,
M.; Parker, M. G.; Wahli, W. Fatty acids eicosanoids, and
hypolipidemic agents identified as ligands of peroxisome prolif-
erator-activated receptors by coactivator-dependent receptor
ligand assay. Mol. Endocrinol. 1997, 11, 779-91.
(24) Ricote, M.; Li, A. C.; Willson, T. M.; Kelly, C. J .; Glass, C. K.
The peroxisome proliferator-activated receptor-γ is a negative
regulator of macrophage activation. Nature 1998, 391, 79-82.
(25) Lehmann, J . M.; Moore, L. B.; Simth-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, 12953-6.
(26) Lambe, K. G.; Tugwood, J . D. A human peroxisome-proliferator-
activated receptor-γ is activated by inducers of adipogenesis,
including thiazolidinedione drugs. Eur. J . Biochem. 1996, 239,
1-7.
Refer en ces
(1) Greene, M. E.; Blumberg, B.; McBride, O. W.; Yi, H. F.;
Kronquist, K.; Kwan, K.; Hsieh, L.; Greene, G.; Nimer, S. D.
Isolation of the human peroxisome proliferator activated receptor
gamma cDNA: expression in hematopoietic cells and chromo-
somal mapping. Gene Express. 1995, 4, 281-99.
(2) Issemann, I.; Green, S. Activation of a member of the steroid
hormone receptor superfamily by peroxisome proliferators.
Nature 1990, 347, 645-50.
(3) Kliewer, S. A.; Forman, B. M.; Blumberg, B.; Ong, E. S.;
Borgmeyer, U.; Mangelsdorf, D. J .; Umesono, K.; Evans, R. M.
Differential expression and activation of a family of murine
peroxisome proliferator-activated receptors. Proc. Natl. Acad.
Sci. U.S.A. 1994, 91, 7355-9.
(4) 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.
(27) Berger, J .; Bailey, P.; Biswas, C.; Cullinan, C. A.; Doebber, T.
W.; Hayes, N. S.; Saperstein, R.; Smith, R. G.; Leibowitz, M. D.
Thiazolidinediones produce a conformational change in peroxi-
somal proliferator-activated receptor-γ: binding and activation
correlate with antidiabetic actions in db/db mice. Endocrinology
1996, 137, 4189-95.
(5) Zhu, Y.; Alvares, K.; Huang, Q.; Rao, M. S.; Reddy, J . K. Cloning
of
a new member of the peroxisome proliferator-activated
(28) 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. The Structure-Activity Relationship between
Peroxisome Proliferator-Activated Receptor γ. Agonism and the
Anti-Hyperglycemic Activity of Thiazolidinediones. J . Med.
Chem. 1996, 39, 665-8.
receptor gene family from mouse liver. J . Biol. Chem. 1993, 268,
26817-20.
(6) Tontonoz, P.; Hu, E.; Graves, R. A.; Budavari, A. I.; Spiegelman,
B. M. mPPARγ2: tissue-specific regulator of an adipocyte
enhancer. Genes Dev. 1994, 8, 1224-34.
(7) Tontonoz, P.; Hu, E.; Devine, J .; Beale, E. G.; Spiegelman, B.
M. PPAR-γ2 regulates adipose expression of the phospho-
enolpyruvate carboxykinase gene. Mol. Cell. Biol. 1995, 15, 351-
7.
(29) Spencer, C. M.; Markham, A. Troglitazone. Drugs 1997, 54, 89-
101.
(30) Grossman, S. L.; Lessem, J . Mechanisms and clinical effects of
thiazolidinediones. Exp. Opin. Invest. Drugs 1997, 6, 1025-40.

N-(2-Benzoylphenyl)-L-tyrosine PPARγ Agonists. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25 5035
(31) Leutenegger, M.; Sacca, L.; Alderton, C.; Eckland, D.; Lettis, S.
Double-masked, placebo-controlled, dose-ranging study of tro-
glitazone 10 to 200 mg once daily in non-insulin-dependent
diabetes mellitus. Curr. Ther. Res. 1997, 58, 403-416.
(32) Levin, K.; Thye-Roenn, P.; Foot, E.; Beck-Nielsen, H. Metabolic
effects of troglitazone in NIDDM patients. Endocrinol. Metab.
1997, 4, 255-64.
(33) Sironi, A. M.; Vichi, S.; Gastaldelli, A.; Pecori, N.; Anichini, R.;
Foot, E.; Seghieri, G.; Ferrannini, E. Effects of troglitazone on
insulin action and cardiovascular risk factors in patients with
noninsulin-dependent diabetes. Clin. Pharmacol. Ther. 1997, 62,
194-202.
(34) Fujiwara, T.; Yoshioka, S.; Yoshioka, T.; Ushiyama, I.; Horikoshi,
H. Characterization of new oral antidiabetic agent CS-045.
Studies in KK and ob/ob mice and Zucker fatty rats. Diabetes
1988, 37, 1549-58.
(35) Chaiken, R. L.; Eckert-Norton, M.; Pasmantier, R.; Boden, G.;
Ryan, I.; Gelfand, R. A.; Lebovitz, H. E. Metabolic effects of
darglitazone, an insulin sensitizer, in NIDDM subjects. Diabe-
tologia 1995, 38, 1307-12.
(36) Colca, J . R.; Morton, D. R. In Antihyperglycemic thiazolidinedi-
ones: ciglitazone and its analogues; Colca, J . R., Morton, D. R.,
Eds.; Smith-Gordon: New York, 1990; pp 255-261.
(37) Yoshioka, T.; Fujita, T.; Kanai, T.; Aizawa, Y.; Kurumada, T.;
Hasegawa, K.; Horikoshi, H. Studies on hindered phenols and
analogues. 1. Hypolipidemic and hypoglycemic agents with
ability to inhibit lipid peroxidation. J . Med. Chem. 1989, 32,
421-8.
(38) Ikeda, H.; Taketomi, S.; Sugiyama, Y.; Shimura, Y.; Sohda, T.;
Meguro, K.; Fujita, T. Effects of pioglitazone on glucose and lipid
metabolism in normal and insulin resistant animals. Arzneim.-
Forsch. 1990, 40, 156-62.
(39) Sohda, T.; Momose, Y.; Meguro, K.; Kawamatsu, Y.; Sugiyama,
Y.; Ikeda, H. Studies on antidiabetic agents. Synthesis and
hypoglycemic activity of 5-[4-(pyridylalkoxy)benzyl]-2,4-thiazo-
lidinediones. Arzneim.-Forsch. 1990, 40, 37-42.
(40) Momose, Y.; Meguro, K.; Ikeda, H.; Hatanaka, C.; Oi, S.; Sohda,
T. Studies on antidiabetic agents. X. Synthesis and biological
activities of pioglitazone and related compounds. Chem. Pharm.
Bull. 1991, 39, 1440-5.
(41) 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, 3977-85.
(42) Cantello, B. C. C.; Cawthorne, M. A.; Haigh, D.; Hindley, R. M.;
Smith, S. A.; Thurlby, P. The synthesis of BRL 49653 - a novel
and potent antihyperglycemic agent. Bioorg. Med. Chem. Lett.
1994, 4, 1181-4.
(43) Horikoshi, H.; Yoshioka, T.; Kawasaki, T.; Nakamura, K.-i.;
Matsunuma, N.; Yamaguchi, K.; Sasahara, K. Troglitazone (CS-
045), a New Antidiabetic Drug. Annu. Rep. Sankyo Res. Lab.
1994, 46, 1-57.
(44) Fujiwara, T.; Yoshioka, S.; Yoshioka, T.; Ushiyama, I.; Horikoshi,
H. Characterization of New Oral Antidiabetic Agent CS-045.
Diabetes 1988, 37, 1549-58.
(45) Henry, R. R. Thiazolidinediones. Endocrinol. Metab. Clin. North
Am. 1997, 26, 553-73.
(46) Sohda, T.; Mizuno, K.; Kawamatsu, Y. Studies on antidiabetic
agents. VI. Asymmetric transformation of (()-5-[4-(1-methylcy-
clohexylmethoxy)benzyl]-2,4-thiazolidinedione(ciglitazone) with
optically active 1-phenylethylamines. Chem. Pharm. Bull. 1984,
32, 4460-5.
(47) Stork, G.; Brizzolara, A.; Landesman, H.; Szmuszkovicz, J .;
Terrell, R. The Enamine Alkylation and Acylation of Carbonyl
Compounds. J . Am. Chem. Soc. 1963, 85, 207-22.
(48) Westheimer, F. H.; Taguchi, K. Catalysis by molecular sieves
in the preparation of ketimines and enamines. J . Org. Chem.
1971, 36, 1570-2.
(54) Doyle, M. P.; Austin, R. E.; Bailey, A. S.; Dwyer, M. P.; Dyatkin,
A. B.; Kalinin, A. V.; Kwan, M. M. Y.; Liras, S.; Oalmann, C. J .;
et al. Enantioselective Intramolecular Cyclopropanations of
Allylic and Homoallylic Diazoacetates and Diazoacetamides
Using Chiral Dirhodium(II) Carboxamide Catalysts. J . Am.
Chem. Soc. 1995, 117, 5763-75.
(55) Doyle, M. P.; Zhou, Q. L.; H., S. S.; Lynch, V. Dirhodium(II)
tetrakis[alkyl 2-oxaazetidine-4(S)-carboxylates]. A new set of
effective chiral catalysts for asymmetric intermolecular cyclo-
propanation reactions with diazoacetates. Synlett 1996, 697-8.
(56) Bode, J . W.; Doyle, M. P.; Protopopova, M. N.; Zhou, Q.-L.
Intramolecular Regioselective Insertion into Unactivated Prochiral
Carbon-Hydrogen Bonds with Diazoacetates of Primary Alco-
hols Catalyzed by Chiral Dirhodium(II) Carboxamidates. Highly
Enantioselective Total Synthesis of Natural Lignan Lactones.
J . Org. Chem. 1996, 61, 9146-55.
(57) Haigh, D. Heterocyclic compounds and their use in the treatment
of type 2 diabetes. PCT Int. Appl., 1993.
(58) Brieaddy, L. E.; Donaldson, K. H. The synthesis of cis-6,12-
diaryl-6H,12H-dibenzo[b,f][1,5]dithiocin-6,12-imine and cis-6,12-
diphenyl-6,12-epoxy-6H,12H-dibenzo[b,f][1,5]dithiocin. J . Het-
erocycl. Chem. 1995, 32, 1683-6.
(59) Ishiyama, T.; Kizaki, H.; Miyaura, N.; Suzuki, A. Synthesis of
unsymmetrical biaryl ketones via palladium-catalyzed carbon-
ylative cross-coupling reaction of arylboronic acids with iodoare-
nes. Tetrahedron Lett. 1993, 34, 7595-8.
(60) Nichols, J . S.; Parks, D. J .; Consler, T. G.; Blanchard, S. G.
Development of a Scintillation Proximity Assay for Peroxisome
Proliferator-Activated Receptor gamma Ligand Binding Domain.
Anal. Biochem. 1998, 257, 112-9.
(61) Hiragun, A.; Sato, M.; Mitsui, H. Preadipocyte differentiation
in vitro: identification of a highly active adipogenic agent. J .
Cell. Physiol. 1988, 134, 124-30.
(62) Lenhard, J . M.; Kliewer, S. A.; Paulik, M. A.; Plunket, K. D.;
Lehmann, J . M.; Weiel, J . E. Effects of troglitazone and met-
formin on glucose and lipid metabolism. Alterations of two
distinct molecular pathways. Biochem. Pharmacol. 1997, 54,
801-8.
(63) Shafrir, E. Animal models of noninsulin-dependent diabetes.
Diabetes/ Metab. Rev. 1992, 8, 179-208.
(64) Bailey, C. J .; Flatt, P. R. In Animal Models of Diabetes; Bailey,
C. J ., Flatt, P. R., Eds.; Churchill Livingstone: Edinburgh, 1986;
Vol. 2, pp 71-89.
(65) (a) Hu, E.; Kim, J . B.; Sarraf, P.; Spiegelman, B. M. Inhibition
of adipogenesis through MAP kinase-mediated phosphorylation
of PPARγ. Science 1996, 274, 2100-3. (b) Adams, M.; Reginato,
M. J .; Shao, D.; Lazar, M. A.; Chatterjee, V. K. Transcriptional
activation by peroxisome proliferator-activated receptor γ is
inhibited by phosphorylation at a consensus mitogen-activated
protein kinase site. J . Biol. Chem. 1997, 272, 5128-32. (c)
Zhang, B.; Berger, J .; Zhou, G.; Elbrecht, A.; Biswas, S.; White-
Carrington, S.; Szalkowski, D.; Moller, D. E. Insulin- and
mitogen-activated protein kinase-mediated phosphorylation and
activation of peroxisome proliferator-activated receptor γ. J . Biol.
Chem. 1996, 271, 31771-4. (d) Camp, H. S.; Tafuri, S. R.
Regulation of peroxisome proliferator-activated receptor γ activ-
ity by mitogen-activated protein kinase. J . Biol. Chem. 1997,
272, 10811-16.
(66) (a) Puigserver, P.; Wu, Z.; Park, C. W.; Graves, R.; Wright, M.;
Spiegelman, B. M. A cold-inducible coactivator of nuclear recep-
tors linked to adaptive thermogenesis. Cell 1998, 92, 829-9. (b)
Zhu, Y.; Qi, C.; Calandra, C.; Rao, M.; Sambasiva, R.; J anardan,
K. Cloning and identification of mouse steroid receptor coacti-
vator-1 (mSRC-1), as a coactivator of peroxisome proliferator-
activated receptor γ. Gene Express. 1996, 6, 185-95. (c) Direnzo,
J .; Soderstrom, M.; Kurokawa, R.; Ogliastro, M.-H.; Ricote, M.;
Ingrey, S.; Horlein, A.; Rosenfeld, M. G.; Glass, C. K. Peroxisome
proliferator-activated receptors and retinoic acid receptors dif-
ferentially control the interactions of retinoid X receptor het-
erodimers with ligands, coactivators, and corepressors. Mol. Cell.
Biol. 1997, 17, 2166-76.
(67) Brun, R. P.; Kim, J . B.; Hu, E.; Spiegelman, B. M. Peroxisome
proliferator-activated receptor gamma and the control of adipo-
genesis. Curr. Opin. Lipidol. 1997, 8, 212-8.
(68) Ellingboe, J . W.; Alessi, T. R.; Dolak, T. M.; Nguyen, T. T.; Tomer,
J . D.; Guzzo, F.; Bagli, J . F.; McCaleb, M. L. Antihyperglycemic
activity of novel substituted 3H-1,2,3,5-oxathiadiazole 2-oxides.
J . Med. Chem. 1992, 35, 1176-83.
(69) Ellingboe, J . W.; Lombardo, L. J .; Alessi, T. R.; Nguyen, T. T.;
Guzzo, F.; Guinosso, C. J .; Bullington, J .; Browne, E. N. C.; Bagli,
J . F. Antihyperglycemic activity of novel naphthalenylmethyl-
3H-1,2,3,5-oxathiadiazole 2-oxides. J . Med. Chem. 1993, 36,
2485-93.
(70) Sohda, T.; Mizuno, K.; Tawada, H.; Sugiyama, Y.; Fujita, T.;
Kawamatsu, Y. Studies on antidiabetic agents. I. Synthesis of
5-[4-(2-methyl-2-phenylpropoxy)benzyl]thiazolidine-2,4-dione (AL-
321) and related compounds. Chem. Pharm. Bull. 1982, 30,
3563-73.
(49) Adams, J .; Spero, D. M. Rhodium(II) catalyzed reactions of diazo-
carbonyl compounds. Tetrahedron 1991, 47, 1765-808.
(50) Kawamatsu, Y.; Asakawa, H.; Saraie, T.; Imamiya, E.; Nish-
ikawa, K.; Hamuro, Y. Studies on antihyperlipidemic agents.
II. Synthesis and biological activities of 2-chloro-3-arylpropionic
acids. Arzneim.-Forsch. 1980, 30, 585-9.
(51) Denny, W. A.; Cain, B. F. Potential antitumor agents. 27.
Quantitative structure-antileukemic (L1210) activity relation-
ships for the ω[4-(9-acridinylamino)phenyl]alkanoic acids. J .
Med. Chem. 1978, 21, 430-7.
(52) Hulin, B.; Clark, D. A.; Goldstein, S. W.; McDermott, R. E.;
Dambek, P. J .; Kappeler, W. H.; Lamphere, C. H.; Lewis, D. M.;
Rizzi, J . P. Novel thiazolidine-2,4-diones as potent euglycemic
agents. J . Med. Chem. 1992, 35, 1853-64.
(53) Horner, M.; Nissen, A. 2-Hydroxyalkylchromans. Ger. Offen.,
1981.

5036 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 25
Henke et al.
(71) Kees, K. L.; Caggiano, T. J .; Steiner, K. E.; Fitzgerald, J . J ., J r.;
Kates, M. J .; Christos, T. E.; Kulishoff, J . M.; Moore, R. D.;
McCaleb, M. L. Studies on New Acidic Azoles as Glucose-
Lowering Agents in Obese, Diabetic db/db Mice. J . Med. Chem.
1995, 38, 617-28.
(72) Kees, K. L.; Fitzgerald, J . J ., J r.; Steiner, K. E.; Mattes, J . F.;
Mihan, B.; Tosi, T.; Mondoro, D.; McCaleb, M. L. New Potent
Antihyperglycemic Agents in db/db Mice: Synthesis and Struc-
ture-Activity Relationship Studies of (4-Substituted benzyl)-
(trifluoromethyl)pyrazoles and -pyrazolones. J . Med. Chem.
1996, 39, 3920-8.
(73) Dow, R. L.; Bechle, B. M.; Chou, T. T.; Clark, D. A.; Hulin, B.;
Stevenson, R. W. Benzyloxazolidine-2,4-diones as potent hy-
poglycemic agents. J . Med. Chem. 1991, 34, 1538-44.
(77) (a) Naringrekar, V. H.; Stella, V. J . Mechanism of hydrolysis
and structure-stability relationship of enaminones as potential
prodrugs of model primary amines. J . Pharm. Sci. 1990, 79,
138-46. (b) Larsen, J . D.; Bundgaard, H. Prodrugs as drug
delivery systems. 49. Hydrolysis kinetics of enaminones derived
from various amino acids and 1,3-diketones, keto-esters and
keto-amides. Arch. Pharm. Chem., Sci. Ed. 1986, 14, 52-63.
(78) Sohda, T.; Mizuno, K.; Momose, Y.; Ikeda, H.; Fujita, T.; Meguro,
K. Studies on antidiabetic agents. 11. Novel thiazolidinedione
derivatives as potent hypoglycemic and hypolipidemic agents.
J . Med. Chem. 1992, 35, 2617-26.
(79) X-ray crystallographic analysis of ligands from each series
cocrystallized with the ligand binding domain of hPPARγ
indicates that the carboxylic acid moiety of the N-(2-benzoylphen-
yl)tyrosines and the 2,4-thiazolidinedione moiety of the TZDs
interact with the same portion of the ligand binding domain in
the hPPARγ receptor, with the benzophenone occupying a pocket
which remains empty in the PPARγ-TZD complex (M. Milburn,
unpublished results from these laboratories).
(80) Brown, K. K.; Henke, B. R.; Blanchard, S. G.; Cobb, J . E.; Kaldor,
I.; Kliewer, S. A.; Lehmann, J . M.; Lenhard, J . M.; Harrington,
W.; Novak, P. J .; Binz, J .; Hashim, M. A.; Oliver, W. O.; Brown,
H. R.; Parks, D. J .; Plunket, K. D.; Tong, T.; Menius, A.; Adkison,
K. K.; Noble, S. A.; Willson, T. M. Novel N-aryl tyrosine
activators of PPARγ reverse the diabetic phenotype of the zucker
diabetic fatty rat. Diabetes, submitted for publication.
(81) Kumar, S.; Boulton, A. J . M.; Beck-Nielsen, H.; Berthezene, F.;
Muggeo, M.; Persson, B.; Spinas, G. A.; Donoghue, S.; Lettis,
S.; Stewart-Long, P. Troglitazone, an insulin action enhancer,
improves metabolic control in NIDDM patients. Diabetologia
1996, 39, 701-709.
(74) Hulin, B.; Newton, L. S.; Lewis, D. M.; Genereux, P. E.; Gibbs,
E. M.; Clark, D. A. Hypoglycemic Activity of a Series of
R-Alkylthio and R-Alkoxy Carboxylic Acids Related to Ciglita-
zone. J . Med. Chem. 1996, 39, 3897-907.
(75) (a) Buckle, D. R.; Cantello, B. C. C.; Cawthorne, M. A.; Coyle,
P. J .; Dean, D. K.; Faller, A.; Haigh, D.; Hindley, R. M.; J efcott,
L. J .; Lister, C. A.; Pinto, I. L.; Rami, H. K.; Smith, D. G.; Smith,
S. A. Nonthiazolidinedione antihyperglycemic agents. 1: R-Het-
eroatom substituted â-phenylpropanoic acids. Bioorg. Med.
Chem. Lett. 1996, 6, 2121-6. (b) Buckle, D. R.; Cantello, B. C.
C.; Cawthorne, M. A.; Coyle, P. J .; Dean, D. K.; Faller, A.; Haigh,
D.; Hindley, R. M.; Lefcott, L. J .; Lister, C. A.; Pinto, I. L.; Rami,
H. K.; Smith, D. G.; Smith, S. A. Nonthiazolidinedione antihy-
perglycemic agents. 2: R-Carbon substituted â-phenylpropanoic
acids. Bioorg. Med. Chem. Lett. 1996, 6, 2127-30.
(76) Young, P. W.; Buckle, D. R.; Cantello, B. C. C.; Chapman, H.;
Clapham, J . C.; Coyle, P. J .; Haigh, D.; Hindley, R. M.; Holder,
J . C.; Kallender, H.; Latter, A. J .; Lawrie, K. W. M.; Mossa-
kowska, D.; Murphy, G. J .; Cox, L. R.; Smith, S. A. Identification
of high-affinity binding sites for the insulin sensitizer Rosigli-
tazone (BRL-49653) in rodent and human adipocytes using a
radioiodinated ligand for Peroxisomal Proliferator-Activated
Receptor γ. J . Pharmacol. Exp. Ther. 1998, 284, 751-9.
(82) For an excellent discussion on the possible downstream mech-
anism of action of PPARγ activators, see: Spiegelman, B. M.
PPAR-γ: Adipogenic regulator and thiazolidinedione receptor.
Diabetes 1998, 47, 507-4.
J M9804127